The geographies of community disaster resilience

Abstract

There is increasing policy and research interest in disaster resilience, yet the extant literature is still mired in definitional debates, epistemological orientations of researchers, and differences in basic approaches to measurement. As a consequence, there is little integration across domains and disciplines on community resilience assessment, its driving forces, and geographic variability. Using US counties as the study unit, this paper creates an empirically-based resilience metric called the Baseline Resilience Indicators for Communities (BRIC) that is both conceptually and theoretically sound, yet, easy enough to compute for use in a policy context. A common set of variables were used to measure the inherent resilience of counties in the United States according to six different domains or capitals as identified in the extant literature – social, economic, housing and infrastructure, institutional, community, and environmental. Data were from public and freely accessible data sources. Counties in the US Midwest and Great Plains states have the most inherent resilience, while counties in the west, along the US-Mexico border, and along the Appalachian ridge in the east contain the least resilience. Further, it was found that inherent resilience is not the opposite of social vulnerability, but a distinctly different construct both conceptually and empirically. While understanding the overall variability in resilience, the BRIC is easily deconstructed to its component parts to provide guidance to policy makers on where investments in intervention strategies may make a difference in the improvement of scores. Such evidence-based research has an opportunity to influence public policy focused on disaster risk.

Full text

The
geographies
of
community
disaster
resilience
§
Susan
L.
Cutter *,
Kevin
D.
Ash,
Christopher
T.
Emrich
University
of
South
Carolina,
USA
1.
Introduction
Resilience,
especially
the
concept
of
community
resilience
is
becoming
the
de
facto
framework
for
enhancing
community-level
disaster
preparedness,
response,
and
recovery
in
the
short
term,
and
climate
change
adaptation
in
the
longer
term.
The
enhance-
ment
of
disaster
resilience
is
a
topic
of
significant
importance
as
evidenced
by
recent
high
level
reports
in
the
United
States
(US
NRC,
2012),
the
United
Kingdom
(UK
Foresight,
2012),
and
for
the
United
Nations
(UNISDR,
2012).
While
not
agreeing
on
a
precise
definition
of
disaster
resilience,
all
three
reports
provide
a
consensus
view
that
disaster
resilience
enhances
the
ability
of
a
community
to
prepare
and
plan
for,
absorb,
recover
from,
and
more
successfully
adapt
to
actual
or
potential
adverse
events
in
a
timely
and
efficient
manner
including
the
restoration
and
improvement
of
basic
functions
and
structures.
In
its
original
ecological
context,
the
notion
of
bouncing
back
to
the
pre-impact
state
defined
resilience,
but
in
the
disaster
context,
this
has
been
expanded
to
include
measures
of
betterment
–
bouncing
forward,
not
merely
just
bouncing
back
(Manyena
et
al.,
2011).
Conceptual
models
of
disaster
resilience
are
plentiful,
ranging
from
those
that
consider
resilience
as
a
set
of
networked
capacities
(Norris
et
al.,
2008;
Sherrieb
et
al.,
2010,
2012),
to
those
that
consider
it
as
a
set
of
distinct
capitals
such
as
economic
or
social
capital
(Alawiyah
et
al.,
2011;
Aldrich,
2012;
Ritchie
and
Gill,
2007),
community
capital
(Miles
and
Chang,
2011)
or
attributes
of
a
particular
system
such
as
infrastructure
(Flynn,
2007),
the
economy
(Rose,
2007),
or
governance
(Tierney,
2012).
Place-based
models
of
resilience
also
appear
in
the
literature
(Cutter
et
al.,
2008a;
Zhou
et
al.,
2010;
Frazier
et
al.,
2013).
Resilience
as
a
core
construct
has
been
applied
to
different
types
of
human
environ-
ments
such
as
urban
areas
(see
Leichenko,
2011
for
an
excellent
summary)
and
rural
areas
(McManus
et
al.,
2012),
and
to
different
thematic
areas
such
as
climate
change
(O‘Brien
et
al.,
2009)
and
sustainability
science
(Turner,
2010).
There
has
been
work
on
disaster
resilience
for
specific
cities
(Pelling,
2003;
Vale
and
Campanella,
2004),
coastal
regions
(Adger,
2005),
and
for
Global
Environmental
Change
29
(2014)
65–77
A
R
T
I
C
L
E
I
N
F
O
Article
history:
Received
25
March
2014
Received
in
revised
form
10
July
2014
Accepted
26
August
2014
Available
online
Keywords:
Resilience
Indicators
SoVI
US
BRIC
A
B
S
T
R
A
C
T
There
is
increasing
policy
and
research
interest
in
disaster
resilience,
yet
the
extant
literature
is
still
mired
in
definitional
debates,
epistemological
orientations
of
researchers,
and
differences
in
basic
approaches
to
measurement.
As
a
consequence,
there
is
little
integration
across
domains
and
disciplines
on
community
resilience
assessment,
its
driving
forces,
and
geographic
variability.
Using
US
counties
as
the
study
unit,
this
paper
creates
an
empirically-based
resilience
metric
called
the
Baseline
Resilience
Indicators
for
Communities
(BRIC)
that
is
both
conceptually
and
theoretically
sound,
yet,
easy
enough
to
compute
for
use
in
a
policy
context.
A
common
set
of
variables
were
used
to
measure
the
inherent
resilience
of
counties
in
the
United
States
according
to
six
different
domains
or
capitals
as
identified
in
the
extant
literature
–
social,
economic,
housing
and
infrastructure,
institutional,
community,
and
environmental.
Data
were
from
public
and
freely
accessible
data
sources.
Counties
in
the
US
Midwest
and
Great
Plains
states
have
the
most
inherent
resilience,
while
counties
in
the
west,
along
the
US-Mexico
border,
and
along
the
Appalachian
ridge
in
the
east
contain
the
least
resilience.
Further,
it
was
found
that
inherent
resilience
is
not
the
opposite
of
social
vulnerability,
but
a
distinctly
different
construct
both
conceptually
and
empirically.
While
understanding
the
overall
variability
in
resilience,
the
BRIC
is
easily
deconstructed
to
its
component
parts
to
provide
guidance
to
policy
makers
on
where
investments
in
intervention
strategies
may
make
a
difference
in
the
improvement
of
scores.
Such
evidence-based
research
has
an
opportunity
to
influence
public
policy
focused
on
disaster
risk.
ß
2014
Elsevier
Ltd.
All
rights
reserved.
§
This
research
was
supported
by
the
U.S.
National
Science
Foundation
under
Grant
SES-1132755.
Any
opinions,
findings,
and
conclusions
are
those
of
the
authors
who
take
full
responsibility
for
the
paper’s
content.
*Corresponding
author
at:
Hazards
&
Vulnerability
Research
Institute,
Depart-
ment
of
Geography,
University
of
South
Carolina,
Columbia,
SC
29208,
USA.
Tel.:
+1
803
777
1590;
fax:
+1
803
777
4972.
E-mail
addresses:
scutter@sc.edu
(S.L.
Cutter),
ashkd@email.sc.edu
(K.D.
Ash),
emrich@mailbox.sc.edu
(C.T.
Emrich).
Contents
lists
available
at
ScienceDirect
Global
Environmental
Change
jo
ur
n
al
h
o
mep
ag
e:
www
.elsevier
.co
m
/loc
ate/g
lo
envc
h
a
http://dx.doi.org/10.1016/j.gloenvcha.2014.08.005
0959-3780/ß
2014
Elsevier
Ltd.
All
rights
reserved.

particular
threat
sources
like
earthquakes
(Bruneau
et
al.,
2003;
Whitman
et
al.,
2013),
brushfires
(Paton
and
Tedim,
2012),
and
hurricanes
(Frazier
et
al.,
2013).
Despite
this
robust
literature,
there
is
still
considerable
disagreement
about
the
characteristics
defining
disaster
resilience,
the
analytical
frameworks
most
useful
for
measuring
it,
and
the
target
of
reference
for
the
resilience
(a
person,
an
individual
structure,
households,
infrastructure,
or
broader
based
systems).
For
each,
there
are
differences
in
the
methods,
models
and
metrics
employed
based
on
the
epistemological
orientations
of
the
researchers.
As
a
result,
there
is
little
integration
of
approaches
in
assessing
community
disaster
resilience
or
its
geographic
variability
from
place
to
place.
In
particular,
conceptual
orienta-
tions
on
the
linkages
between
vulnerability,
resilience,
and
adaptive
capacity
are
quite
diverse
and
depend
on
whether
viewed
from
the
socio-ecological
systems
frameworks
inherent
in
global
change
research
or
from
an
environmental
hazards
perspective.
For
example,
Turner
et
al.
(2003)
looking
from
a
sustainability
science
orientation
view
resilience
as
a
subset
of
vulnerability.
This
paper,
on
the
other
hand
takes
a
hazards
perspective
which
views
vulnerability
and
resilience
as
separate,
but
linked
concepts
with
some
overlap.
It
follows
the
theoretical
framework
of
the
disaster
resilience
of
place
(DROP)
model
(Cutter
et
al.,
2008a).
In
that
model,
human
systems,
environmental
systems,
and
the
built
environment
interact
to
produce
antecedent
conditions
which
contain
both
inherent
vulnerabilities
as
well
as
inherent
resilience.
The
inherent
vulnerabilities
are
those
attri-
butes
of
populations
that
influence
their
ability
to
prepare
for,
respond
to,
and
recovery
from
disasters.
Inherent
resilience
refers
to
qualities
of
a
community,
stemming
from
everyday
processes,
that
might
enhance
or
detract
from
its
ability
to
prepare
for,
respond
to,
recover
from
and
mitigate
environmental
hazard
events.
These
inherent
resilience
properties
are
assumed
to
be
in
place
prior
to
the
onset
of
a
disruptive
hazard
event.
Furthermore,
for
purposes
of
quantification
we
use
a
static
snapshot
of
inherent
resilience,
recognizing
that
the
production
of
resilient
community
character-
istics
is
dynamic
and
can
vary
on
an
annual,
monthly,
weekly,
daily,
or
even
hourly
basis.
Because
we
are
interested
in
the
antecedent
conditions,
adaptive
disaster
resilience
is
not
explicitly
addressed
in
the
indicator
construct
presented
in
this
paper,
although
it
does
appear
in
the
DROP
model.
Finally,
as
our
focus
is
on
resilience
in
a
disaster-specific
setting,
we
do
not
purport
to
include
all
aspects
or
attributes
of
coupled
human-environment
systems,
only
those
most
relevant
to
the
disaster
context.
The
role
of
disaster
resilience
in
the
enhancement
of
risk
reduction
also
is
not
well
understood
(Cutter
et
al.,
2008a;
Miller
et
al.,
2010;
Zhou
et
al.,
2010).
For
example,
there
is
no
single
or
widely
accepted
method
to
quantitatively
characterize
the
level
of
disaster
resilience
that
exists
in
communities
prior
to
an
adverse
event
and
how
this
pre-existing
or
inherent
resilience
relates
to
pre-event
levels
of
social
vulnerability
or
capabilities
to
respond.
Nor
do
we
fully
understand
how
to
estimate
the
capacity
that
exists
within
a
community
to
increase
or
enhance
disaster
resilience
during
or
in
the
aftermath
of
a
disruptive
hazard
event
(adaptive
resilience).
The
inability,
thus
far,
to
reconcile
frame-
works
of
disaster
resilience
and
social
vulnerability
with
available
quantitative
indicators
hampers
the
ability
to
improve
communi-
ty-level
disaster
resilience.
Composite
quantitative
measures
are
needed
that
would
permit
examination
and/or
comparison
among
places
as
to
their
present
levels
of
both
social
vulnerability
and
disaster
resilience,
pointing
decision-makers
to
possible
targets
for
intervention
and
improvement
(Barnett
et
al.,
2008;
Cutter
et
al.,
2008a).
Several
quantitative
vulnerability
indicators
have
been
devel-
oped,
however
quantitative
methods
for
constructing
resilience
metrics
are
in
their
infancy,
and
it
remains
unclear
whether
such
indices
can
meaningfully
capture
the
outcomes
or
processes
of
disaster
resilience
(Prior
and
Hagmann,
2014).
Previous
attempts
at
developing
resilience
metrics
start
from
different
conceptual
orientations
(Cutter
et
al.,
2010;
Miles
and
Chang,
2011;
Peacock
et
al.,
2010;
Renschler
et
al.,
2010;
Sherrieb
et
al.,
2010)
and
when
applied,
they
generally
focus
on
local
to
regional
study
areas.
While
building
a
diverse
catalog
of
case
studies
is
a
necessary
step
in
the
resilience
indicators
research
process,
gaps
persist
in
the
observational
knowledge
about
how
disaster
resilience
varies
from
place
to
place
within
a
larger
political
geographic
unit
such
as
the
United
States
(US).
Why
are
some
communities
in
the
US
more
resilient
to
disasters
than
others?
How
can
we
explain
differences
in
disaster
resilience
between
and
among
places?
Is
it
related
to
the
relative
impact
of
disaster
losses,
social
vulnerability,
historic
experience,
or
some
other
factor?
These
voids
in
the
resilience
indicators
literature
lead
us
to
ask
three
broad
questions
that
provide
the
focus
for
this
paper.
First,
is
there
a
common
set
of
resilience
indicators
that
can
be
applied
across
the
US
to
assess
similarities
and
differences
in
the
inherent
disaster
resilience?
Second,
what
factors
drive
inherent
disaster
resilience
at
the
county-level
and
are
there
distinct
spatial
patterns
to
these
drivers?
Finally,
what
are
the
similarities
in
our
indicators
of
inherent
social
vulnerability
and
disaster
resilience?
Are
they
simply
the
obverse
of
one
another,
is
there
some
overlap,
or
are
they
distinctly
different
constructs?
Our
goal
is
to
extend
the
metrics-based
approach
for
inherent
community
disaster
resilience
(Cutter
et
al.,
2010)
to
include
a
more
comprehensive
suite
of
variables
and
a
much
larger
and
heteroge-
neous
study
area
(the
contiguous
US).
As
an
ideal,
we
view
inherent
community
disaster
resilience
as
a
complex
process
of
interactions
between
various
social
systems,
each
with
their
own
form
and
function,
but
working
in
tandem
to
provide
for
the
betterment
of
the
whole
community
(US
NRC,
2012).
Our
place-based
metrics
approach
aims
to
capture
a
snapshot
of
all
facets
of
a
community
that
can
be
integrated
toward
the
goal
of
enhancing
disaster
resilience:
its
infrastructure,
its
governance
structures,
economy,
natural
resources
and
attributes,
and
its
demographic
character
and
social
interactions.
Such
a
metric
approach
is
needed
to
benchmark
inherent
resilience
so
that
the
effects
of
future
policies
and
programs
aiming
to
enhance
community
resilience
can
be
evaluated.
2.
Data
and
methods
This
paper
details
the
construction
of
a
composite
index
of
community
resilience
to
disasters
and
its
geographic
variability
when
applied
to
specific
places.
The
term
‘community’
may
refer
to
a
geographic
place
such
as
a
neighborhood,
municipality,
or
rural
area;
it
may
also
refer
to
a
relational
unit
such
as
a
club,
congregation,
or
workplace
(Kloos
et
al.,
2012).
In
building
a
community
resilience
index,
this
research
operationalizes
‘com-
munity’
as
a
locality
rather
than
as
relational
examining
interactions
among
populations,
organizations,
or
other
entities.
The
choice
of
US
counties
as
the
spatial
units
of
analysis
is
driven
by
a
number
of
factors.
Counties
are
the
smallest
level
of
aggregation
for
which
a
wide
range
of
human
and
physical
data
are
consistently
and
completely
collected
and
archived.
Furthermore,
county
governments
are
heavily
involved
in
emergency
manage-
ment
activities,
serving
important
roles
as
intermediaries
between
municipalities
and
state
governments
(Bowman
and
Parsons,
2009;
McGuire
and
Silvia,
2010;
Waugh,
1994).
They
are
the
primary
local
administrative
unit
for
national
emergency
manage-
ment
authorities.
Finally,
county
boundaries
change
relatively
little
over
time
in
comparison
to
municipal
or
other
census
unit
boundaries.
Although
this
analysis
uses
only
one
time
step
(2010),
the
ability
to
collect
the
same
suite
of
variables
for
both
the
past
and
future
is
another
desirable
quality
of
county-level
data.
Alaska
S.L.
Cutter
et
al.
/
Global
Environmental
Change
29
(2014)
65–77
66

and
Hawaii
were
omitted
from
the
analysis
due
to
lack
of
data
for
all
the
variables,
as
well
as
changes
in
the
Alaskan
county
census
boundaries
over
time.
The
principal
drawback
of
using
counties
as
communities
is
the
wide
range
of
county
sizes,
in
terms
of
both
population
and
geographic
area
in
the
contiguous
US
(US
NRC,
2012).
2.1.
Data
sources
Data
for
this
research
were
collected
from
30
different
sources
(Table
1).
Mostly
free
and
open
data
sources
were
used
intentionally
so
that
the
indicator
set
could
be
replicated
with
a
reasonable
amount
of
effort.
Over
20
datasets
were
obtained
from
the
US
federal
government
via
online
data
portals,
websites,
and
hard
copies
accessed
at
a
university
library.
Hazard
events
and
loss
data
(SHELDUS
1
)
were
provided
by
the
Hazards
and
Vulnerability
Research
Institute
at
the
University
of
South
Carolina.
Four
datasets
were
obtained
from
nonprofit
groups’
websites,
two
via
a
contact
with
the
American
Red
Cross,
and
one
from
an
open
access
data
portal
of
a
major
news
outlet.
At
the
time
of
writing,
only
one
data
source
out
of
the
30
used
–
Dun
and
Bradstreet’s
Million
Dollar
Database
–
required
a
paid
subscription
to
acquire
data.
This
paper
focuses
on
the
year
2010
for
construction
and
analysis
of
the
composite
resilience
indicator.
Accordingly,
the
2010
decennial
census
is
the
most
used
dataset.
However,
data
were
not
strictly
available
for
the
year
2010
for
every
desired
variable
(Table
1).
For
example,
the
2010
decennial
census
did
not
include
a
long
form,
meaning
that
some
demographic
and
economic
data
were
not
available
for
2010.
Thus,
the
American
Community
Survey
five-year
estimates
from
2006
to
2010
were
the
second
most
used
source
of
data
behind
the
2010
census.
These
data
represent
the
closest
estimates
to
2010
levels
as
possible
for
each
variable.
In
other
cases,
the
closest
available
data
to
approximate
reality
in
2010
were
from
as
early
as
2005
or
as
recent
as
2012.
Other
variables,
such
as
disaster
aid
experience,
are
constructed
using
ten-year
averages
from
the
years
2000
to
2009.
2.2.
Data
preprocessing
After
collecting
the
raw
variables,
the
data
items
then
underwent
a
process
of
transformation,
normalization,
and
theoretical
orientation.
For
the
purpose
of
this
research,
transfor-
mation
is
the
conversion
of
raw
count
variables
into
percentages,
rates,
differences,
or
averages.
This
is
a
necessary
step
so
communities
of
varying
sizes
and
characteristics
can
be
compared
(as
in
Cutter
et
al.,
2010).
For
example,
there
are
roughly
380,000
households
with
access
to
a
vehicle
in
the
large
urbanized
area
of
Philadelphia
County,
PA
compared
to
about
8000
in
more
rural
Holmes
County,
OH,
home
to
a
large
Amish
population.
When
these
figures
are
transformed
into
percentages
of
all
households
with
access
to
a
vehicle,
we
see
that
the
two
counties
are
roughly
comparable
on
this
variable
despite
differences
in
population,
with
about
66%
of
households
in
both
places
having
access
to
a
vehicle.
Normalization
refers
to
scaling
all
the
variables
using
one
method
so
that
all
the
data
have
comparable
reference
points.
We
used
min–max
scaling,
a
straightforward
normalization
technique
common
in
social
indicators
research
(Tarabusi
and
Guarini,
2012).
Min–max
normalization
assigns
a
value
of
0
to
the
minimum
value
and
1
to
the
maximum
value.
All
other
values
are
scaled
between
zero
and
one
by
subtracting
the
minimum
value
and
dividing
by
the
range
(the
minimum
subtracted
from
the
maximum).
While
Table
1
Data
sources
and
time
periods
used
to
construct
resilience
indicator
set.
Number
Dataset
Time
Period
Data
Provider
United
States
Federal
Government
1
USA
Counties
Database
2007
Census
Bureau
2
County
and
City
Data
Book
2007
3
County
Business
Patterns
2009–2010
4
Decennial
Census
2010
5
Small
Area
Health
Insurance
Estimates
2010
6
Tiger/Line
2010
7
Current
Population
Estimate
2005,
2012
8
American
Community
Survey
Three-Year
Estimates
2008–2010
9
American
Community
Survey
Five-Year
Estimates
2006–2010
10
Hazard
Mitigation
Grant
Program
2000–2009
Federal
Emergency
Management
Agency
11
Presidential
Disaster
Declarations
Database
2000–2009
12
Citizen
Corps
Councils
2010
13
National
Flood
Insurance
Program
2010
14
National
Land
Cover
Dataset
2006
US
Geological
Survey
15
National
Atlas
2010
16
Quarterly
Census
of
Employment
and
Wages
2010
Bureau
of
Labor
Statistics
17
Census
of
Agriculture
2007
Department
of
Agriculture
18
National
Center
for
Education
Statistics
2009–2010
Department
of
Education
19
Electricity
Consumption
2010
Energy
Information
Administration
20
Broadband
Internet
Access
2010
Federal
Communications
Commission
21
Water
Supply
Stress
Index
2005
Forest
Service
22
Nuclear
Power
Plants
Database
2010
Nuclear
Regulatory
Commission
23
Railroad
Network
2010
Oak
Ridge
National
Laboratory
Academic
24
Spatial
Hazard
Events
and
Losses
Database
for
the
US
(SHELDUS)
2000–2009
Univ.
South
Carolina
Hazards
and
Vulnerability
Research
Institute
Nonprofit/Open
Access
25
Religious
Congregations
and
Membership
Study
2010
Association
of
Religion
Data
Archives
26
Farm
Subsidies
2010
Environmental
Working
Group
27
Map
the
Meal
Gap
2010
Feeding
America
28
US
2012
Presidential
Election
2012
The
Guardian
29
Volunteers
and
Preparedness
Training
2013
American
Red
Cross
Proprietary
30
Million
Dollar
Database
2010
Dun
and
Bradstreet
S.L.
Cutter
et
al.
/
Global
Environmental
Change
29
(2014)
65–77
67

min–max
scaling
does
not
address
problems
related
to
outlier
or
extreme
observations
in
the
data,
it
does
make
comparison
and
interpretation
of
a
large
number
of
variables
much
easier.
One
disadvantage
of
using
normalization,
however,
is
that
the
final
score
is
not
an
absolute
measurement
of
community
resilience
for
a
single
location,
but
rather
a
relative
value
in
which
multiple
places
can
be
compared.
Such
a
relative
estimation
of
inherent
resilience
possibly
over
or
under
estimates
local
resilience
through
the
normalization
process,
on
the
other
hand
it
provides
easily
understood
comparisons
between
places
at
a
particular
point
in
time.
Utilizing
such
normalized
values
is
useful
for
benchmarking
progress
in
enhancing
resilience
over
time
and
across
space.
Finally,
the
orientation
of
each
variable
was
adjusted
so
that
larger
values
corresponded
theoretically
to
higher
resilience.
The
performance
regimes
variable
using
the
distance
from
a
county
seat
to
the
nearest
metropolitan
area
is
a
good
example
of
the
process
of
inverting
the
values
to
match
the
theoretical
orientation.
According
to
Bowman
and
Parsons
(2009),
rural
counties
tend
to
reach
out
to
nearby
urban
areas
to
form
partnerships
to
enhance
their
ability
to
manage
emergency
situations.
As
a
proxy
to
measure
the
potential
for
these
partnerships
to
form,
we
calculated
a
performance
regime
score
where
higher
resilience
was
equal
to
1
–
the
least
distance
from
each
county
seat
in
the
US
to
the
nearest
county
seat
within
a
Metropolitan
Statistical
Area
such
that
proximity
increased
resilience.
2.3.
Construction
of
the
baseline
resilience
indicators
for
communities
(BRIC)
Our
theoretical
orientation
combines
the
notion
of
distinct
capitals
(Ritchie
and
Gill,
2007)
and
the
Disaster
Resilience
of
Place
(DROP)
model’s
concept
of
inherent
resilience
(Cutter
et
al.,
2008a).
An
initial
set
of
61
variables
were
chosen
through
conceptual,
theoretical,
and/or
empirical
justification
from
previous
research
to
represent
each
of
the
six
capitals
or
types
of
resilience
–
social,
economic,
community,
institutional,
housing/infrastructure,
and
environmental
(see
Cutter
et
al.,
2008a,
2010;
Gunderson,
2010;
Kulig
et
al.,
2013;
Mowbray
et
al.,
2007;
Norris
et
al.,
2008;
Rose
and
Krausmann,
2013;
Sherrieb
et
al.,
2010).
Several
variables
were
eliminated
after
correlation
analyses
revealed
a
high
degree
of
collinearity
with
other
variables,
while
others
were
eliminated
after
further
consideration
because
they
were
not
conceptually
congruent
with
the
majority
of
the
remaining
variables.
One
example
of
the
latter
was
a
variable
based
on
Florida’s
(2002)
classification
of
creative
class
occupations.
While
having
a
higher
proportion
of
persons
employed
in
creative
occupations
–
designers,
engineers,
artists,
architects
–
could
be
used
to
estimate
the
potential
for
innovation,
this
variable
did
not
fit
well
in
any
of
our
six
resilience
categories.
Following
elimination
of
conceptually
and
statistically
extraneous
variables,
the
final
set
contained
49
indicators
(Table
2).
Since
disaster
resilience
is
an
abstract
concept,
an
important
caveat
for
this
work
is
that
we
estimate
resilience
levels
using
existing
data
collected
originally
for
a
variety
of
purposes.
Therefore,
the
reasoning
behind
variable
inclusion
should
be
explained.
The
relevant
citations
that
provide
either
empirical
or
conceptual
justification
are
listed
in
Table
2.
The
10
indicators
in
the
social
resilience
category
are
intended
to
capture
demographic
qualities
of
a
community’s
population
that
tend
to
associate
with
physical
and
mental
wellness
leading
to
increased
comprehension,
communication,
and
mobility.
For
example,
having
a
greater
concentration
of
physicians,
mental
health
support
facilities,
and
health
insurance
enrollees
tends
to
increase
the
overall
physical
and
mental
health
of
communities.
These
are
useful
qualities
for
preparing
for,
responding
to,
and
recovering
from
disasters.
Likewise,
having
more
proficient
English
speakers
and
greater
access
to
telephones
enables
communication
which
is
vital
during
disaster
events.
Community
capital
can
be
related
to
demographic
qualities,
yet
is
separated
from
the
social
resilience
category
in
our
index
structure
in
order
to
estimate
the
propensity
for
a
community
to
call
on
the
good
will
of
local
citizens
to
assist
their
neighbors
and
fellow
citizens
–
a
whole
community
approach
to
emergencies.
A
community
that
would
seem,
demographically,
to
be
resilient
to
disasters
may
not
be
particularly
conscientious
and
helpful
to
one
another.
Thus,
our
seven
community
capital
indicators
conceptu-
ally
represent
the
level
of
community
engagement
and
involve-
ment
in
local
organizations
and
the
potential
for
local
ties
and
social
networks
that
can
be
critical
for
survival
and
recovery
during
disasters.
Persons
who
participate
in
local
civic
or
religious
organizations
are
likely
to
have
some
level
of
community
social
capital
to
draw
upon
for
help
beyond
their
family
and/or
close
neighbors
during
disasters.
High
levels
of
Red
Cross
training
and
volunteerism
further
indicate
a
community
that
is
likely
to
cope
well
during
adverse
events.
In
addition,
having
a
large
portion
of
the
local
population
that
has
lived
there
for
more
than
a
few
years
increases
the
likelihood
of
having
a
community
that
is
engaged
and
invested
in
its
own
well-being,
in
both
short-
and
long-term
contexts.
Eight
indicators
are
in
the
economic
resilience
category.
The
indicators
are
intended
to
represent
community
economic
vitality,
diversity,
and
equality
in
compensation,
but
not
to
represent
resilience
of
individual
businesses
per
se.
Rather,
our
interest
is
in
how
the
general
economic
profile
and
character
of
a
community
can
be
of
benefit
in
a
disaster
context.
Thus,
general
economic
vitality
is
related
to
employment
and
homeownership
rates.
Diversity
is
also
important
for
long-term
economic
resilience,
meaning
that
the
local
economy
is
not
overly
dependent
upon
continuing
success
in
just
one
sector.
Extractive
industries
in
the
primary
sector
and
tourism
economies
are
perhaps
the
sectors
most
prone
to
being
undermined
by
a
disaster,
hence
their
inclusion
in
this
analysis.
To
represent
equality
in
compensation,
we
included
inverted
Gini
coefficients
indicating
equality
in
income
distribution
across
races
and
ethnicities.
Additionally,
we
included
an
indicator
of
equality
for
income
by
gender.
Communities
can
enhance
disaster
resilience
through
non-discriminatory
wage
policies,
ensuring
that
all
groups
have
fair
access
to
resources.
Finally,
we
included
several
indicators
of
economic
ties
outside
of
communities.
Businesses
with
more
employees,
or
businesses
that
are
part
of
regional
or
national
chains,
are
often
able
to
bring
in
resources
from
outside
to
assist
a
community
during
a
disaster.
Likewise,
communities
with
a
large
proportion
of
federal
government
jobs
are
often
efficient
at
garnering
more
economic
resources
to
cope
with
disasters.
The
nine
housing/infrastructure
indicators
estimate
the
quality
of
housing
construction
and
a
myriad
of
physical
capacities
within
a
county
to
house
the
displaced,
provide
emergency
medical
care,
facilitate
evacuations,
and
maintain
schooling
activities,
among
other
disaster-relevant
infrastructural
capacities.
Environmental
resilience
indicators
conceptually
relate
to
qualities
of
the
environment
that
enhance
absorptive
capacity
of
coastal
surges
and
freshwater
flooding
in
particular.
The
remaining
indicators
estimate
the
efficiency
with
which
a
community
uses
natural
resources.
The
latter
is
important
for
longer
term,
slower
onset
disasters
whereas
the
former
is
important
for
shorter
term
disasters.
Finally,
the
10
institutional
indicators
are
meant
to
capture
aspects
related
to
programs,
policies,
and
governance
of
disaster
resilience.
Three
variables
are
associated
with
coordination
between
governments.
When
there
are
fewer
jurisdictions
within
a
county,
it
is
easier
to
manage
and
assign
resources
during
a
disaster
than
where
there
are
many
jurisdictions
or
governments
that
must
constantly
coordinate
with
one
another.
On
the
other
S.L.
Cutter
et
al.
/
Global
Environmental
Change
29
(2014)
65–77
68

Table
2
Indicator
sets
with
descriptions
of
calculation
for
individual
indicators.
Dataset
numbers
correspond
to
those
in
Table
1.
Resilience
concept
Variable
description
Justification
Dataset
Social
resilience
Educational
attainment
equality
Negative
absolute
difference
between
%
population
with
college
education
and
%
population
with
less
than
high
school
education
Morrow
(2008)
and
Sherrieb
et
al.
(2010)
9
Pre-retirement
age
%
Population
below
65
years
of
age
Morrow
(2008)
and
Peek
(2010)
4
Transportation
%
Households
with
at
least
one
vehicle
Peacock
et
al.
(2010)
and
Tierney
(2009)
9
Communication
capacity
%
Households
with
telephone
service
available
Burger
et
al.
(2013)
and
Strawderman
et
al.
(2012)
9
English
language
competency
%
Population
proficient
English
speakers
Messias
et
al.
(2012)
and
Senkbeil
et
al.
(2013)
9
Non-special
needs
%
Population
without
sensory,
physical,
or
mental
disability
Davis
and
Phillips
(2009)
and
Matherly
and
Mobley
(2012)
8
Health
insurance
%
Population
under
age
65
with
health
insurance
Chandra
et
al.
(2011)
and
Plough
et
al.
(2013)
4,
5
Mental
health
support
Psychosocial
support
facilities
per
10,000
persons
Pietrzak
et
al.
(2012)
and
Springgate
et
al.
(2011)
4,
30
Food
provisioning
capacity
Food
security
rate
Pingali
et
al.
(2005)
and
Tobin
and
Whiteford
(2012)
28
Physician
access
Physicians
per
10,000
persons
Chandra
et
al.
(2011)
and
Norris
et
al.
(2008)
9
Economic
resilience
Homeownership
%
Owner-occupied
housing
units
Haveman
and
Wolff
(2005)
and
Pendall
et
al.
(2012)
9
Employment
rate
%
Labor
force
employed
Rose
and
Krausmann
(2013)
and
Sherrieb
et
al.
(2010)
9
Race/ethnicity
income
equality
Negative
Gini
coefficient
Norris
et
al.
(2008)
and
Sherrieb
et
al.
(2010)
9
Non-dependence
on
primary/tourism
sectors
%
Employees
not
in
farming,
fishing,
forestry,
extractive
industry,
or
tourism
Rose
and
Krausmann
(2013)
and
Sherrieb
et
al.
(2010)
9
Gender
income
equality
Negative
absolute
difference
between
male
and
female
median
income
Enarson
(2012)
and
Sherrieb
et
al.
(2010)
9
Business
size
Ratio
of
large
to
small
businesses
Rose
and
Krausmann
(2013)
and
Wein
and
Rose
(2011)
3
Large
retail-regional/national
geographic
distribution
Large
retail
stores
per
10,000
persons
Rose
and
Krausmann
(2013)
and
Wein
and
Rose
(2011)
4,
30
Federal
employment
%
Labor
force
employed
by
federal
government
Heinz
Center
(2002)
and
Rose
and
Krausmann
(2013)
9,
16
Community
capital
Place
attachment-not
recent
immigrants
%
Population
not
foreign-born
persons
who
came
to
US
within
previous
five
years
Norris
et
al.
(2008)
and
Sherrieb
et
al.
(2010)
9
Place
attachment-native
born
residents
%
Population
born
in
state
of
current
residence
Norris
et
al.
(2008)
and
Sherrieb
et
al.
(2010)
9
Political
engagement
%
Voting
age
population
participating
in
presidential
election
Peacock
et
al.
(2010)
and
Sherrieb
et
al.
(2010)
7,
29
Social
capital-religious
organizations
Persons
affiliated
with
a
religious
organization
per
10,000
persons
Sherrieb
et
al.
(2010)
and
Walsh
(2007)
25
Social
capital-civic
organizations
Civic
organizations
per
10,000
persons
Sherrieb
et
al.
(2010)
and
Walsh
(2007)
3,
4
Social
capital-disaster
volunteerism
Red
cross
volunteers
per
10,000
persons
Paton
and
Johnston
(2006)
4,
29
Citizen
disaster
preparedness
and
response
skills
Red
cross
training
workshop
participants
per
10,000
persons
Godschalk
(2003)
4,
29
Institutional
resilience
Mitigation
spending
Ten
year
average
per
capita
spending
for
mitigation
projects
Rose
(2007)
and
Godschalk
et
al.
(2009)
4,
10
Flood
insurance
coverage
%
Housing
units
covered
by
National
Flood
Insurance
Program
Cheong
(2011)
and
Michel-Kerjan
et
al.
(2012)
4,
13
Jurisdictional
coordination
Governments
and
special
districts
per
10,000
persons
Murphy
(2007)
and
Ansell
et
al.
(2010)
1,
4
Disaster
aid
experience
Presidential
disaster
declarations
divided
by
number
of
loss-causing
hazard
events
from
2000
to
2009
Cutter
et
al.
(2008a)
and
Tierney
and
Bruneau
(2007)
11,
24
Local
disaster
training
%
Population
in
communities
with
Citizen
Corps
program
Godschalk
(2003)
and
Simonovich
and
Sharabi
(2013)
4,
12
Performance
regimes-state
capital
Proximity
of
county
seat
to
state
capital
Bowman
and
Parsons
(2009)
6,
15
Performance
regimes-nearest
metro
area
Proximity
of
county
seat
to
nearest
county
seat
within
a
Metropolitan
Statistical
Area
Bowman
and
Parsons
(2009)
6,
15
S.L.
Cutter
et
al.
/
Global
Environmental
Change
29
(2014)
65–77
69

hand,
communities
often
benefit
from
being
located
in
close
proximity
to
seats
of
political
and
economic
power.
These
situations
present
opportunities
to
benefit
from
access
to
decision-makers
and
resources
typically
found
in
large
urban
areas
or
state
capital
cities.
Several
indicators
also
approximate
the
value
of
experience
in
various
programs
and
policies
that
may
benefit
communities
before,
during,
and
after
disasters.
These
range
from
insurance
programs
to
mitigation
and
disaster
aid
programs.
In
this
context,
therefore,
resilience
stems
not
directly
from
dollar
amounts,
but
from
the
institutional
knowledge
and
experience
of
navigating
such
programs
in
order
to
obtain
vital
resources.
A
third
aspect
of
institutional
resilience
is
local
disaster
training,
which
is
estimated
via
two
variables.
For
example,
all
active
nuclear
power
plants
are
required
to
disseminate
safety
and
evacuation
information
to
persons
living
within
ten
miles.
The
last
aspect
of
institutional
resilience
is
population
stability.
It
is
placed
within
this
category
because
rapid
population
change
(a
five
year
period
is
used
here)
places
strain
on
local
institutions.
A
massive
influx
of
population
can
overwhelm
existing
infrastructures,
while
a
precipitous
drop
in
population
results
in
large
reductions
of
local
tax
incomes.
The
latter
can
have
deleterious
effects
on
local
government
disaster
preparedness
and
mitigation
budgets.
To
test
whether
the
total
set
of
variables
measures
the
abstract
concept
of
resilience,
we
used
Cronbach’s
alpha,
a
widely
recognized
statistic
for
diagnosing
composite
indicator
construc-
tion
for
internal
consistency
(Martin
and
Savage-McGlynn,
2013;
Nardo
et
al.,
2008).
The
alpha
value
for
all
49
variables
was
0.65,
which
suggests
a
moderate
level
of
interrelatedness
amongst
the
variables
in
the
indicator
set.
This
level
of
internal
consistency
is
acceptable
for
building
a
composite
index
for
a
baseline
national
resilience
metric
(Jaeger
et
al.,
2013;
Setbon
and
Raude,
2010).
Because
our
research
design
and
theoretical
orientation
incorporated
different
types
of
resilience
as
defined
in
the
extant
literature,
we
do
not
expect
the
individual
sub-indices
to
be
internally
consistent
as
measured
using
Cronbach’s
alpha
(Table
3).
None
of
the
sub-indices
show
high
degrees
of
internal
consistency.
Social
resilience
variables
are
the
most
internally
consistent
while
the
institutional
and
environmental
variables
are
the
least
internally
consistent.
When
taken
as
a
group,
however,
the
six
aggregated
sub-
indices
do
show
a
moderate
level
of
internal
consistency
(Cronbach’s
alpha
is
0.55).
Further,
the
sub-indices
are
generally
independent
of
one
another
with
correlations
less
than
r
=
0.6.
The
highest
linear
correlation
was
between
the
social
and
economic
sub-indices
(r
=
0.54),
but
there
was
little
association
between
economic
and
community
capacity,
or
housing/infrastructure
and
environmental.
The
environmental
sub-index
showed
little
relationship
to
the
community
capital
and
institutional
sub-indices,
and
even
had
weak
negative
associations
with
the
social,
economic,
and
housing/
infrastructure
sub-indices.
This
is
not
unexpected.
Our
binning
of
variables
into
sub-indices
does
approximate
the
broad
facets
of
resilience
as
determined
a
priori.
However,
each
of
the
sub-indices
may
only
represent
a
small
part
of
the
multi-dimensional
nature
of
resilience
and
the
variables
within
in
them
may
well
interact
with
variables
from
different
sub-indices
to
produce
the
broader
dimensions
in
ways
that
have
yet
to
be
fully
explored.
At
the
final
step,
the
BRIC
was
constructed
by
summing
the
composites
of
the
six
resilience
sub-indexes.
Potential
scores
range
from
zero
to
six,
with
higher
scores
corresponding
to
more
resilience,
and
lower
scores,
less
resilience.
Table
3
Cronbach’s
alpha
results
for
indicators
within
each
resilience
category.
Resilience
category
Number
of
indicators
Cronbach’s
alpha
Social
10
0.533
Housing/infrastructural
9
0.411
Community
capital
7
0.317
Economic
8
0.242
Institutional
10
0.074
Environmental
5
0.028
Table
2
(Continued
)
Resilience
concept
Variable
description
Justification
Dataset
Population
stability
Population
change
over
previous
five
year
period
USNRC
(2012)
and
Sherrieb
et
al.
(2010)
4,
7
Nuclear
plant
accident
planning
%
Population
within
10
miles
of
nuclear
power
plant
USNRC
(2012)
and
Urbanik
II
(2000)
4,
22
Crop
insurance
coverage
Crop
insurance
policies
per
square
mile
Glauber
(2013)
and
Gladwin
and
Smith
(2013)
6,
27
Housing/infrastructural
resilience
Sturdier
housing
types
%
Housing
units
not
manufactured
homes
Sutter
and
Simmons
(2010)
and
Tierney
(2009)
9
Temporary
housing
availability
%
Vacant
units
that
are
for
rent
Fe
´lix
et
al.
(2013)
and
Johnson
(2007)
9
Medical
care
capacity
Hospital
beds
per
10,000
persons
Birkmann
et
al.
(2013)
and
Cimellaro
et
al.
(2010)
2,
4
Evacuation
routes
Major
road
egress
points
per
10,000
persons
Emmer
et
al.
(2008)
and
Harrald
(2012)
15
Housing
stock
construction
quality
%
Housing
units
built
prior
to
1970
or
after
2000
Mileti
(1999)
and
Theckethil
(2006)
9
Temporary
shelter
availability
Hotels/motels
per
10,000
persons
Johnson
(2007)
and
Tierney
(2009)
3,
4
School
restoration
potential
Public
schools
per
10,000
persons
Cross
(2014)
and
Ronan
and
Johnston
(2005)
4,
18
Industrial
re-supply
potential
Rail
miles
per
square
mile
Cutter
et
al.
(2008b)
6,
23
High
speed
internet
infrastructure
%
Population
with
access
to
broadband
internet
service
UNDESA
(2007)
4,
20
Environmental
resilience
Local
food
suppliers
Farms
marketing
products
through
Community
Supported
Agriculture
per
10,000
persons
Berardi
et
al.
(2011)
4,
17
Natural
flood
buffers
%
Land
in
wetlands
Beatley
and
Newman
(2013)
and
Brody
et
al.
(2012)
6,
14
Efficient
energy
use
Megawatt
hours
per
energy
consumer
UNDESA
(2007)
19
Pervious
surfaces
Average
percent
perviousness
Brody
et
al.
(2012,
2014)
14
Efficient
Water
Use
Inverted
water
supply
stress
index
UNDESA
(2007)
21
S.L.
Cutter
et
al.
/
Global
Environmental
Change
29
(2014)
65–77
70

We
also
constructed
an
additional
index
inductively
from
the
49
variables
using
principal
components
to
see
whether
they
would
be
markedly
different
from
the
BRIC
overall
sub-index
construction
method.
Based
on
the
eigenvalue
Scree
plot,
we
retained
six
components
explaining
39.4%
of
the
variance
in
the
data.
The
six
components
were
characterized
by
different
sets
of
indicators
than
our
deductive
binning
categories.
The
six
compo-
nents
were:
(1)
younger
populations,
less
concentration
of
employment
in
the
primary
sector,
and
closer
proximity
to
large
urban
areas
(combining
elements
of
social,
economic,
and
institu-
tional
resilience);
(2)
combination
of
several
indicators
of
social
and
economic
equality
and
capacity;
(3)
urban
density
(housing/
infrastructure
and
environmental
resilience,
regardless
of
the
statistical
procedures
used);
(4)
English
proficiency,
voter
turnout,
and
lack
of
recent
immigrants
(social
and
community
capital);
(5)
native
born
residents,
sturdier
housing,
areas
with
crop
insurance
(housing/infrastructure,
community
capital,
and
institutional
resil-
ience);
and
(6)
emergency
medical
capacity,
temporary
shelters,
and
civic
organizations
(housing/infrastructure
and
community
capital).
We
constructed
an
index
from
these
six
components
by
adding
the
factor
scores
for
components
one
through
six.
We
summed
the
components
because
their
respective
variable
loadings
suggested
they
would
enhance
resilience
rather
than
detract
from
it.
The
Pearson
correlation
of
the
PCA-style
index
with
BRIC
sub-index
is
r
=
0.77.
In
response
to
our
first
research
question,
it
is
clear
that
a
common
set
of
resilience
indicators
can
be
applied
to
examine
the
variability
in
resilience
among
US
counties.
3.
Spatial
patterns
and
drivers
There
were
3108
counties/parishes
in
our
study;
Broomfield,
Colorado
was
omitted
because
it
did
not
exist
before
2001
and
some
of
the
data
needed
to
calculate
individual
indicators
were
not
available.
The
average
value
of
the
BRIC
was
3.18,
with
a
standard
deviation
of
0.37,
a
minimum
value
of
1.67
(least
resilient)
and
a
maximum
value
of
4.39
(most
resilient).
The
BRIC
scores
were
classified
using
standard
deviations
into
five
categories
for
visualization
purposes
(Fig.
1).
Regionally,
the
Midwest
had
the
highest
resilience
index
values
(scores
>
3.7)
centered
in
Iowa
and
southern
Minnesota
counties,
and
extending
north
into
the
eastern
Dakotas
and
southwest
into
Nebraska
(Fig.
1).
Other
areas
with
high
values
are
in
eastern
Wisconsin,
central
Illinois,
and
northwestern
Ohio.
Concentrations
of
moderately
high
(between
3.3
and
3.7)
resilience
are
seen
throughout
the
Great
Plains
states
from
Kansas
to
the
Canadian
border
and
eastward
into
the
Ohio
River
Valley.
Moderately
high
levels
of
resilience
are
also
seen
along
the
eastern
seaboard
from
Virginia
to
Maine
as
well
as
interior
Pennsylvania
and
New
York.
The
lowest
values
(scores
<
2.6)
of
the
resilience
index
were
located
in
three
main
regions:
the
western
US,
especially
Nevada,
and
the
Four
Corners
region;
southern
and
west
Texas,
especially
along
the
Rio
Grande,
and
in
Appalachia,
particularly
in
eastern
Kentucky
and
West
Virginia.
Moderately
low
values
(between
2.6
and
3.0)
dominate
in
the
West,
the
South,
and
in
the
upper
peninsula
of
Michigan.
Decomposing
BRIC
based
on
the
six
conceptual
capitals
allows
further
exploration
of
geographic
trends
in
the
data
(Fig.
2).
Higher
values
for
social
resilience
were
concentrated
primarily
in
the
upper
Midwest,
the
Great
Lakes,
and
northeast-
ern
US
(Fig.
2a).
Lower
social
resilience
was
found
across
the
southern
tier
of
states
from
the
Carolinas
to
southern
Nevada
and
California,
as
well
as
in
eastern
Kentucky
and
into
West
Virginia.
Higher
economic
resilience
is
also
very
concentrated
in
the
Midwest,
Mid-Atlantic,
and
the
Northeast.
Pockets
of
high
economic
resilience
are
found
in
Nashville,
Atlanta,
Dallas,
Denver,
and
Salt
Lake
City
metropolitan
areas
(Fig.
2b).
Much
of
the
US
west
of
the
high
plains
had
moderate
to
low
economic
resilience,
as
did
the
lower
Mississippi
River
valley.
Community
capital
presented
a
different
pattern
from
the
social
and
economic
categories
(Fig.
2c).
Here,
the
highest
scores
were
found
in
three
areas:
Minnesota,
Iowa,
and
the
Dakotas;
northern
Ohio
and
western
Pennsylvania;
and
Mississippi,
Alabama,
and
Louisiana.
These
are
easily
explained
by
the
strong
place
attachment
of
residents
and
participation
in
organized
religious
groups.
The
lowest
levels
of
community
capital
seemed
to
be
focused
in
the
far
western
US,
including
Fig.
1.
Disaster
resilience
index
for
the
contiguous
United
States,
2010.
S.L.
Cutter
et
al.
/
Global
Environmental
Change
29
(2014)
65–77
71

Washington,
Oregon,
California,
Nevada,
and
New
Mexico.
The
second
prominent
area
with
low
community
capital
scores
was
the
Florida
peninsula.
Institutional
resilience
is
again
strongest
in
the
Midwestern
and
plains
regions,
and
along
the
Gulf
Coast
(Fig.
2d).
Much
of
the
rural
high
plains
and
intermountain
western
US
had
lower
scores.
Pockets
of
moderately
high
institutional
resilience
appear
in
the
Pacific
Northwest
in
western
Oregon
and
the
Puget
Sound
region
of
Washington.
Housing
&
infrastructure
resilience
scores
were
similar
to
the
social
scores
–
higher
in
portions
of
the
Plains,
the
Midwest,
and
the
Northeast
(Fig.
2e).
Lower
levels
of
housing
and
infrastructural
resilience
were
found
in
the
South
and
the
Four
Corners
region
of
the
Southwest.
Finally,
riverine
and
coastal
areas
of
the
Southeast,
mid-Atlantic,
the
lower
Mississippi
River
valley,
and
the
upper
Midwest
scored
higher
for
environmental
resilience
(Fig.
2f)
than
other
regions.
Lowest
scores
were
found
in
the
western
third
of
the
US
and
in
densely
populated
urban
areas
east
of
the
Rocky
Mountains.
Fig.
2.
Resilience
scores
for
six
categories:
(A)
social,
(B)
economic,
(C)
community
capital,
(D)
institutional,
(E)
housing/infrastructural,
(F)
environmental.
Data
classified
into
low,
medium,
and
high
using
the
standard
deviation
method,
as
in
Fig.
1.
S.L.
Cutter
et
al.
/
Global
Environmental
Change
29
(2014)
65–77
72

In
order
to
understand
which
composite
and
individual
indicators
were
most
influential
in
the
overall
resilience
index,
it
is
useful
to
focus
on
some
of
the
very
highest
and
lowest
scoring
counties
in
the
US.
No
county
received
the
maximum
possible
score
of
six,
yet
the
counties
with
the
ten
highest
scores
were
near
or
above
4.0
(Table
4).
The
high
composite
scores
for
St.
Charles
and
St.
John
the
Baptist,
Louisiana
were
driven
principally
by
their
high
institutional
scores
(ranked
first
and
second),
and
five
indicators
in
particular:
access
to
resources
within
the
New
Orleans
metropolitan
area,
inclusion
in
the
emergency
planning
zone
of
the
Waterford
3
nuclear
plant,
high
rates
of
participation
in
the
National
Flood
Insurance
Program,
high
levels
of
experience
with
Presidential
Disaster
Declarations,
and
high
average
annual
spending
on
mitigation
projects.
St.
Charles
and
St.
John
the
Baptist
parishes
also
had
very
high
environmental
scores
(ranked
23
and
20,
respectively),
due
to
the
existence
of
wetlands
serving
as
natural
buffers
for
flooding.
Despite
more
average
scores
on
the
other
four
sub-indices,
especially
housing/infrastructure,
the
strength
of
the
institutional
and
environmental
capital
was
sufficient
to
propel
St.
Charles
to
the
top
position.
Seven
of
the
10
most
resilient
counties
were
located
in
Minnesota,
Wisconsin,
or
Iowa.
The
Minnesota
counties
generally
scored
very
well
(top
20%)
in
the
social,
economic,
and
community
capital
categories.
Key
indicators
that
contributed
to
higher
scores
were
higher
percentages
of
persons
with
health
insurance,
higher
levels
of
food
security,
higher
rates
of
employment
and
homeownership,
and
high
percentages
of
residents
who
are
not
recent
immigrants.
The
Iowa
and
Wisconsin
counties
generally
scored
in
the
top
twenty
percent
(or
better)
in
all
categories
except
environmental,
with
an
exceptional
institutional
score
for
Shelby,
Iowa
and
an
exceptional
community
capital
score
for
Winnebago,
Wisconsin,
both
in
the
top
one
percent
nationally.
The
10
lowest
resilience
scores
were
all
near
or
below
2.0
(Table
4).
Though
the
composite
scores
for
the
six
resilience
categories
were
generally
low
for
all
10
counties,
the
social,
institutional,
and
community
capital
scores
were
especially
weak
in
these
places
–
consistently
in
the
bottom
ten
percent
nationally.
The
least
resilient
counties
are
all
in
the
southwestern
US.
Imperial
County,
California
had
the
lowest
BRIC
score,
1.67.
The
data
suggest
the
drivers
of
the
lower
level
of
resilience
here
were
consistently
low
scores
(in
the
bottom
90%)
on
all
sub-indices
except
housing/
infrastructure.
Some
of
the
key
indicators
were
low
levels
of
equality
in
educational
attainment,
a
lower
English
proficiency,
fewer
households
with
access
to
a
vehicle,
fewer
physicians,
high
levels
of
food
insecurity,
a
high
number
of
recent
immigrants,
lower
levels
of
voter
participation
and
hazard
experience,
and
a
high
level
of
water
stress.
Presidio
County,
Texas
had
the
second
lowest
score,
driven
by
very
low
composite
scores
for
all
the
sub-indices,
except
environ-
mental.
For
example,
the
county
ranked
in
the
lowest
percentile
on
social
resilience
with
low
percentages
of
households
with
access
to
a
vehicle
and
a
telephone,
large
numbers
of
non-English
speakers
and
persons
without
health
insurance,
a
low
level
of
equality
in
educational
attainment,
few
physicians,
and
a
high
level
of
food
insecurity.
Our
second
research
question
posits
geographical
differentiation
of
community
resilience
that
is
based
on
the
spatial
variability
in
the
drivers
at
the
county
scale.
As
illustrated
above,
there
are
distinct
spatial
patterns
of
the
drivers
of
inherent
resilience,
which
in
turn
influence
the
geographical
distribution
of
places
with
high
and
low
levels
of
overall
community
resilience.
4.
Relationship
between
the
resilience
index
and
a
social
vulnerability
metric
Our
third
research
question
asked
how
disaster
resilience
and
social
vulnerability
relate
to
one
another
(Miller
et
al.,
2010;
Prior
and
Hagmann,
2014).
As
noted
early,
Cutter
et
al.
(2008a)
summarized
several
conceptualizations
of
resilience
and
vulnera-
bility
common
in
environmental
change
and
hazards
research:
resilience
as
a
component
of
vulnerability
(Turner
et
al.,
2003);
resilience
and
vulnerability
as
opposite
ends
of
a
spectrum
(Manyena,
2006;
Wilson,
2012);
or
resilience
and
vulnerability
as
discrete
concepts
with
some
unspecified
level
of
overlap
(Cutter
et
al.,
2008a;
Sherrieb
et
al.,
2010).
Construction
of
the
BRIC
provides
the
opportunity
to
demonstrate
quantitatively
the
specifics
of
the
proposed
relationship
between
inherent
social
vulnerability
and
inherent
disaster
resilience.
Previous
quantitative
resilience
indicator
research
provided
preliminary
findings
about
this
proposed
relationship.
Sherrieb
et
al.
(2010)
measured
the
adaptive
capacities
for
community
resilience
using
the
Norris
et
al.
(2008)
framework
for
Mississippi.
They
compared
their
resilience
model
to
a
measure
of
social
vulnerability,
the
Social
Vulnerability
Index
(SoVI
1
),
and
found
using
a
simple
linear
regression
model
that
there
was
some
overlap
Table
4
Ten
highest
and
lowest
resilience
scores
in
contiguous
US,
with
ranks
in
parentheses.
Rank
County
&
state
Resilience
score
Social
Economic
Housing
&
infrastructure
Community
capital
Institutional
Environmental
Most
resilient
1
St.
Charles,
LA
4.389
0.81
(181)
0.70
(176)
0.47
(1571)
0.61
(723)
1.00
(1)
0.80
(23)
2
Shelby,
IA
4.267
0.81
(196)
0.71
(121)
0.63
(247)
0.69
(163)
0.94
(3)
0.48
(2370)
3
St.
John
the
Baptist,
LA
4.176
0.66
(1499)
0.68
(267)
0.46
(1642)
0.60
(867)
0.97
(2)
0.81
(20)
4
McLeod,
MN
4.133
0.88
(20)
0.75
(42)
0.62
(304)
0.71
(117)
0.68
(94)
0.49
(1872)
5
Yellow
Medicine,
MN
4.122
0.78
(414)
0.65
(507)
0.65
(164)
0.70
(146)
0.78
(22)
0.56
(600)
6
Putnam,
OH
4.119
0.89
(13)
0.76
(33)
0.58
(544)
0.80
(32)
0.62
(229)
0.48
(2175)
7
Big
Stone,
MN
4.114
0.80
(274)
0.60
(1434)
0.82
(25)
0.74
(71)
0.66
(125)
0.54
(868)
8
Red
Lake,
MN
4.092
0.82
(167)
0.76
(32)
0.60
(384)
0.80
(33)
0.50
(1356)
0.63
(241)
9
Goodhue,
MN
4.087
0.86
(39)
0.69
(192)
0.56
(680)
0.70
(141)
0.75
(31)
0.52
(1219)
10
Winnebago,
WI
4.085
0.83
(125)
0.69
(195)
0.60
(429)
0.86
(11)
0.60
(286)
0.51
(1523)
Least
resilient
3108
Imperial,
CA
1.670
0.31
(3076)
0.40
(2763)
0.42
(2025)
0.32
(3081)
0.23
(3033)
0.0005
(3107)
3107
Presidio,
TX
1.674
0.10
(3106)
0.24
(3075)
0.27
(2843)
0.36
(3021)
0.17
(3083)
0.53
(1059)
3106
Esmeralda,
NV
1.776
0.42
(2972)
0.07
(3106)
0.27
(2850)
0.29
(3097)
0.24
(3023)
0.49
(2084)
3105
La
Paz,
AZ
1.816
0.40
(3014)
0.32
(2984)
0.09
(3090)
0.22
(3105)
0.37
(2701)
0.41
(2949)
3104
Hudspeth,
TX
1.875
0.06
(3107)
0.43
(2630)
0.31
(2699)
0.39
(2965)
0.20
(3068)
0.49
(1919)
3103
Daggett,
UT
1.925
0.66
(1536)
0.00
(3108)
0.25
(2918)
0.46
(2646)
0.27
(2970)
0.28
(3059)
3102
Catron,
NM
1.936
0.39
(3015)
0.18
(3094)
0.30
(2749)
0.39
(2950)
0.17
(3082)
0.50
(1563)
3101
Nye,
NV
1.944
0.43
(2944)
0.30
(3025)
0.20
(3011)
0.29
(3093)
0.27
(2972)
0.45
(2761)
3100
Eureka,
NV
1.987
0.68
(1387)
0.32
(2990)
0.25
(2916)
0.36
(3038)
0.003
(3107)
0.38
(2994)
3099
Greenlee,
AZ
2.017
0.73
(905)
0.09
(3104)
0.40
(2200)
0.46
(2636)
0.32
(2866)
0.02
(3106)
S.L.
Cutter
et
al.
/
Global
Environmental
Change
29
(2014)
65–77
73

between
the
two
constructs.
Significant
and
negative
correlations
existed
between
SoVI
1
and
their
measures
of
community
resilience,
economic
development,
and
social
capital;
the
greater
the
social
vulnerability
the
less
economic
development,
social
capital,
and
community
resilience.
The
linear
correlations
accounted
for,
at
best,
only
half
the
variance
in
the
relationships
between
social
vulnerability
and
resilience
measures.
Therefore,
their
indicators
were
consistent
with
the
assertion
that
vulnera-
bility
and
resilience
have
some
level
of
overlap,
but
are
not
simply
opposites.
Since
their
analysis
only
covered
Mississippi,
different
results
might
be
found
when
analyzing
the
contiguous
US.
We
also
compared
the
BRIC
score
to
the
same
social
vulnerability
metric
to
determine
the
extent
of
overlap
between
these
two
composite
indices.
The
metric
used
in
the
analysis
was
the
Social
Vulnerability
Index
(SoVI
1
)
for
the
years
2006–2010
(HVRI,
2014),
an
updated
version
of
the
SoVI
1
first
outlined
by
Cutter
et
al.
(2003).
It
incorporates
29
socioeconomic
variables
to
form
an
overall
index
of
vulnerability
to
environmental
hazards
in
the
US
at
the
county
level.
Using
principal
component
analysis
(PCA),
the
29
variables
are
condensed
into
seven
distinct
components,
each
capturing
a
dimension
of
social
vulnerability
in
the
US:
race/class;
wealth;
elderly
residents;
Latino/Hispanic
ethnicity;
special
needs
individuals;
Native
American
ethnicity;
and
service
sector
employment.
Each
county
has
a
score
for
each
component,
and
those
seven
scores
are
summed
to
comprise
the
overall
SoVI
1
scores.
There
is
a
statistically
significant
(
a
=
0.01)
negative
linear
relationship
between
the
resilience
index
and
SoVI
1
(Pearson
r
=
0.5;
R
2
=
0.25).
This
result
is
similar
in
direction,
though
somewhat
less
in
magnitude,
to
the
relationships
shown
by
Sherrieb
et
al.
(2010).
In
general,
when
the
social
vulnerability
metric
is
higher,
BRIC
tends
to
be
lower.
However,
the
relationship
is
relatively
weak
as
75%
of
the
variability
in
the
resilience
index
remains
unaccounted
for
by
SoVI
1
.
Clearly,
the
composite
resilience
index
(BRIC)
and
SoVI
1
are
not
simply
opposites
as
we
would
expect
a
stronger
negative
correlation,
but
do
contain
some
common
information.
Geographically,
the
results
indicate
that
BRIC
values
are
underestimated
relative
to
levels
of
social
vulnerability
over
much
of
the
central
and
northern
plains
and
into
the
Midwest,
as
well
as
along
the
lower
Mississippi
River
valley
(results
not
shown).
This
suggests
that
there
is
more
inherent
resilience
in
these
areas
than
what
would
be
predicted
based
on
the
current
levels
of
social
vulnerability.
Conversely,
in
the
western
US
the
BRIC
is
overestimated
using
social
vulnerability
suggesting
less
inherent
resilience
than
what
might
be
expected
given
the
levels
of
social
vulnerability.
To
further
explore
the
quality
of
the
relationship
between
our
disaster
resilience
and
social
vulnerability
indices,
a
regression
model
was
run
using
the
individual
SoVI
1
factors
instead
of
the
composite
score.
The
prediction
of
BRIC
was
much
better,
(R
2
=
0.64,
a
=
0.000),
with
all
seven
factors
statistically
significant
(Table
5).
The
most
important
predictors
were
ethnicity,
race,
and
class,
and
special
needs
populations.
The
Latino/Hispanic
ethnicity
component
had
a
strong
negative
linear
relationship
with
overall
resilience.
In
general,
resilience
scores
tended
to
be
higher
in
counties
where
there
were
lower
percentages
of
Latino/Hispanic
persons
and
higher
percentages
of
persons
proficient
in
the
English
language.
Other
components
also
had
negative
relationships
with
the
resilience
metric:
race
(black)
and
class,
elderly
residents,
service
sector
employment,
and
Native
American
ethnicity.
Broadly
speaking,
resilience
scores
were
higher
in
areas
where
there
were
lower
percentages
of
socially-disadvantaged
popula-
tions
–
black
persons,
persons
living
below
the
poverty
level,
persons
older
than
65
years
of
age,
employees
working
in
service
sector
jobs,
and
persons
of
Native
American
ethnicity.
Interesting-
ly,
there
is
a
positive
relationship
between
counties
with
higher
resilience
levels
and
those
with
higher
percentages
of
persons
with
special
needs,
including
persons
living
in
nursing
facilities
and
hospital
patients.
Finally,
there
is
also
a
weak
positive
association
between
the
wealth
component
of
SoVI
1
and
the
BRIC
score;
places
with
higher
incomes,
home
values,
and
rents
tended
to
have
higher
resilience
scores
than
places
with
lesser
values
on
these
variables.
By
examining
the
linear
statistical
relationships
between
BRIC
and
SoVI
1
,
we
show
that
our
indicator
structure
is
consistent
with
the
assertion
in
the
DROP
framework
that
inherent
vulnerability
and
resilience
contain
some
common
information
but
are
not
opposite
ends
of
a
continuous
spectrum.
When
using
the
contiguous
US
as
a
study
region
and
2010
data,
there
is
approximately
a
25%
overlap
in
our
composite
indicators,
a
finding
similar
to
Sherrieb
et
al.
(2010).
Breaking
out
the
vulnerability
components
for
a
separate
analysis
suggests
that
25%
overlap
consists
principally
of
social
and
economic
factors,
whereas
factors
related
to
infrastructure,
institutional,
environmental
resilience,
and
community
capital
are
less
consistent
with
social
vulnerability
indicators.
5.
Discussion
BRIC
provides
a
reference
point
or
baseline
for
examining
the
current
status
of
inherent
resilience
at
the
county
level.
Nearly
all
of
the
variables
can
be
freely
reproduced
by
state/local
entities
and
easily
manipulated
within
standard
computer
software.
While
there
is
no
critical
threshold
value
of
high
or
low
inherent
resilience,
communities
could
compare
their
values
to
neighboring
places
based
on
their
ranking
or
on
the
percentile
equivalent
of
the
score.
They
could
also
compare
the
BRIC
values
over
time
(2000,
2005,
2010,
etc.)
as
a
means
for
charting
progress
for
their
individual
county.
More
significantly,
BRIC
can
enable
community
leaders
to
ascertain
on
which
dimension
they
are
weakest,
providing
some
basis
for
targeted
interventions
for
improvement.
For
example,
take
several
counties
in
southern
Indiana
–
Martin,
Harrison,
Scott,
and
Pike.
Why
expend
resources
in
the
economic
domain,
in
which
these
counties
are
strong
(in
the
top
15%
nationally),
when
assets
could
be
used
to
improve
housing
and
infrastructure
resilience
(in
the
lowest
5%
nationally)?
The
BRIC
can
aid
decision-making
by
highlighting
where
certain
types
of
initiatives
or
programs
could
best
enhance
disaster
resilience.
It
should
be
noted
that
some
disaster
resilience
indicators
used
in
this
analysis
could
be
directly
targeted
by
policy
and
decision-
makers
for
improvement,
whereas
others
would
require
more
indirect
policy
measures.
In
other
words,
some
variables
are
actionable
and
can
improve
scores,
while
others
are
not.
The
social
resilience
category
is
rife
with
relevant
examples
of
the
latter.
No
one
would
suggest
that
Warren,
North
Carolina
–
having
the
lowest
percentage
of
non-disabled
persons
in
the
contiguous
US
–
should
focus
their
efforts
to
reduce
the
number
of
persons
with
a
disability.
Rather,
they
could
enact
programs
to
ensure
that
disabled
persons
have
the
knowledge
and
resources
to
better
prepare
for,
cope
with,
and
recover
from
disasters.
Such
a
program,
Table
5
SoVI
1
component
predictors
of
overall
composite
resilience
index.
SoVI
1
component
Beta
t
Sig.
Latino/hispanic
ethnicity
0.59
54.92
0.000
Race
(black)/class
0.31
28.52
0.000
Special
needs
persons
0.30
25.91
0.000
Elderly
residents
0.22
20.65
0.000
Service
sector
emp.
0.20
18.97
0.000
Native
American
Ethnicity
0.14
12.85
0.000
Wealth
0.09
8.30
0.000
S.L.
Cutter
et
al.
/
Global
Environmental
Change
29
(2014)
65–77
74

even
if
highly
successful,
would
not
alter
or
be
reflected
by
the
indicator
set
presented
in
this
paper.
On
the
other
hand,
indicators
of
educational
and
income
equality
may
be
more
readily
targeted
for
improvement.
Aware-
ness
programs
and
legislation
aimed
at
improving
race/ethnicity
or
gender-based
wage
gaps,
if
successful
over
time
would
be
directly
reflected
in
the
social
and
economic
resilience
categories.
Similarly,
many
of
the
infrastructure
indicators
could
be
directly
targeted
for
improvement.
Other
indicators
are
extremely
context-
specific
and
would
be
unlikely
targets
for
betterment.
If
a
large
rapid
onset
disaster
befell
Wayne
County,
Michigan
(Detroit),
there
would
be
a
relatively
large
number
of
vacant
rental
units
where
displaced
residents
could
temporarily
reside.
Yet
Detroit
and
surrounding
areas
are
not
likely
to
enact
policies
to
sustain
or
increase
the
number
of
vacant
rental
units
because
this
disaster
resilience
indicator
is
only
relevant
in
a
specific
context,
and
is
generally
undesirable
otherwise.
Thus,
while
BRIC
can
be
useful
in
guiding
policy
decisions,
not
every
individual
indicator
could
or
should
be
targeted
directly
for
improvement.
Future
work
should
investigate
ways
to
account
for
community
disaster
resilience
interventions
and
programs
that
differentiate
actionable
ways
to
improve
the
indicator
scores
from
those
indirect
efforts.
As
for
the
geography
of
BRIC,
the
Midwest
and
Northeast
US
have
higher
levels
of
inherent
resilience
than
counties
in
the
west
or
the
south.
The
region
known
as
the
Rust
Belt
–
the
former
manufacturing
region
including
western
New
York,
Pennsylvania,
Ohio,
southern
Michigan,
and
Indiana,
as
well
as
Chicago,
Illinois,
and
Milwaukee,
Wisconsin
–
shows
up
with
high
disaster
resilience
scores
relative
to
most
of
the
United
States.
This
may
seem
counterintuitive
given
the
economic
struggles
of
this
region
since
the
1970s.
Yet
this
region
possesses
many
inherent
characteristics
that
can
enhance
disaster
resilience,
especially
in
the
social
and
housing/infrastructure
resilience
categories,
as
well
as
community
capital.
In
addition,
there
are
some
positive
economic
aspects
such
as
income
equality
and
homeownership
that
serve
to
counteract
the
impact
of
lower
employment
rates
within
the
economic
resilience
category.
Another
notable
region
is
southeastern
Louisiana,
which
comes
out
as
moderately-high
to
high
in
terms
of
overall
disaster
resilience
despite
the
lower
scores
found
across
most
of
the
southern
US.
Within
this
region,
county
scores
for
institutional
resilience
are
nearly
all
above
the
85th
percentile
nationally,
largely
due
to
its
frequent
disasters
and
the
influx
of
mitigation
spending,
local
disaster
training,
and
flood
insurance,
all
of
which
help
to
build
the
institutional
capacity
for
resilience.
Coupled
with
reasonably
high
scores
on
community
capital
(many
above
the
66th
percentile
nationally),
southeast
Louisiana
ranks
higher
on
the
BRIC
than
would
be
expected
given
its
higher
social
vulnerability.
In
fact,
this
region
is
a
good
example
of
the
notion
that
increased
social
vulnerability
does
not
always
indicate
decreased
resilience.
While
the
social
and
economic
factors
contribute
to
higher
social
vulnerability
and
rather
average
social
and
economic
resilience
measures,
community
disaster
resilience
also
includes
other
dimensions
not
captured
in
social
vulnerability
–
environmental,
institutional,
housing/infrastructure,
and
com-
munity
capital.
In
the
case
of
southeastern
Louisiana,
great
assets
in
two
of
these
capitals
–
institutional
and
environmental
–
offset
the
lower
scores
on
social
and
economic
resilience
in
providing
an
integrated
snapshot
of
inherent
disaster
resilience.
As
a
group
of
indicators,
the
revised
BRIC
presented
here
approximates
the
abstract
concept
of
resilience
reasonably
well.
There
are
however,
some
discrepancies
between
the
conceptual
structure
of
resilience
(as
a
set
of
distinct
capitals
that
were
derived
from
the
literature)
and
the
quantitative
indicator
structure,
as
judged
by
statistical
tests
of
internal
consistency.
This
suggests
that
disaster
resilience
manifests
itself
in
a
myriad
of
ways
with
many
subtle
nuances,
and
this
can
be
difficult
to
adequately
represent
with
composite
indicators.
It
also
suggests
that
future
work
constructing
disaster
resilience
indicators
should
perhaps
reconsider
the
six
categories
drawn
from
previous
research
literature
and
used
in
this
analysis.
A
different
a
priori
resilience
category
schema
could
enhance
the
internal
validity
of
the
indicator
subsets.
Previously,
Cutter
et
al.
(2010)
noted
the
difficulty
in
specifying
indicators
for
the
entire
US.
For
example,
some
assumptions
about
community
resilience
in
a
national
context
will
be
less
applicable
in
some
local
contexts.
A
lower
rate
of
English
language
proficiency
is
less
important
in
Miami,
El
Paso,
or
San
Diego
where
emergency
information
is
available
in
Spanish
than
in
Nashville
or
Des
Moines
where
information
in
Spanish
is
less
available.
Similarly,
indicators
of
community
capital
such
as
civic
and
religious
organizations
and
Red
Cross
volunteers
and
training
participants
can
be
enhanced
if
additional
information
is
known
about
their
tendency
to
work
together
during
crises
or
to
address
otherwise
important
local
issues
relevant
during
multiple
phases
of
the
disaster
cycle.
Such
nuanced
information
will
seldom
appear
in
national
data
sources
or
analyses,
but
it
may
be
more
feasible
to
collect
and
use
such
data
at
the
sub-county
or
county
level
within
a
single
state.
The
collection
of
proxies
for
environmental/ecological
resil-
ience
was
very
problematic
at
the
national
scale.
Using
wetlands
as
a
proxy
indicator
for
natural
hazard
buffers
favors
those
places
exposed
to
coastal
or
riverine
flooding.
A
natural
buffer
indicator
for
the
contiguous
US
that
better
encompasses
all
environmental
hazards
would
improve
the
environmental
subset.
Furthermore,
using
a
single
time
period
snapshot
does
not
take
into
account
trends
which
may
be
vital
for
long-term
resilience.
Some
coastal
regions
may
have
large
areas
of
wetlands
to
serve
as
natural
flood
buffers,
but
what
about
those
regions
where
wetlands
have
been
receding
over
the
past
several
decades?
A
temporal
disaster
resilience
indicator
analysis
could
potentially
incorporate
trends
such
as
these.
The
indicator
for
local
food
supplies
likewise
has
some
deficiencies.
It
only
accounts
for
the
number
of
farms
marketing
products
via
community
supported
agriculture
in
relation
to
the
local
population.
This
does
not
account
for
the
sizes
of
farms,
the
volume
of
products,
or
the
level
of
diversity
of
agricultural
products
marketed
and
supported
locally.
These
other
aspects
could
be
important
to
consider
in
relation
to
the
quantity
and
quality
of
local
food
products
which
might
enhance
resilience
to
environmental
hazards.
Indeed,
specifying
and
finding
data
for
the
environmental
category
proved
to
be
one
of
the
most
challenging
aspects
of
this
research
and
is
an
aspect
that
would
benefit
from
further
refinement
in
future
iterations
of
BRIC.
6.
Conclusions
This
paper
had
three
main
objectives.
First,
it
defined
a
set
of
49
indicators
that
adequately
capture
the
multi-dimensional
and
multi-attribute
nature
of
inherent
disaster
resilience.
We
obtained
similar
results
in
the
final
composite
scores
using
two
different
methods:
a
deductive
binning
of
indicators
into
resilience
categories
and
subsequent
summations
within
and
across
the
categories,
and
an
inductive
indicator
method
constructed
via
principal
components
analysis
and
subsequent
summation
of
factor
scores
across
the
first
six
components.
Second,
there
is
a
distinct
spatial
pattern
to
inherent
disaster
resilience
with
counties
in
the
Midwest
and
eastern
Great
Plains
showing
the
greatest
resilience,
and
border
counties
in
Texas,
counties
in
Appalachia,
and
those
in
the
western
US
showing
the
least
resilience.
The
predominant
drivers
of
the
lower
inherent
resilience
are
lower
rankings
on
housing/infrastructure,
institution-
al,
community
capital,
and
environmental
resilience.
In
fact,
many
S.L.
Cutter
et
al.
/
Global
Environmental
Change
29
(2014)
65–77
75

of
the
least
resilient
counties
are
in
the
lower
20%
on
all
of
these
components,
suggesting
a
myriad
of
opportunities
for
enhancing
resilience.
Finally,
the
paper
demonstrated
quantitatively
that
our
social
vulnerability
and
disaster
resilience
indicators
are
not
the
inverse
of
one
another.
While
there
is
some
statistical
overlap
between
the
two
constructs
(roughly
25%),
inherent
disaster
resilience
is
a
broader
and
distinct
construct
compared
to
social
vulnerability
(as
quantified
using
SoVI
1
).
This
is
consistent
with
the
disaster
resilience
of
place
model,
our
conceptual
framework
for
this
research.
Social
vulnerability
most
closely
parallels
disaster
resilience
in
the
social
and
economic
resilience
categories.
It
is
also
of
note
that
BRIC
is
a
place-based
construct
and
as
such
allows
for
visualization
of
the
spatial
variability
in
the
underlying
dimensions
as
they
are
manifest
differently
across
the
US.
A
final
caveat
is
that
this
paper
did
not
address
the
persistent
difficulty
in
validating
whether
quantitative
indicator
models
of
social
vulnerability
or
disaster
resilience
truly
represent
such
abstract
concepts.
This
remains
an
important
open
question.
Nevertheless,
such
a
quantitative
analysis
has
a
wider
application
for
the
policy
world.
In
addition
to
providing
a
benchmark
for
measuring
progress,
BRIC
is
useful
for
assessing
needs
and
targeting
intervention
or
mitigation
programs.
It
provides
a
rationale
for
policy
makers
to
use
scarce
resources
more
strategically
to
maximize
their
effect
rather
than
equally
distributing
them
across
all
places
and
programs.
At
the
same
time,
it
affords
the
opportunity
to
see
whether
or
not
deployment
of
said
resources
actually
make
a
difference
in
enhancing
resilience
by
recalculating
the
BRIC
score
a
few
years
later
to
monitor
change.
In
this
way,
accountability
for
tracking
disaster
resilience
efforts
can
be
maintained
and
compared
by
state,
region,
or
nationally
along
with
the
resources
provided
for
such
enhancements.
References
Adger,
W.N.,
2005.
Social-ecological
resilience
to
coastal
disasters.
Science
309,
1036–1039.
Alawiyah,
T.,
Bell,
H.,
Pyles,
L.,
Runnels,
R.C.,
2011.
Spirituality
and
faith-based
interventions:
pathways
to
disaster
resilience
for
African
American
Hurri-
cane
Katrina
survivors.
J.
Relig.
Spiritual.
Soc.
Work:
Soc.
Thought
30
(3)
294–319.
Aldrich,
D.P.,
2012.
Building
Resilience:
Social
Capital
in
Post-Disaster
Recovery.
University
of
Chicago
Press,
Chicago.
Ansell,
C.,
Boin,
A.,
Keller,
A.,
2010.
Managing
transboundary
crises:
identifying
the
building
blocks
of
an
effective
response
system.
J.
Conting.
Crisis
Manag.
18,
195–207.
Barnett,
J.,
Lambert,
S.,
Fry,
I.,
2008.
The
hazards
of
indicators:
insights
from
the
Environmental
Vulnerability
Index.
Ann.
Assoc.
Am.
Geogr.
98,
102–119.
Beatley,
T.,
Newman,
P.,
2013.
Biophilic
cities
are
sustainable,
resilient
cities.
Sustainability
5,
3328–3345.
Berardi,
G.,
Green,
R.,
Hammond,
B.,
2011.
Stability,
sustainability,
and
catastro-
phe:
applying
resilience
thinking
to
U.S.
agriculture.
Hum.
Ecol.
Rev.
18,
115–125.
Birkmann,
J.,
Cardona,
O.D.,
Carren
˜o,
M.L.,
Barbat,
A.H.,
Pelling,
M.,
Schneiderbauer,
S.,
Kienberger,
S.,
Keiler,
M.,
Alexander,
D.,
Zeil,
P.,
Welle,
T.,
2013.
Framing
vulnerability,
risk
and
societal
responses:
the
MOVE
framework.
Nat.
Hazards
67,
193–211.
Bowman,
A.O’M.,
Parsons,
B.M.,
2009.
Vulnerability
and
resilience
in
local
govern-
ment:
assessing
the
strength
of
performance
regimes.
State
Local
Gov.
Rev.
41,
13–24.
Brody,
S.D.,
Peacock,
W.G.,
Gunn,
J.,
2012.
Ecological
indicators
of
flood
risk
along
the
Gulf
of
Mexico.
Ecol.
Indic.
18,
493–500.
Brody,
S.,
Blessing,
R.,
Sebastian,
A.,
Bedient,
P.,
2014.
Examining
the
impact
of
land
use/land
cover
characteristics
on
flood
losses.
J.
Environ.
Plan.
Manag.
57
(8)
1252–1265.
Bruneau,
M.,
Chang,
S.E.,
Eguchi,
R.T.,
Lee,
G.C.,
O’Rourke,
T.D.,
Reinhorn,
A.M.,
Shinozuka,
M.,
Tierney,
K.,
Wallace,
W.A.,
von
Winterfeldt,
D.,
2003.
A
frame-
work
to
quantitatively
assess
and
enhance
the
seismic
resilience
of
communi-
ties.
Earthq.
Spectra
19,
733–752.
Burger,
J.,
Gochfeld,
M.,
Jeitner,
C.,
Pittfield,
T.,
Donio,
M.,
2013.
Trusted
information
sources
used
during
and
after
Superstorm
Sandy:
TV
and
radio
were
used
more
often
than
social
media.
J.
Toxicol.
Environ.
Health
A:
Curr.
Issues
76,
1138–1150.
Chandra,
A.,
Acosta,
J.,
Stern,
S.,
Uscher-Pines,
L.,
Williams,
M.,
Yeung,
D.,
Garnett,
J.,
Meredith,
L.,
2011.
Building
Community
Resilience
to
Disasters:
A
Way
Forward
to
Enhance
National
Health
Security.
RAND
Corporation
Technical
Report
US
Dept.
of
Health
and
Human
Services,
Washington,
DC.
Cheong,
S.-M.,
2011.
Policy
solutions
in
the
US.
Clim.
Change
106,
57–70.
Cimellaro,
G.P.,
Reinhorn,
A.M.,
Bruneau,
M.,
2010.
Seismic
resilience
of
a
hospital
system.
Struct.
Infrastruct.
Eng.:
Maint.
Manag.
Life-Cycle
Des.
Perform.
6,
127–144.
Cross,
J.A.,
2014.
Disaster
devastation
of
US
communities:
long-term
demographic
consequences.
Environ.
Hazards
13,
73–91.
Cutter,
S.L.,
Boruff,
B.J.,
Shirley,
W.L.,
2003.
Social
vulnerability
to
environmental
hazards.
Soc.
Sci.
Q.
84,
242–261.
Cutter,
S.L.,
Barnes,
L.,
Berry,
M.,
Burton,
C.,
Evans,
E.,
Tate,
E.,
Webb,
J.,
2008a.
A
place-based
model
for
understanding
community
resilience
to
natural
disas-
ters.
Glob.
Environ.
Change
18,
598–606.
Cutter,
S.L.,
Barnes,
L.,
Berry,
M.,
Burton,
C.G.,
Evans,
E.,
Tate,
E.C.,
Webb,
J.,
2008b.
Community
and
Regional
Resilience:
Perspectives
from
Hazards,
Disasters,
and
Emergency
Management.
CARRI
Research
Report
1
Community
and
Regional
Resilience
Institute,
Oak
Ridge,
TN.
Cutter,
S.L.,
Burton,
C.G.,
Emrich,
C.T.,
2010.
Disaster
resilience
indicators
for
benchmarking
baseline
conditions.
J.
Homel.
Secur.
Emerg.
Manag.
7,
Article
51,
1–22.
Davis,
E.,
Phillips,
B.,
2009.
Effective
Emergency
Management:
Making
Improve-
ments
for
Communities
and
People
with
Disabilities.
National
Council
on
Disability,
Washington,
DC.
Emmer,
R.,
Swann,
L.,
Schneider,
M.,
Sempier,
S.,
Sempier,
T.,
Sanchez,
T.,
2008.
Coastal
Resilience
Index:
A
Community
Self-Assessment.
A
Guide
to
Examining
How
Prepared
Your
Community
is
for
a
Disaster.
NOAA
Publ.
No.
MAS
GP-08-014.
Available
at
http://research.fit.edu/sealevelriselibrary/documents/doc_mgr/434/
Gulf_Coast_Coastal_Resilience_Index_-_SeaGrant.pdf.
Enarson,
E.,
2012.
Women
Confronting
Natural
Disaster:
From
Vulnerability
to
Resilience.
Lynne
Rienner,
Boulder,
CO.
Fe
´lix,
D.,
Branco,
J.M.,
Feio,
A.,
2013.
Temporary
housing
after
disasters:
a
state
of
the
art
survey.
Habitat
Int.
40,
136–141.
Florida,
R.,
2002.
The
Rise
of
the
Creative
Class.
Basic
Books,
New
York,
NY.
Flynn,
S.E.,
2007.
The
Edge
of
Disaster:
Rebuilding
a
Resilient
Nation.
Random
House,
New
York.
Frazier,
T.G.,
Thompson,
C.M.,
Dezzani,
R.J.,
Butsick,
D.,
2013.
Spatial
and
temporal
quantification
of
resilience
at
the
community
scale.
Appl.
Geogr.
42,
95–107.
Gladwin,
B.K.,
Smith,
V.H.,
2013.
What
harm
is
done
by
subsidizing
crop
insurance?
Am.
J.
Agric.
Econ.
95,
489–497.
Glauber,
J.W.,
2013.
The
growth
of
the
federal
crop
insurance
program,
1990–2011.
Am.
J.
Agric.
Econ.
95,
482–488.
Godschalk,
D.R.,
2003.
Urban
hazard
mitigation:
creating
resilient
cities.
Nat.
Hazards
Rev.
4,
136–143.
Godschalk,
D.R.,
Rose,
A.,
Mittler,
E.,
Porter,
K.,
West,
C.T.,
2009.
Estimating
the
value
of
foresight:
aggregate
analysis
of
natural
hazard
mitigation
benefits
and
costs.
J.
Environ.
Plan.
Manag.
52,
739–756.
Gunderson,
L.,
2010.
Ecological
and
human
community
resilience
in
response
to
natural
disasters.
Ecol.
Soc.
15
,
Article
18,
1–11.
Harrald,
J.,
2012.
The
case
for
resilience:
a
comparative
analysis.
Int.
J.
Crit.
Infra-
struct.
8,
3–21.
Haveman,
R.,
Wolff,
E.N.,
2005.
The
concept
and
measurement
of
asset
poverty:
levels,
trends
and
composition
for
the
U.S.,
1983–2001.
J.
Econ.
Inequal.
2,
145–169.
Hazards
and
Vulnerability
Research
Institute,
2014.
Social
Vulnerability
Index
for
the
United
States
–
2006–10.
Hazards
and
Vulnerability
Research
Institute,
In:
http://webra.cas.sc.edu/hvri/products/sovi.aspx.
Heinz
Center,
2002.
Human
Links
to
Coastal
Disasters.
The
H.
John
Heinz
III
Center
for
Science,
Economics
and
the
Environment,
Washington,
DC.
Jaeger,
M.,
Konrad,
A.,
Rueegg,
S.,
Rabenschlag,
F.,
2013.
Measuring
recovery:
validity
of
the
‘‘Recovery
Process
Inventory’’
and
the
‘‘Recovery
Attitudes
Questionnaire’’.
Psychiatr.
Res.
210,
363–367.
Johnson,
C.,
2007.
Strategic
planning
for
post-disaster
temporary
housing.
Disasters
31,
435–458.
Kloos,
B.,
Hill,
J.,
Thomas,
E.,
Wandersman,
A.,
Elias,
M.J.,
Dalton,
J.H.,
2012.
Community
Psychology:
Linking
Individuals
and
Communities.
Wadsworth,
Belmont,
CA.
Kulig,
J.C.,
Edge,
D.S.,
Townshend,
I.,
Lightfoot,
N.,
Reimer,
W.,
2013.
Community
resiliency:
emerging
theoretical
insights.
J.
Community
Psychol.
41,
758–775.
Leichenko,
R.,
2011.
Climate
change
and
urban
resilience.
Curr.
Opin.
Environ.
Sustain.
3,
164–168.
Manyena,
S.B.,
2006.
The
concept
of
resilience
revisited.
Disasters
30,
433–450.
Manyena,
S.B.,
O‘Brien,
G.,
O‘Keefe,
P.,
Rose,
J.,
2011.
Disaster
resilience:
a
bounce
back
or
a
bounce
forward
ability.
Local
Environ.
16,
417–424.
Martin,
C.R.,
Savage-McGlynn,
E.,
2013.
A
‘good
practice’
guide
for
the
reporting
of
design
and
analysis
for
psychometric
evaluation.
J.
Reprod.
Infant
Psychol.
31,
449–455.
Matherly,
D.,
Mobley,
J.,
2012.
Transportation
and
emergency
management
tool
kit
for
communications
with
vulnerable
populations.
Transp.
Res.
Rec.:
J.
Transp.
Res.
Board
2234,
62–70.
McGuire,
M.,
Silvia,
C.,
2010.
The
effect
of
problem
severity,
managerial
and
organizational
capacity,
and
agency
structure
on
intergovernmental
collabo-
ration:
evidence
from
local
emergency
management.
Public
Admin.
Rev.
70,
279–288.
McManus,
P.,
Walmsley,
J.,
Argent,
N.,
Baum,
S.,
Bourke,
L.,
Martin,
J.,
Pritchard,
B.,
Sorensen,
T.,
2012.
Rural
community
and
rural
resilience:
what
is
important
to
farmers
in
keeping
their
country
towns
alive?
J.
Rural
Stud.
28,
20–29.
S.L.
Cutter
et
al.
/
Global
Environmental
Change
29
(2014)
65–77
76

Messias,
D.K.H.,
Barrington,
C.,
Lacy,
E.,
2012.
Latino
social
network
dynamics
and
the
Hurricane
Katrina
disaster.
Disasters
36,
101–121.
Michel-Kerjan,
E.,
de
Forges,
S.L.,
Kunreuther,
H.,
2012.
Policy
tenure
under
the
U.S.
National
Flood
Insurance
Program
(NFIP).
Risk
Anal.
32,
644–658.
Miles,
S.B.,
Chang,
S.E.,
2011.
ResilUS:
a
community
based
disaster
resilience
model.
Cartogr.
Geogr.
Inf.
Sci.
38,
36–51.
Mileti,
D.S.,
1999.
Disasters
by
Design.
Joseph
Henry
Press,
Washington,
DC.
Miller,
F.,
Osbahr,
E.,
Boyd,
E.,
Thomalla,
F.,
Bharwani,
S.,
Ziervogel,
G.,
Walker,
B.,
Birkmann,
J.,
van
der
Leeuw,
S.,
Rockstro
¨m,
J.,
Hinkel,
J.,
Downing,
T.,
Folke,
C.,
Nelson,
D.,
2010.
Resilience
and
vulnerability:
complementary
or
conflicting
concepts?
Ecol.
Soc.
15,
Article
11,
1–25.
Morrow,
B.,
2008.
Community
Resilience:
A
Social
Justice
Perspective.
CARRI
Research
Report
4
Community
and
Regional
Resilience
Institute,
Oak
Ridge,
TN.
Mowbray,
C.T.,
Woolley,
M.E.,
Grogan-Kaylor,
A.,
Gant,
L.M.,
Gilster,
M.E.,
Shanks,
T.R.W.,
2007.
Neighborhood
research
from
a
spatially
oriented
strengths
perspective.
J.
Community
Psychol.
35,
667–680.
Murphy,
B.L.,
2007.
Locating
social
capital
in
resilient
community-level
emergency
management.
Nat.
Hazards
41,
297–315.
Nardo,
M.,
Saisana,
M.,
Saltelli,
A.,
Tarantola,
S.,
2008.
Handbook
on
Constructing
Composite
Indicators:
Methodology
and
User
Guide.
OECD
Publishing,
Paris,
France.
Norris,
F.H.,
Stevens,
S.P.,
Pfefferbaum,
B.,
Wyche,
K.F.,
Pfefferbaum,
R.L.,
2008.
Community
resilience
as
a
metaphor,
theory,
set
of
capacities,
and
strategy
for
disaster
readiness.
Am.
J.
Community
Psychol.
41,
127–150.
O‘Brien,
K.,
Hayward,
B.,
Berkes,
F.,
2009.
Rethinking
social
contracts:
building
resilience
in
a
changing
climate.
Ecol.
Soc.
14,
Article
12,
1–17.
Paton,
D.,
Johnston,
D.,
2006.
Disaster
Resilience:
An
Integrated
Approach.
Charles
C.
Thomas,
Springfield,
IL.
Paton,
D.,
Tedim,
F.
(Eds.),
2012.
Wildfire
and
Community:
Facilitating
Prepared-
ness
and
Resilience.
Charles
C.
Thomas,
Springfield,
IL.
Peacock,
W.G.,
Brody,
S.D.,
Seitz,
W.A.,
Merrell,
W.J.,
Vedlitz,
A.,
Zahran,
S.,
Harriss,
R.C.,
Stickney,
R.,
2010.
Advancing
the
resilience
of
coastal
localities:
develop-
ing,
implementing
and
sustaining
the
use
of
coastal
resilience
indicators.
Final
report
for
NOAA
CSC
grant
no.
NA07NOS4730147.
Peek,
L.,
2010.
Age.
In:
Phillips,
B.D.,
Thomas,
D.S.K.,
Fothergill,
A.,
Blinn-Pike,
L.
(Eds.),
Social
Vulnerability
to
Disasters.
CRC
Press,
Boca
Raton,
FL,
pp.
155–185.
Pelling,
M.,
2003.
The
Vulnerability
of
Cities:
Social
Resilience
and
Natural
Disaster.
Earthscan,
London.
Pendall,
R.,
Theodos,
B.,
Franks,
K.,
2012.
Vulnerable
people,
precarious
housing,
and
regional
resilience:
an
exploratory
analysis.
Hous.
Policy
Debate
22,
271–296.
Pietrzak,
R.H.,
Tracy,
M.,
Galea,
S.,
Kilpatrick,
D.G.,
Ruggiero,
K.J.,
Hamblen,
J.L.,
Southwick,
S.M.,
Norris,
F.H.,
2012.
Resilience
in
the
face
of
disaster:
prevalence
and
longitudinal
course
of
mental
disorders
following
Hurricane
Ike.
PLoS
ONE
7,
e38964,
http://dx.doi.org/10.1371/journal.pone.0038964.
Pingali,
P.,
Alinovi,
L.,
Sutton,
J.,
2005.
Food
security
in
complex
emergencies:
enhancing
food
system
resilience.
Disasters
29,
S5–S24.
Plough,
A.,
Fielding,
J.E.,
Chandra,
A.,
Williams,
M.,
Eisenman,
D.,
Wells,
K.B.,
Law,
G.Y.,
Fogleman,
S.,
Magan
˜a,
A.,
2013.
Building
community
disaster
resilience:
perspectives
from
a
large
urban
county
Department
of
Public
Health.
Am.
J.
Public
Health
103,
1190–1197.
Prior,
T.,
Hagmann,
J.,
2014.
Measuring
resilience:
methodological
and
political
challenges
of
a
trend
security
concept.
J.
Risk
Res.
17,
281–298.
Renschler,
C.S.,
Frazier,
A.E.,
Arendt,
L.A.,
Cimellaro,
G.P.,
Reinhorn,
A.M.,
Bruneau,
M.,
2010.
A
Framework
for
Defining
and
Measuring
Resilience
at
the
Commu-
nity
Scale:
The
PEOPLES
Resilience
Framework.
MCEER
Technical
Report-
MCEER-10-0006
Multidisciplinary
Center
for
Earthquake
Engineering
Research,
Buffalo,
NY.
Ritchie,
L.A.,
Gill,
D.A.,
2007.
Social
capital
theory
as
an
integrating
framework
for
technological
disaster
research.
Sociol.
Spectr.
27,
1–26.
Ronan,
K.R.,
Johnston,
D.M.,
2005.
Promoting
Community
Resilience
in
Disaster:
The
Role
for
Schools,
Youth
and
Families.
Springer,
New
York.
Rose,
A.,
2007.
Economic
resilience
to
natural
and
man-made
disasters:
multidisci-
plinary
origins
and
contextual
dimensions.
Environ.
Hazards
7
(4)
383–398.
Rose,
A.,
Krausmann,
E.,
2013.
An
economic
framework
for
the
development
of
a
resilience
index
for
business
recovery.
Int.
J.
Disaster
Risk
Reduct.
5,
73–83.
Senkbeil,
J.C.,
Scott,
D.A.,
Guinazu-Walker,
P.,
Rockman,
M.S.,
2013.
Ethnic
and
racial
differences
in
tornado
hazard
perception,
preparedness,
and
shelter
lead
time
in
Tuscaloosa.
Prof.
Geogr.,
http://dx.doi.org/10.1080/00330124.2013.826562.
Setbon,
M.,
Raude,
J.,
2010.
Factors
in
vaccination
intention
against
the
pandemic
influenza
A/H1N1.
Eur.
J.
Public
Health
20,
490–494.
Sherrieb,
K.,
Norris,
F.H.,
Galea,
S.,
2010.
Measuring
capacities
for
community
resilience.
Soc.
Indic.
Res.
99,
227–247.
Sherrieb,
K.,
Louis,
C.A.,
Pfefferbaum,
R.L.,
Pfefferbaum,
J.D.B.,
Diab,
E.,
Norris,
F.H.,
2012.
Assessing
community
resilience
on
the
US
coast
using
school
principals
as
key
informants.
Int.
J.
Disaster
Risk
Reduct.
2,
6–15.
Simonovich,
J.,
Sharabi,
M.,
2013.
Dealing
with
environmental
disaster:
the
inter-
vention
of
community
emergency
teams
(CET)
in
the
2010
Israeli
forest
fire
disaster.
J.
Sustain.
Dev.
6,
86–93.
Springgate,
B.F.,
Wennerstrom,
A.,
Meyers,
D.,
Allen
III,
C.E.,
Vannoy,
S.D.,
Bentham,
W.,
Wells,
K.B.,
2011.
Building
community
resilience
through
mental
health
infrastructure
and
training
in
post-Katrina
New
Orleans.
Ethnic.
Dis.
21
,
S1-1–
S1-29.
Strawderman,
L.,
Salehi,
A.,
Babski-Reeves,
K.,
Thornton-Neaves,
T.,
Cosby,
A.,
2012.
Reverse
911
as
a
complementary
evacuation
warning
system.
Nat.
Hazards
Rev.
13,
65–73.
Sutter,
D.,
Simmons,
K.M.,
2010.
Tornado
fatalities
and
mobile
homes
in
the
United
States.
Nat.
Hazards
53,
125–137.
Tarabusi,
E.C.,
Guarini,
G.,
2012.
An
unbalance
adjustment
method
for
development
indicators.
Soc.
Indic.
Res.
112,
19–45.
Theckethil,
R.,
2006.
Building
codes:
a
regulatory
mechanism
for
reducing
the
vulnerability
of
urban
areas.
J.
Secur.
Educ.
1,
95–106.
Tierney,
K.,
2009.
Disaster
Response:
Research
Findings
and
Their
Implications
for
Resilience
Measurement.
CARRI
Research
Report
6
Community
and
Regional
Resilience
Institute,
Oak
Ridge,
TN.
Tierney,
K.,
2012.
Disaster
governance:
social,
political,
and
economic
dimensions.
Ann.
Rev.
Environ.
Resour.
37,
341–363.
Tierney,
K.,
Bruneau,
M.,
2007.
Conceptualizing
and
Measuring
Resilience:
A
Key
to
Disaster
Loss
Reduction.
TR
News
(May-June)
14–17.
Tobin,
G.A.,
Whiteford,
L.M.,
2012.
Provisioning
capacity:
a
critical
component
of
vulnerability
and
resilience
under
chronic
volcanic
eruptions.
In:
Pfeifer,
K.,
Pfeifer,
N.
(Eds.),
Forces
of
Nature
and
Cultural
Responses.
Springer,
New
York,
pp.
136–166.
Turner,
B.L.,
2010.
Vulnerability
and
resilience:
coalescing
or
paralleling
approaches
for
sustainability
science?
Glob.
Environ.
Change
20,
570–576.
Turner,
B.L.,
Kasperson,
R.E.,
Matson,
P.A.,
McCarthy,
J.J.,
Corell,
R.W.,
Christensen,
L.,
Eckley,
N.,
Kasperson,
J.X.,
Luers,
A.,
Martello,
M.L.,
Polsky,
C.,
Pulsipher,
A.,
Schiller,
A.,
2003.
A
framework
for
vulnerability
analysis
in
sustainability
science.
Proc.
Natl.
Acad.
Sci.
U.
S.
A.
100,
8074–8079.
United
Kingdom
Foresight,
2012.
Foresight
Reducing
Risks
of
Future
Disasters:
Priorities
for
Decision
Makers,
Final
Project
Report.
Government
Office
for
Science,
London.
United
Nations
Department
of
Economic
and
Social
Affairs
(UNDESA),
2007.
Indi-
cators
of
Sustainable
Development:
Guidelines
and
Methodologies.
United
Nations
Press,
New
York.
United
Nations
Office
for
Disaster
Risk
Reduction
(UNISDR),
2012.
Making
Cities
Resilient
Report
2012.
United
Nations
Office
for
Disaster
Risk
Reduction,
Geneva.
United
States
National
Research
Council
(US
NRC),
2012.
Disaster
Resilience:
A
National
Imperative.
The
National
Academies
Press,
Washington,
DC.
Urbanik
II,
T.,
2000.
Evacuation
time
estimates
for
nuclear
power
plants.
J.
Hazard.
Mater.
75,
165–180.
Vale,
L.J.,
Campanella,
T.J.
(Eds.),
2004.
The
Resilient
City:
How
Modern
Cities
Recover
from
Disaster.
Oxford
University
Press,
New
York.
Walsh,
F.,
2007.
Traumatic
loss
and
major
disasters:
strengthening
family
and
community
resilience.
Fam.
Process
46,
207–227.
Waugh,
W.L.,
1994.
Regionalizing
emergency
management:
counties
as
state
and
local
government.
Public
Admin.
Rev.
54,
253–258.
Wein,
A.,
Rose,
A.,
2011.
Economic
resilience
lessons
from
the
ShakeOut
earthquake
scenario.
Earthq.
Spectra
27,
559–573.
Whitman,
Z.R.,
Wilson,
T.M.,
Seville,
E.,
Vargo,
J.,
Stevenson,
J.R.,
Kachali,
H.,
Cole,
J.,
2013.
Rural
organizational
impacts,
mitigation
strategies,
and
resilience
to
the
2010
Darfield
earthquake,
New
Zealand.
Nat.
Hazards
69,
1849–1875.
Wilson,
G.A.,
2012.
Community
resilience,
globalization,
and
transitional
pathways
of
decision-making.
Geoforum
43,
1218–1231.
Zhou,
H.,
Wang,
J.,
Wan,
J.,
Jia,
H.,
2010.
Resilience
to
natural
hazards:
a
geographic
perspective.
Nat.
Hazards
53,
21–41.
S.L.
Cutter
et
al.
/
Global
Environmental
Change
29
(2014)
65–77
77