See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/241382670

Measuring resilience: Methodological and political challenges of a trend

security concept

Article  in  Journal of Risk Research · June 2013

DOI: 10.1080/13669877.2013.808686

CITATIONS READS

87 1,014

2 authors, including:

Tim Prior
Swisscom
68 PUBLICATIONS   1,632 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Migration and Civil Protection in Europe – Current challenges and long-term implications View project

Mapping Social Vulnerability in Affluent Contexts View project

All content following this page was uploaded by Tim Prior on 26 May 2014.

The user has requested enhancement of the downloaded file.

This article was downloaded by: [ETH Zurich]
On: 25 June 2013, At: 02:22
Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Risk Research

Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/rjrr20

Measuring resilience: methodological

and political challenges of a trend
security concept
a a
Timothy Prior & Jonas Hagmann
a
Center for Security Studies, ETH Zürich , Zürich , Switzerland
Published online: 25 Jun 2013.

To cite this article: Timothy Prior & Jonas Hagmann (2013): Measuring resilience:
methodological and political challenges of a trend security concept, Journal of Risk Research,
DOI:10.1080/13669877.2013.808686

To link to this article: http://dx.doi.org/10.1080/13669877.2013.808686

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-

conditions

This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation
that the contents will be complete or accurate or up to date. The accuracy of any
instructions, formulae, and drug doses should be independently verified with primary
sources. The publisher shall not be liable for any loss, actions, claims, proceedings,
demand, or costs or damages whatsoever or howsoever caused arising directly or
indirectly in connection with or arising out of the use of this material.
 Journal of Risk Research, 2013
http://dx.doi.org/10.1080/13669877.2013.808686

Measuring resilience: methodological and political challenges of a

trend security concept
Timothy Prior* and Jonas Hagmann

Center for Security Studies, ETH Zürich, Zürich, Switzerland

(Received 9 May 2012; final version received 30 April 2013)

Modern societies are characterised by global connectedness and complexity. At

the same time society, and the various infrastructures that connect and define it,
are understood to be increasingly threatened by unpredictable and uncertain (or
unknown) global risks. With this, the conceptualisation and development of resil-
Downloaded by [ETH Zurich] at 02:22 25 June 2013

ience has become a dominant, yet enigmatic preoccupation: dominant because it

is seen as a fundamental component of devolved proactive approaches to mitigat-
ing complex threats whatever their nature; and enigmatic because its practical
application is as diverse as its definitions. Today, however, a significant challenge
still lies in the accurate characterisation and quantification of resilience, and thus
also the ability to provide a systematic basis for policy-making in resilience-based
threat mitigation. This article examines the methodological challenges of opera-
tionalising resilience. It draws on several cases that detail ways of measuring resil-
ience, reflecting on the development, benefits and limitations of these and
highlighting important considerations pertinent in the construction of resilience
indices. Doing so, however, the article also maintains that resilience should not be
reduced to a methodological problem only, given that the methodological opera-
tionalisation of resilience also connects with analytical ideas of what and whose
kind of responsibility should be measured and political conceptions of who
assumes what tasks and responsibility in a resilience framework.
Keywords: resilience; measurement; politics; concept operationalisation;
security

1. Introduction
Modern western societies are characterised by global connectedness. This connect-
edness has increased the number and complexity of services supporting societies,
and has increased societal sensitivity to disturbances that might threaten the delivery
of those services. Information, communication, energy and trading systems, for
instance, are based on a variety of transnational networks today, meaning national
societies are increasingly, and directly dependent on the uninterrupted performance
of such systems. Considering this realisation yields firstly a hint of the possible con-
sequences for society if these society-supporting features are lost; and secondarily,
the necessity to ensure that these features are preserved or maintained – in order to
avoid disruptions in economic production or even social order. Taking action to mit-
igate the possible consequences of disruptions to society (whether resulting from

*Corresponding author. Email: tim.prior@sipo.gess.ethz.ch

Ó 2013 Taylor & Francis

 2 T. Prior and J. Hagmann

natural, social, economic, cultural or technological threats or shocks) by whatever

mechanisms appropriate and available (whether through crisis management or other
critical infrastructure protection policies) then becomes a fundamental goal for the
members of society with the mandate or capacity for protection and preservation
(Hagmann, Dunn-Cavelty, and Wenger 2011).
However, at a time when societies are characterised by global connections and
interdependencies and the social obligations of central governments seem to be
diminishing (Coaffee and Murakami Wood 2006; Florin 2011; Lund Petersen
2012a), threat is evermore characterised by global risks and unforeseen (or
unknown) challenges (Daase and Kessler 2007). Although many of these hazards
can be expected, their actual occurrence is uncertain and systemic planning for pre-
paredness is seen as imperative in the governance of risk and associated social secu-
rity (Adger 2006; Clarke and Chenoweth 2006). This is particularly the case
because most risks (and particularly those with security implications) can arguably
not be prevented, and therefore must be managed proactively, highlighting the need
for a precautionary approach to governing these risks and their implications for
Downloaded by [ETH Zurich] at 02:22 25 June 2013

society, its structures and its interdependencies (Aradau and Van Munster 2007;
Lund Petersen 2012b).
Within the discourse of precautionary risk, social disruption and security, the
concept of resilience has become a dominant and often rhetorical maxim in Europe
and North America. A polysemous and malleable term, it has worked its way from
disciplines as diverse as engineering and ecology to psychology and security – but
with differing depths of exploration and theorisation at each stop. Resilience is
derived from the Latin ‘resiliere’, meaning to spring or bounce back, and at its most
basic interpretation, is often applied in relation to an entity or system’s ability to
return to normal functioning quickly following a disturbance. A more nuanced con-
ception of resilience moves away from the idea of unquestioned equilibrium,
embracing change through adaptation and learning to ensure the system’s functions
persist. Studies on the meanings and expression of resilience are legion, highlight-
ing the conceptual vagueness of the term and the difficulty in applying resilience in
a robust and meaningful manner (Waller 2001; Walker et al. 2004; Folke 2006;
Haimes 2009; Zhou et al. 2010; Walker and Cooper 2011).
Although confusing, and at times contradictory, the popularity of the concept
has not been diminished. It is used to characterise individuals and communities,
organisations and infrastructure and cities and ecological systems. Such varied use,
but normative application, has resulted in a broad resilience vernacular – one that is
rarely made explicit, whether in the context of academic or policy settings. Confu-
sion need not taint the utility of resilience as a concept (Strunz 2012),1 but given
that many threats or disruptions to society must increasingly be addressed in a sys-
temic manner, across disciplines, confusion and contradiction pose methodological
limits for the operationalisation of resilience approaches (Gallopín 2006; Hinkel
2011). Here, the challenge lies in the accurate characterisation and quantification of
resilience, and thus the ability to provide a systematic basis for policy-making.
This article examines the methodological challenges of resilience as a new secu-
rity concept. It aims to initiate, but not predetermine, a discussion about the way in
which the concept can be operationalised, thus moving the agenda forward from
conceptual rhetoric to actual policy-making. Doing so, the article first seeks to stim-
ulate the process of evolving the trend concept of resilience into a tangible one.
Yet, it also considers how the operationalisation of this concept or approach has
 Journal of Risk Research 3

political implications itself, given that different operationalisation of the term allows
some material and social processes and challenges to be addressed or captured, but
not others (Hagmann 2012).
The article is structured as follows: In Section 2, we provide an overview of the
concept, its introduction into security affairs and its policy-making utility. Section 3
draws on examples of approaches to measure resilience, exploring three case studies
where resilience has been, or could be, measured. Section 4 illustrates the methodo-
logical considerations highlighted by these attempts to measure resilience, which
should be acknowledged in the event of a resilience index (RI) being developed.
The concluding Section 5 puts this discussion into perspective by acknowledging
the need and benefit of the methodological discussion and importance of the con-
cept in security, but noting that an exploration of resilience should not only be
reduced to a debate on operationalisation only. Effectively, we argue resilience to
include distinct implicit understandings of the role of the state, the economy and
civil society/citizens. Considering these, the idea of a resilient society not only
poses major methodological, but also major political challenges.
Downloaded by [ETH Zurich] at 02:22 25 June 2013

2. Resilience in policy and practice

Resilience has become a popular term in the international language of risk and
security in recent times, being called upon by the national security, civil protection
and emergency preparedness doctrines of countries, such as the USA, the UK, Aus-
tralia, Singapore, Canada or Germany (Bara and Brönnimann 2011). Resilience is
used widely to denote a general state of an entity – this community is ‘resilient’,
infrastructural redundancy increases ‘resilience’ to disruption, the organisation’s
‘resilience’ has fallen, etc. The Oxford English Dictionary gives two fundamental
meanings for resilience:

(1) the ability of a substance or object to spring back into shape; elasticity
(2) the capacity to recover quickly from difficulties; toughness.

These meanings become considerably more nuanced between the academic tra-
ditions and disciplines that have encouraged the use of the term in policy and prac-
tice. In the engineering and materials science fields, resilience is considered to be
‘the capacity of a material or system to return to equilibrium after displacement’
(Norris et al. 2008, 127). Likewise, in economic systems and businesses, resilience
is characterised by the continuation of operations during a crisis and the ability to
quickly ‘bounce back’ so that service provision is maintained (Herbane, Elliot, and
Swartz 2004), and this is mirrored by the traditional psychological view of individ-
ual resilience (Waller 2001; Werner and Smith 2001; Bonanno 2005). A somewhat
different take on resilience has been advanced by Norris and other scholars (2008),
who examine community resilience in the context of disasters. They considered
resilience a process linking resources like adaptive capacity, to outcomes like adap-
tation, readiness and response. This systemic interpretation of resilience matches
that articulated within studies of socio-ecological systems, where much of the more
recent and progressive scholarship has been made relating to the assessment and
characterisation of resilience and the implications for policy-making (Walker et al.
2009).
 4 T. Prior and J. Hagmann

The incorporation of resilience into national, international and even private

security doctrines has mirrored the popularity of the concept’s application in other
disciplinary areas (Walker and Cooper 2011). Since the 1990s, and particularly
since the 11 September terrorist attacks in New York, the unpredictability and
‘unknowability’ of threat has increased the difficulty of ensuring (and assuring)
societal security. The discourse concerning security has been pushed from terrorism
to terror, mainstreaming fear by highlighting that threat is ‘everywhere and
nowhere’ (Coaffee and Murakami Wood 2006, 514). Not only has this push driven
the ‘business’ of security, it has emphasised the need for human security and shifted
concerns about security from governments, to institutions and to society. The recog-
nised need to build societal resilience is also a response the real and systemic nature
of risks posed by many hazards (whether natural, social, economic, cultural or tech-
nological) that threaten society and its institutions and the affirmed connections
between these risks and security. This highlights the necessity to systematically
establish frameworks that both theoretically and empirically implant resilience
within security governance processes.
Downloaded by [ETH Zurich] at 02:22 25 June 2013

The discussion about whether to measure resilience is as old and as fraught as

the concept’s meaning. Many indices of resilience have been developed across a
range of disciplines. In general, these measures employ different definitions of resil-
ience, they are constructed using dissimilar constituents (indicators or variables),
they are utilised for different purposes – and as a result they ultimately measure dif-
ferent things. Even a basic exploration of what might constitute a measure (or
index) of resilience, for example, reveals the difficulty in establishing a measure
that is both accurate and ‘fit for purpose’ (Hinkel 2011). Measurement requires a
phenomenon to be observable and permit systematic attribution of value (like heat
or height), but the conceptual nature of resilience makes this difficult. Ultimately,
scientists are yet to agree on clear-cut conventions for measuring resilience and
there is, consequently, a significant literature discussing both how and whether the
phenomena can and should be measured (Hinkel 2011).
In a security context, understanding how resilient societies, infrastructure or
economies are to terror attack, for instance, and how this resilience manifests, has
significant implications for security policy and its development. It is also central to
the development of preparedness or mitigation strategies. Several reasons for mea-
suring resilience are articulated below:

2.1. To characterise resilience

Given its multidimensionality, developing a measure for resilience can be a step
towards characterising resilience in a particular context (such as security). Establish-
ing an adequate and usable measure of resilience first requires an articulation of of
the constituents of resilience, which is a valuable process in its own right – particu-
larly for the academic disciplines.

2.2. To raise awareness

Methods of communicating the need to be resilient to at-risk entities can be assisted
with an observable measure of resilience. Observations of resilience can help man-
agers to direct resilience-related information to entities whose resilience is lower
than some predetermined threshold.
 Journal of Risk Research 5

2.3. To allocate resources for resilience

Measurement allows the quantitative comparison of resilience (Figure 1) between
entities and this becomes useful in the allocation of risk management resources. A
RI could be used (in the same way as risk matrices: see Hagmann 2012) to argue
that allocation of funds is made in a transparent manner.

2.4. To build resilience

Identifying entities with low resilience is important for managing disruption and its
consequences. Knowing the extent of resilience can help risk management agencies
to best direct their assistance measures. The development of resilience ‘thresholds’
can permit better resilience-related policy decisions. Additionally, without measuring
resilience, it is difficult to gauge how resilience changes as a result of disruption or
following the implementation of resilience-building practices or processes.
Downloaded by [ETH Zurich] at 02:22 25 June 2013

2.5. To monitor policy performance

An index could be used to assess the effectiveness of resilience-building policy
through longitudinal comparisons of resilience in those entities targeted by the pol-
icy. To this end, the development of resilience-building policy should integrate the
identification of policy goals and targets against which efficacy, or ‘on-the-ground’
outcomes, might be assessed.

3. How is resilience measured in practice?

This section examines three resilience assessment techniques. In choosing cases, no
attempt is made to restrict the discussion to a narrow conception of security. We
describe three measurement approaches, their different applications and implica-
tions. The first two are examples of quantitive measures of resilience – the
Enhanced Critical Infrastructure Protection Program (US DHS) and the disaster
resilience of place (DROP) model. The last example illustrates a mixed-methods
approach to measurement that incorporates both qualitative and quantitative assess-
ment techniques. In each case, the development, benefits and limitations of the
approach are discussed (and summarised in Table 1 at the end of Section 3). These
examples were chosen to demonstrate that measuring resilience has been attempted
in very different ways and the technique or techniques used depend on the mea-
surer’s requirements and the characteristics of the system they are interested in.

Figure 1. Different resilience profiles and their ranking (‘best’, ‘acceptable’ and ‘worst’
situation). Profiles illustrate extent of affect on function, rapidity of function loss, and return
time to normal function. Image sourced from Department of Homeland Security: Science
and Technology Directorate (2010).
 Downloaded by [ETH Zurich] at 02:22 25 June 2013

6
T. Prior and J. Hagmann
Table 1. Summary and comparison of resilience measures.

Measure Type Operational? Benefits Limitations

ECIP Quantitative Yes • Characterises elements of harm and consequence before and • Time, money and labour intensive.
program after threat.
• Assesses entire spectrum of a risk for a critical • Analysis and interpretation is highly
infrastructure. demanding.
• Allows security planners to capture and analyse longitudinal • Only yields a very generic picture of CI
changes in relative vulnerability and resilience. security.
• Designed around a public-private partnership, with easy • Does not explore external influences on
information sharing. security.
• A relative measure.
DROP Quantitative No, • Connects vulnerability and resilience in a longitudinal • Not tested empirically, and community
model conceptual manner to show their dynamic nature. resilience indicators not articulated.
• Can account for new threat characteristics: multiple and • Up-scaling model would be difficult.
gradual onset events, place specificity, spatial/temporal
dynamics, peoples’ perceptions/attitudes.
• Incorporates antecedent measures of vulnerability and • No inclusion of ecological/environmental
resilience to account for exogenous factors, like policy. resilience.
Mixed Quant./ No, • Mixed methods can capture richness and complexity of • Qualitative analyses can be subjective.
methods qual. conceptual resilience.
• Lends itself to examining the contextual and heterogeneous • Mixed methodologies are time consuming.
nature of resilience in systems.
• Explores the way human elements interact with structural • More difficult to provide ‘clear-cut’
features of social systems. directions to policy makers that can be
ranked for decision-making.
 Journal of Risk Research 7

3.1. Case 1: the enhanced critical infrastructure protection program (Fisher

and Norman 2010; Petit et al. 2011)
Recent research sponsored by the US Department of Homeland security (DHS) has
focussed on the development of several quantitative indices designed to assist in the
risk management of critical infrastructures. The protective measures index (PMI),
RI and criticality index (CI) are intended to be used in an integrated fashion as part
of the DHS’s Enhanced Critical Infrastructure Protection Program. The initiative
draws on this triumvirate of indices to conduct assessments that identify infrastruc-
ture vulnerabilities and build resilience through partnerships with federal, state, local
and private sector stakeholders.

3.1.1. Application
Data is collected during critical infrastructure site visits where an extensive net-
work of Protective Security Advisers (PSAs – DHS employees) uses a survey
Downloaded by [ETH Zurich] at 02:22 25 June 2013

tool comprising more than 1500 data points (variables). This information is used
to characterise a facility in six key areas2: physical security (e.g. fences and
lighting), security management (e.g. emergency action plan and staff background
checks), the security force (e.g. training and patrols), level of information shar-
ing (e.g. threat sources), protective measures assessment (e.g. random security
measures) and dependencies (e.g. electricity and telecommunications). Data col-
lection by the PSA takes between four and eight hours and data is analysed to
calculate the PMI. The PMI is assessed as an aggregation of these ‘key areas’
and the subcomponents within these areas: each subcomponent (fence, threat
source, emergency action plan, etc.) is assigned a relative importance (weight)
by an expert panel based on its contribution to protection, and the PMI is a
weighted sum of the values measured in the six key areas. The resilience (RI)
and criticality (CI) indices are formulated in the same hierarchical way. For
resilience, the overarching components of robustness, recovery and resourceful-
ness are broken into specific subcomponents. Aggregating weighted values for
each set of subcomponents yields the particular index and the protective mea-
sures, resilience and criticality indices can be combined in a risk matrix to give
an overall view of a facility’s security.

3.1.2. Benefits
By integrating the aspects of vulnerability (the PMI is a proxy for vulnerability),
resilience and criticality, the ECIP initiative aims to characterise the elements of
harm and consequence before and after a threat, thereby assessing the entire spec-
trum of a given risk to some form of critical infrastructure or key resource. In addi-
tion, the regular execution of the three-component assessment (in 2010, 674 ECIP
surveys were conducted, with a larger number planned for 2012 (DHS National
Protection and Programs Directorate 2012) allows security planners to capture and
analyse longitudinal changes in security for single infrastructures and between simi-
lar infrastructures classed in the same infrastructure sectors (e.g. critical infrastruc-
ture used to deliver electricity or petroleum is classed in the energy infrastructure
sector). Both features are designed to gain a better understanding of the relationship
between the critical infrastructure and the risk environment.
 8 T. Prior and J. Hagmann

The ECIP programme is designed around a public–private partnership. A key

feature of the programme is the need to strengthen relationships and improve infor-
mation sharing between government departments and the organisations that own
and manage critical infrastructure. The programme gives owners of critical infra-
structure the ability to compare the security of their assets with other similar assets
that might be managed in a different way. ECIP provides the DHS with a mecha-
nism and guidelines to prioritise national protection efforts (allocation of resources
to increase resilience and decrease vulnerability). To facilitate communication of
results, information collated by the programme’s analysts is shared with facility
owners with an easy-to-use ‘dashboard’ or software interface that can represent data
in a readable and understandable way (Figure 2).3

3.1.3. Limitations
The extraordinary comprehensiveness of this set of measures generates its most sig-
nificant limitations, summarised below (based on information provided in Fisher
Downloaded by [ETH Zurich] at 02:22 25 June 2013

and Norman 2010; Petit et al. 2011):

• The methodology established is very comprehensive, but requires a huge net-

work of data collectors (PSAs) and analysts. Of course, this merely reflects
the diversity and number of critical infrastructures supporting today’s society.

Figure 2. A sample screenshot of the PMI dashboard used to communicate resilience

results to infrastructure facility managers. Image sourced from Argonne National Laboratory
(2012).
 Journal of Risk Research 9

In 2010, the DHS employed 93 PSAs at an annual cost of USD 12 million,

which will possibly increase by an extra 15 officers in 2012.
• Because the infrastructure survey collects information on over 1500 variables,
the analysis and interpretation process is highly demanding.
• The methodology is used across all critical infrastructure key resource sectors,
and therefore it can only provide a very generic picture of CI security, resil-
ience and vulnerability
• The analysis is fully systems-internal (looking only at the infrastructure) and
does not explore external factors that might influence security (like the infra-
structure’s geographic location near a large population, or position on a fault
line, for example).
• All in all, the ECIP programme is very expensive. In 2012, the DHS National
Protection and Programs Directorate (under which the ECIP programme falls)
has requested USD 27.5 million to run the programme.
• Finally, like most indices, the methodology yields only a relative measure of
resilience, i.e. a high measure of resilience from this analysis may not trans-
Downloaded by [ETH Zurich] at 02:22 25 June 2013

late to a facility that is impregnable and unwavering in reality. Nor does a

measure of low resilience mean that a similar facility’s operation will be com-
pletely shut down by the same event.

3.2. Case 2: the DROP model (Cutter et al. 2008; Cutter, Burton, and Emrich
2010)
The DROP model is a conceptual framework for community disaster resilience that
is yet to be fully operationalised. The model draws on and integrates resilience mea-
surement processes and practices from a range of literatures to establish a complex
resilience schema (Figure 3). It also identifies key candidate variables that might be
used to measure community disaster resilience.

3.2.1. Development
The authors contest that the DROP model fills a gap in the resilience measurement
literature, which they (rightly) point out is littered with inconsistent indicators and

Figure 3. The DROP model. Image sourced from Cutter et al. (2008).
 10 T. Prior and J. Hagmann

non-standard metrics. This is particularly the case in disaster literature, where resil-
ience models are often focused on engineered systems (like the ECIP model
described previously – robustness, recovery and resourcefulness), but largely fail to
capture social factors, inherent resilience and the antecedent processes that promote
or degrade resilience. Five dimensions (with underlying candidate variables) are
proposed as prospective resilience indicators, including: social resilience (e.g. aver-
age age and health coverage), economic resilience (e.g. employment and income
equality), institutional resilience (e.g. community mitigation actions and insurance),
infrastructure resilience (e.g. housing type and shelters) and community capital (e.g.
social networks and place attachment). There is a large literature discussing the dif-
ference between individual and community resilience, and the type of resilience
under examination must be made explicit in the development of a measurement
index (see e.g. Boon et al. 2012).
The model is articulated in a longitudinal process: (1) Inherent (antecedent, thus
existing before an event) levels of vulnerability and resilience combine, (2) with an
event, (3) event-specific coping abilities/capacities in the community and resulting
Downloaded by [ETH Zurich] at 02:22 25 June 2013

in (4) the disaster consequences (Figure 3). Recovery following the event is moder-
ated by absorptive capacity and adaptive resilience (5). Smit and Wandel (2006,
286) explore systemic ‘coping ability’ in some detail, reflecting on the relationship
between recovery and adaptive capacity.

3.2.2. Benefits
The model aims to explicitly connect vulnerability and resilience in a longitudi-
nal manner in order to capture the dynamic nature of these processes/concepts.
This then allows the model to better account for some of the challenges or frus-
trations that have plagued resilience and vulnerability measurement: multiple or
gradual onset events, place specificity and context/circumstance, spatial and tem-
poral dynamics of vulnerability and resilience and the perceptions or attitudes of
those people affected. The model’s longitudinal nature and the fact it incorpo-
rates antecedent measures of vulnerability and resilience also allows it to account
for the influence of exogenous factors like policy regulation. Well-executed pre-
paredness or mitigation policy could have a significant influence on antecedent
vulnerability or resilience, coping responses and processes that support adaptive
resilience post event and capturing this change is of great importance from a
policy perspective.

3.2.3. Limitations
No attempt (that could be found) has been made to operationalise this model as a
whole. A first step has been made to create composite indicators of community
resilience to assess baseline (inherent or antecedent) community resilience (Cutter
et al. 2010), however, and this provides a useful tool to assess and compare
underlying community resilience between geographic locations (in this case, here
the south-eastern states of the USA). The DROP model’s theoretical application is
powerful and meaningful, but practically very difficult to apply. Some general
limitations of the model include:
 Journal of Risk Research 11

• At present, the model does not articulate how the suggeted community resil-
ience indicators might be utilised in the conceptual model, which has to date
not been tested in an empirical case study (but see Cutter et al. 2010)
• The model is designed for application at the community scale, yet most disas-
ters occur at the meso- and macro-scales. Upscaling this model would likely
result in as resource intensive a process as decribed for the ECIP programme
above (Holling 2001).4
• An important feature in mitigation (institutional resilience) is government pol-
icy and proactivness towards mitigation, yet this aspect is not included in the
resilience indicator set.
• Likewise, there is no inclusion of ecological/environmental resilience either as
a resilience category or variable, despite clear documentation that social or
community resilience is closely connected to the environment and resources
derived from ecological processes. For example, there is a huge literature dis-
cussing ecosystem services and the well-being people derive from these natu-
ral resources, which connects closely with the resilience literature (see e.g.
Downloaded by [ETH Zurich] at 02:22 25 June 2013

Haines-Young and Potschin 2010).

3.3. Case 3: assessing resilience with mixed qualitative and quantitative

techniques
A mixed-methodology approach is one where both qualitative and quantitative tech-
niques are used to analyse a situation or process. It allows a triangulation of two
forms of information as a means of improving analytical rigour and deepening
understandings. The mixed methodology is particularly applicable to assessments of
community resilience, and the social interaction between communities and their
environments, and the structural aspects they rely on. Many authors argue that the
complexity of phenomena, like resilience and vulnerability, can best be captured in
a qualitative manner. The two models described subsequently are representative of
such an approach to measuring resilience.

3.3.1. Development
The mixed methodology resilience measure, or conceptual model is designed to
capture both the structural and experiential response of a community to a disaster,
disturbance or perturbation. One such model is ‘Bronfenbrenner’s bio-ecological
theory of development and resilience’, which is designed to relate individual and
community resilience within a hierarchy of systems (Boon et al. 2012). This model
is used to illustrate how individual resilience factors like adaptive coping and self-
efficacy are influenced by external processes like health provision or government
financial support and is somewhat analogous to Holling’s panarchy theory of resil-
ience, but with a strong social focus. Boon and colleagues provide an example of
how the proposed theory might be applied to measure community disaster resilience
(Boon et al. 2012) (Figure 4).
A second mixed-methods resilience model proposed by Flint and Luloff aims to
bridge the divide between theoretical debate and community experience with a
‘Mid-range model of community response to disaster’ (Flint and Luloff 2005). This
model aims to describe the interactional characteristics of a community and their
ability to act collectively to respond to problems facing the community. The model
 12 T. Prior and J. Hagmann
Downloaded by [ETH Zurich] at 02:22 25 June 2013

Figure 4. Bronfrenbrenner’s bio-ecological theory. Image sourced from Boon and

colleagues (2012).
is premised on the assumption that closely interacting communities are more capa-
ble of mobilising resources for response and recovery (Figure 5).

3.3.2. Benefits
Resilience is both contextual and heterogeneous, and the mixed-methods approach
to measurement aims to capture more of the complexity and richness of this con-
cept/process. By focusing on individuals, communities and places in analyses of

Figure 5. The mid-range model of community action in response to risk. Image sourced
from Flint and Luloff (2005).
 Journal of Risk Research 13

resilience, this approach is designed to explore the human elements of exposure to

threat and the way these elements interact with the structural features of social sys-
tems. A purely quantitative approach to measuring community resilience (that uses
income, age, education and access to information, for example, as the measures of
resilience) will be unable to explore and describe the social determinants of, and
influence on, community and individual resilience features like experience, well-
being, adaptive capacity, sense of community and collective efficacy – all of which
are highlighted as important resilience features.

3.3.3. Limitations
While mixed methodologies seem to lend themselves particularly well to measuring
the interaction between social and structural features in a community or social sys-
tem, their application in practice is limited. To some extent, this is due to the cur-
rent state of development in mixed methods RI-making.
Downloaded by [ETH Zurich] at 02:22 25 June 2013

• Qualitative approaches to analysis are sometimes difficult to compare with

each other, given that different standards are used and that coding rules may
evolve or shift.
• Applying mixed methodologies in data collection and analyses is time
consuming, though not hugely resource intensive.
• Ultimately, authorities prefer assessments that provide ‘clear-cut’ guidance for
policy-making, which is often typically reduced to quantitative inquiry.
According to such conceptions of science, a qualitative approach to measuring
resilience, even if potentially more contextualised and detailed than other
approaches, might not be accepted to yield a sufficiently valid ranking of
resilience profiles.

4. General considerations for Resilience Index-making

As noted above, there are many reasons to measure resilience. Yet, the diverse, and
often contradictory, literature highlights both the difficulty and dilemmas associated
with measuring this trait. For this reason, it is very important to think critically
about indices, considering what problems measurement might raise or encounter
and where caution should be taken in the development of a RI. Several of the main
issues for an index are outlined below in no order of importance.

4.1. Simplifying complexity

Einstein once noted that theory should be ‘as simple as possible, but not simpler’.
This pointer is highly valuable for the resilience measurer, but also illustrates a
conundrum – how can you simplify a complex process or concept so that it is under-
standable and measureable without losing the very complexity and deepness of
meaning you are trying to capture? Resilience is inherently complex and with
increasing complexity comes greater difficulties in establishing measures and inter-
preting results. Making a measure that satisfies the necessity of interpretation and
captures the complexity of resilience as a policy-relevant phenomena will be time-
consuming and expensive. The issue of simplification leads into the following point.
 14 T. Prior and J. Hagmann

4.2. Absolute vs. relative evaluation

An index is a way of simplifying the complex as duly noted. Indeed, many vari-
ables used to indicate phenomena are proximal representations of the actual subject
of measurement and only assumed to be representative.5 For this reason, most indi-
ces only yield a relative measure (Mori and Christodoulou 2012), rather than an
absolute measure. As long as the measure is calculated consistently between enti-
ties, all a relative measure allows is comparisons between places, between entities,
or over time. This is good only if you want a relational understanding of resilience
(for example, to allocate resilience development funding): it will not tell you if the
river-side community you are concerned about can be resilient to a major flooding
event, only whether it will be better off than the neighbouring community. Develop-
ing resilience/vulnerability benchmarks for at-risk entities is a step towards avoiding
this problem (Cutter et al. 2010; Mori and Christodoulou 2012).

4.3. Indicator arbitrariness and weightings

Downloaded by [ETH Zurich] at 02:22 25 June 2013

The fact that there is considerable discussion about the exact meanings of resilience
in the risk-related literature results in arbitrary or subjective application of indica-
tors. A measure for resilience will likely be an amalgam of several indicators, but it
is almost impossible to identify and choose indicators for resilience that can be
measured on the same scales. In order to integrate these indicators into a coherent
index, transformation of the indicators to the same scale is necessary. Additionally,
different indicators are assumed to contribute differently to resilience and are
assigned weightings to differentiate these contributions. Assigning weightings is a
time-consuming and difficult process, especially without an in-depth knowledge of
the way particular behaviours, structures, policies, etc. contribute to the resilience of
the entity under examination.6

4.4. Data quality, availability and suitability

A valid measure of resilience could be limited by data quality, availability or suit-
ability. In most cases, the indicators (and therefore the data) used to measure resil-
ience would not have been established to explicitly do so – as noted at point b)
above, indicators are often proxies. As such, data collected for these indicators may
be easily available, but not specifically applicable to the situation or measure they
are intended to be used in. Also, where data is easily available, there lies the temp-
tation to use it, whether or not it encapsulates, or is suitable for measuring the phe-
nomenon under question. Using proxies highlights the possibility that these, and the
data that represent them, may ultimately define resilience, rather than the needs of
the measurer.

4.5. Context, place and threat specificity

Almost all authors recognise that resilience and vulnerability vary dramatically
between places and with respect to the events they are examined in relation to. This
means that it is very difficult (if not impossible) to construct a generic and direct
measure of resilience. This difficulty is evidenced by the huge variety of proposed
measures and theoretical conceptions of resilience presented in the academic
literature.
 Journal of Risk Research 15

4.6. Fit for purpose

Given its vagueness, clear framing of resilience is likely the most important factor
in developing a measure of resilience. From a policy perspective, three consider-
ations can yield clarity:

(1) a sound definition;

(2) explicit policy linked to the definition; and
(3) explicit articulation of scale and context.

Creating a useable and meaningful measure of resilience requires a sound articu-

lation of the definition with respect to the reason the measurement is required (e.g.
disaster mitigation policy). Connecting the definition to the policy will assist the
development of a measure of resilience that is fit for the purpose it is intended. A
measure can only be as good as the articulation of the problem (or policy question)
to be explored and relies on the early and explicit determination of policy goals and
Downloaded by [ETH Zurich] at 02:22 25 June 2013

targets.

5. Conclusion
Resilience is a popular new security concept that is being introduced into increas-
ingly numerous national and international, but also private security strategies. Mea-
suring resilience will become more important in operationalising this new security
agenda and for turning this concept into an actual policy tool. The three models
presented in this article outline various approaches to measure resilience. Having
shown the origins, utilities and limitations of the three approaches, the chapter has
distilled some general lessons from existing experiences with resilience modelling
that could also inform the construction of future RI (or indices), projects that are
envisaged in a number of European and North American countries.
These methodological lessons notwithstanding, it remains important to bear in
mind that a ‘resilience policy’ must not be reduced to issues of measurement alone.
On the one hand, this means that analysts and policy-makers alike must reflect
carefully on the ways in which the choice of measures themselves affects the way
in which resilience can and is being understood – and that such conceptual defini-
tion via particular methodologies may not necessarily be accepted by different
social groups. On the other hand, it also means that like any policy concept, resil-
ience policy-making must also consider the distinct political configuration of actors
that might influence the policy’s development and implementation. This is to say
that a ‘resilience policy’ should not only reflect agreed-on measures and objects,
but that its development and implementation must also acknowledge the roles and
responsibilities of the state, the economy, civil society and even individuals. Cur-
rently, initial evidence from risk and resilience research suggests that the new secu-
rity agenda indeed tends to affect established divisions of labour in the societal
balance. In its implementation, private companies, for instance, are seen to become
more closely integrated into the provision of national security – for example, in the
case of critical infrastructures that provide services to the wider society, many of
which are owned and operated by private companies. By the same token, state
authority is transformed to assume the role of a ‘task supervisor’, or authority in
charge of defining, rather than implementing, the meta-governance of the security
sector (Lund Petersen 2012a).
 16 T. Prior and J. Hagmann

Beyond discussions of measurement, such an implicit reconfiguration of security

policy-making either ‘by method’ or ‘by concept’ is, thus, highly political and nec-
essarily requires transparency. Seen in this perspective, the endeavour of creating an
increasingly resilient society not only refers to a goal that should be attained.
Instead, the resilience agenda effectively also refers to a larger political process,
namely societal decisions about whose and what kind of resilience is addressed,
and political conclusions about who assumes what task and what responsibility in
the production of a secure and resilient society. With this, the operationalisation of
resilience not only represents a major methodological challenge today, but also
highlights a political challenge that should not be underestimated as this trend
concept is increasingly called into practice.

Notes
1. Strunz argues that the conceptual vagueness of resilience could encourage creativity, per-
mit empirical testability and foster transdiciplinarity for pragmatic problem-solving of
Downloaded by [ETH Zurich] at 02:22 25 June 2013

complex social issues.

2. For a full list of the components within this survey refer to Petit and colleagues (2011).
3. The top image gives an overall view of the resilience of the particular facility, with dif-
ferent coloured dots indicating the facility’s relative resilience when compared to other
similar facilities. The lower image breaks down physical security into its components,
indicating again how the facility (in this case an office building) compares to 100 other
similar buildings.
4. C.S. Holling (widely considered the progenitor of modern resilience thinking) postulated
that systemic resilience was an interaction between and within multiple spatial and tem-
poral scales. Without this interaction, the change in resilience could not be adequately
described.
5. Though in the measurement of psychological resilience a central necessity is an estima-
tion of how reliably a measurement scale can represent the construct it is designed to
measure. Windle and colleagues (2011) provide a good overview of the measurement of
psychological resilience.
6. The Department of Homeland Security and Argonne National Laboratories engaged an
expert panel to determine the weightings for each indicator used in the PMI described at
Section 3.1.

References
Adger, W. N. 2006. “Vulnerability.” Global Environmental Change 16 (3): 268–281.
doi:10.1016/j.gloenvcha.2006.02.006.
Aradau, C., and R. Van Munster. 2007. “Governing Terrorism through Risk: Taking Precau-
tions, (Un)Knowing the Future.” European Journal of International Relations 13 (1):
89–115. doi:10.1177/1354066107074290.
Argonne National Laboratory. 2012. Decision and Information Sciences: Better Infrastruc-
ture Risk and Resilience. Argonne National Laboratory: US Department of Energy,
Office of Science 2012 [cited April 7 2012]. http://www.dis.anl.gov/projects/ri.html.
Bara, C., and G. Brönnimann. 2011. Resilience – Trends in Policy and Research. Zurich,
Switzerland: Center for Security Studies (CSS), ETH Zürich.
Bonanno, G. A. 2005. “Resilience in the Face of Potential Trauma.” Current Directions in
Psychological Science 14 (3): 135–138.
Boon, H. J., A. Cottrell, D. King, R. B. Stevenson, and J. Millar. 2012. “Bronfenbrenner’s
Bioecological Theory for Modelling Community Resilience to Natural Disasters.” Natu-
ral Hazards 60 (2): 381–408.
Clarke, S. E., and E. Chenoweth. 2006. “The Politics of Vulnerability: Constructing Local
Performance Regimes for Homeland Security.” Review of Policy Research 23 (1):
95–114.
 Journal of Risk Research 17

Coaffee, J., and D. Murakami Wood. 2006. “Security is Coming Home: Rethinking Scale
and Constructing Resilience in the Global Urban Response to Terrorist Risk.” Interna-
tional Relations 20 (4): 503–517. doi:10.1177/0047117806069416.
Cutter, S. L., L. Barnes, M. Berry, C. Burton, E. Evans, E. Tate, and J. Webb. 2008. “A
Place-based Model for Understanding Community Resilience to Natural Disasters.” Glo-
bal Environmental Change 18 (4): 598–606.
Cutter, S. L., C. G. Burton, and C. T. Emrich. 2010. “Disaster Resilience Indicators for
Benchmarking Baseline Conditions.” Journal of Homeland Security and Emergency
Management 7 (1). doi:10.2202/1547-7355.1732.
Daase, C., and O. Kessler. 2007. “Knowns and Unknowns in the ‘War on Terror’: Uncer-
tainty and the Political Construction of Danger.” Security Dialogue 38 (4): 411–434.
Department of Homeland Security: Science and Technology Directorate. 2010. Risk and
Resilience: Exploring the Relationship. Washington: National Infrastructure Advisory
Council: Homeland Security Studies & Analysis Institute.
DHS National Protection and Programs Directorate. 2012. NPPD Fiscal Year 2012 Budget
Request – Cyber Security and Infrastructure Protection Programs. Department of Home-
land Security 2011 [cited April 7 2012]. http://www.dhs.gov/ynews/testimony/testi-
mony_1301595025263.shtm.
Fisher, R. E., and M. Norman. 2010. “Developing Measurement Indices to Enhance Protec-
Downloaded by [ETH Zurich] at 02:22 25 June 2013

tion and Resilience of Critical Infrastructure and Key Resources.” Journal of Business
Continuity & Emergency Planning 4 (3): 191–206.
Flint, C. G., and E. Luloff. 2005. “Natural Resource-based Communities, Risk, and Disaster:
An Intersection of Theories.” Society and Natural Resources 18 (5): 399–412.
Florin, M.-V. 2011. “Risk Governance by Governments.” In Commonwealth Finance Minis-
ters Reference Report 2011, edited by Commonwealth Secretariat, 2–4. London: Publica-
tions Section, Commonwealth Secretariat.
Folke, C. 2006. “Resilience: The Emergence of a Perspective for Social-Ecological Systems
Analyses.” Global Environmental Change 16 (3): 253–267.
Gallopín, G. C. 2006. “Linkages between Vulnerability, Resilience, and Adaptive Capacity.”
Global Environmental Change 16 (3): 293–303.
Hagmann, J. 2012. “Fukushima: Probing the Analytical and Epistemological Limits of Risk
Analysis.” Journal of Risk Research 15 (7): 801–815.
Hagmann, J., M. Dunn-Cavelty, and A. Wenger. 2011. “Versorgungspolitik und Vers-
orgungssicherheit im Kontext globaler Risiken und Interessen.” [National Economic Sup-
ply and Supply Security in a Context of Global Risks and Interests.] In Bulletin zur
Schweizer Sicherheitspolitik, edited by A. Wenger and D. Trachsler, 9–35. Zürich: Center
for Security Studies, ETH Zürich.
Haimes, Y. Y. 2009. “On the Definition of Resilience in Systems.” Risk Analysis 29 (4):
498–501.
Haines-Young, R., and M. Potschin. 2010. “The Links Between Biodiversity, Ecosystem Ser-
vices and Human Well-Being.” In Ecosystem Ecology: A New Synthesis, edited by C.
Frid and D. Raffaell, 110–139. Cambridge: Cambridge University Press.
Herbane, B., D. Elliot, and E. M. Swartz. 2004. “Business Continuity Management: Time
for a Strategic Role?” Long Range Planning 37 (5): 435–457.
Hinkel, J. 2011. “‘Indicators of Vulnerability and Adaptive Capacity’: Towards a Clarifica-
tion of the Science-Policy Interface.” Global Environmental Change 21 (1): 198–208.
Holling, C. S. 2001. “Understanding the Complexity of Economic, Ecological, and Social
Systems.” Ecosystems 4 (5): 390–405.
Lund Petersen, K. L. 2012a. Corporate Risk and National Security Redefined, Routledge
Advances in International Relations and Global Politics. New York: Routledge.
Lund Petersen, K. L. 2012b. “Risk Analysis: A Field Within Security Studies?” European
Journal of International Relations 18 (4): 693–717.
Mori, K., and A. Christodoulou. 2012. “Review of Sustainability Indices and Indicators:
Towards a New City Sustainability Index (CSI).” Environmental Impact Assessment
Review 32 (1): 94–106.
Norris, F. H., S. P. Stevens, B. Pfefferbaum, K. F. Wyche, and R. L. Pfefferbaum. 2008.
“Community Resilience as a Metaphor, Theory, Set of Capacities, and Strategy for
Disaster Readiness.” American Journal of Community Psychology 41 (1–2): 127–150.
 18 T. Prior and J. Hagmann

Petit, F., W. Buehring, R. Whitfield, R. Fisher, and M. Collins. 2011. “Protective Measures
and Vulnerability Indices for the Enhanced Critical Infrastructure Protection Pro-
gramme.” International Journal of Critical Infrastructures 7 (3): 200–219.
Smit, B., and J. Wandel. 2006. “Adaptation, Adaptive Capacity and Vulnerability.” Global
Environmental Change 16 (3): 282–292.
Strunz, S. 2012. “Is Conceptual Vagueness an Asset? Arguments from Philosophy of Science
Applied to the Concept of Resilience.” Ecological Economics. 76: 112–118,
doi:10.1016/j.ecolecon.2012.02.012.
Walker, B. H., N. Abel, J. M. Anderies, and P. Ryan. 2009. “Resilience, Adaptability, and
Transformability in the Goulburn-Broken Catchment, Australia.” Ecology and Society 14
(1): 12.
Walker, J., and M. Cooper. 2011. “Genealogies of Resilience: From Systems Ecology to the
Political Economy of Crisis Adaptation.” Security Dialogue 42 (2): 143–160.
Walker, B., C. S. Holling, S. R. Carpenter, and A. Kinzig. 2004. “Resilience, Adaptability
and Transformability in Social–Ecological Systems.” Ecology and Society 9 (2): 5.
Waller, M. A. 2001. “Resilience in Ecosystemic Context: Evolution of the Concept.” Ameri-
can Journal of Orthopsychiatry 71 (3): 290–297.
Werner, E. E., and R. S. Smith. 2001. Journeys from Childhood to Midlife: Risk, Resilience,
and Recovery. Ithaca, NY: Cornell Univ Press.
Downloaded by [ETH Zurich] at 02:22 25 June 2013

Windle, G., K. M. Bennett, and J. Noyes. 2011. “A Methodological Review of Resilience

Measurement Scales.” Health and Quality of Life Outcomes 9 (8): 1–18.
Zhou, H., J. Wang, J. Wan, and H. Jia. 2010. “Resilience to Natural Hazards: A Geographic
Perspective.” Natural Hazards 53 (1): 21–41.

View publication stats