Creating an index of social vulnerability
to climate change for Africa
Katharine Vincent
August 2004
Tyndall Centre for Climate Change Research
Working Paper 56
Creating an index of social
vulnerability to climate
change for Africa
Katharine Vincent
Tyndall Centre for Climate Change Research
and
School of Environmental Sciences
University of East Anglia
Norwich NR4 7TJ
Email: katharine.vincent@uea.ac.uk
Tyndall Centre Working Paper No. 56
August 2004
Please note that Tyndall working papers are "work in progress". Whilst they are commented on by
Tyndall researchers, they have not been subject to a full peer review. The accuracy of this work and
the conclusions reached are the responsibility of the author(s) alone and not the Tyndall Centre
Summary
Although shrouded by uncertainties as to its specific nature and manifestations, climate change is a
very real phenomenon that will inevitably affect human populations in the coming decades. Thus far
impacts assessments have been predicated upon a linear model of pressure-impact, which focus on
the biophysical vulnerability of the natural environment to the risk exposure. It has been
increasingly recognised that those impacts are mediated by the social vulnerability, that is the
complex interrelationship of social, economic, political, technological and institutional factors that
renders an exposure unit vulnerable or resilient in the face of a hazard.
The aim of this research was to create an index to empirically assess relative levels of social
vulnerability to climate change-induced variations in water availability and allow cross-country
comparison in Africa. A theory-driven aggregate index of social vulnerability was formed through
the weighted average of five composite sub-indices: economic well-being and stability (20%),
demographic structure (20%), institutional stability and strength of public infrastructure (40%),
global interconnectivity (10%) and dependence on natural resources (10%).
Using current data, Niger, Sierra Leone, Burundi, Madagascar, and Burkina Faso are the most
vulnerable countries in Africa. When mapped in conjunction with the appropriate indicators of
biophysical vulnerability this will allow a more holistic and integrated assessment of the impacts of
climate change-induced changes in water availability. This contributes to the academic debate by
proposing a methodology to overcome the persistent dichotomy in different epistemological
approaches to vulnerability research. It also has policy implications by highlighting priority areas
for aid and building adaptive capacity, as enshrined in the UNFCCC.
1
Introduction
Growing interest in global environmental change has focused attention on the inter-relationships
between natural and human systems. Climate change is increasingly accepted as a major issue
facing human societies in the 21st century. The IPCC has concluded that anthropogenic greenhouse
gas emissions will continue to drive change into the future unless dramatic mitigation measures are
adopted (Houghton et al, 2001). Assessments of future climate change have highlighted the
potential regional differentiation of impacts. Traditionally science has concentrated on projections
of climate change using models based on past analogues of climate variability, and then made some
suggestions as to how such changes might impact on human populations through changing patterns
of weather and coastal flooding. However a limitation of such top down approaches is their failure
to take account of the differential vulnerabilities of human populations to those environmental risks.
Assessing the likely impacts of climate change is thus inextricably linked with an assessment of the
social vulnerability. Understanding how different societies will respond to and adapt to these
changes is thus a key element of research and policy relating to global environmental change.
As a result the field of vulnerability science has emerged comprising more bottom up studies of the
way in which human populations mediate environmental change to produce impacts. This field of
enquiry marks one of a number of emerging research areas of society-nature relations, and is
classified amongst the emerging field of sustainability science, with key policy and other practical
applications. However its development is currently impeded by the variety of paradigms and
conceptual approaches, fragmented empirical studies and therefore a lack of comparability on the
larger scale. The aim of this research is therefore to fill an academic and policy demand for the first
assessment of national level social vulnerability to climate change in Africa. By developing an
index, this puts social vulnerability in a language and format that can be added to the existing
biophysical vulnerability assessments to create holistic and integrated studies of the potential
impacts of climate change.
The paper begins with a brief history of the evolution of vulnerability science, highlighting the two
main epistemological approaches and their effect on the current state of the art, as well as outlining
the conceptual framework for this research. Section 3 introduces the role of indicators and indices in
linking science and policy, drawing attention to their strengths and weaknesses, reviewing existing
attempts at national level vulnerability indices, and defining good practice. Section 4 highlights the
methodological choices involved at the various stages of index production, justifying the methods to
maintain transparency. Section 5 summarises the results of the study, both in terms of the aggregate
scores and ranks of social vulnerability, and the nature of the component sub-indices. Section 6
concludes by elaborating on the intellectual contribution of the research to the field of vulnerability
science, highlighting the policy applications in terms of holistic and integrated impacts assessments,
as well as outlining some directions for further research.
2
The emergence of vulnerability science
2.1
What is vulnerability?
Vulnerability is a contested term which has its origins in the natural hazards and food security
literature, and is increasingly being applied in climate change impacts assessments. A variety of
definitions have been proposed (for a review, see Cutter, 1996a). It is generally taken to be the
ability to anticipate, resist, cope with and respond to a hazard (Blaikie et al, 1994). However, a
meta-analysis of vulnerability definitions reveals a distinction in the literature between the two main
epistemological approaches. The natural hazards school of thought arises out of a positivist vein,
and therefore focuses on the objective studying of hazards. Under this approach emphasis is placed
on a particular environmental stress, and vulnerability refers to the risk of exposure of an ecosystem
to a hazard. In contrast, the human ecology and political economy schools of thought have arisen
out of interpretive social science paradigms based on relativist and constructivist ontologies. In
these cases vulnerability refers to a particular group or social unit of exposure and especially to the
structures and institutions – economic, political and social – that govern human lives.
2.2
What is vulnerability to climate change?
Despite the IPCC’s conclusion that anthropogenic climate change is a real phenomenon, there is a
large amount of uncertainty relating to the nature of these changes. Projections of change are
dependent on global climate models that simulate elements of the climate system and can be forced
according to particular plausible scenarios of emissions (SRES) (e.g. Arnell et al, 2004; Parry et al,
2001; Hulme et al, 1999). In addition to changing distributions of temperature and rainfall, other
potential impacts include changes in the patterns, nature and intensity of climate-related natural
hazards, such as hurricanes and droughts. Whilst uncertainty is an important consideration, the
incremental nature of climate change also differentiates it from natural hazards, most of which are
discrete events after which populations have a chance to recover and reduce their vulnerability
levels.
However, even if the exposure to climate change is similar there will be variation in the impacts due
to the differential vulnerability of ecosystems to such changes. Investigating the potential effects of
changing climate has occurred for different ecosystems and sectors in various locations on a case
study basis, for example coasts (Klein and Nicholls, 1999; Capobianco et al, 1999; Olivio, 1997),
rivers/water resources (Hurd et al, 1999; Mendoza et al, 1997), forests (Dixon et al, 1996), wetlands
(Hartig et al, 1997), agricultural productivity (Chipanshi et al, 2003; Lal et al, 1997; Magrin et al,
1997; Karim et al, 1996), and the carbon sink (Levy et al, 2004). Essentially such studies are
predicated upon a simple linear relationship between hazard and impact, and vulnerability is
referring to the sensitivity of natural environments to projected changes in climate, or their
biophysical vulnerability.
Through influencing resource availability, such impacts might in turn filter through to impact human
populations, particularly those that are geographically proximate to the exposure. Other studies have
explicitly investigated the impacts of climate change on human lives through such parameters as
malaria incidence (van Liesert et al, 2004; Martens et al, 1999), food security (Parry et al, 2004;
Parry et al, 1999), water availability (Arnell, 2004; Arnell, 1999) and coastal flooding (Nicholls,
2004; Nicholls et al, 1999). This trend of assessing impacts based on biophysical vulnerability is
also enshrined in the IPCC process (McCarthy et al, 2001). However, the approach has attracted
criticism through assuming humans are passive recipients of global environmental change, and thus
failing to capture their dynamic ability to mediate such hazards, either through resisting an event or
coping once it occurs (Jones and Boer, 2003, Stonich, 2000). Essentially, as once predominated with
regard to natural hazards, climate change is seen as a problem for society, not of society (Hewitt,
1997). Many impacts assessments have thus been impeded by only considering one side of the
equation (Cutter, 1996b).
Researchers in the field of vulnerability to climate change have also started to embrace more
interactionist approaches to society-nature relations. Political ecology is a synthetic approach that
brings together critical insights from political-economy perspectives with the awareness of physicalhuman systems interaction and place specificity that are the focus of the human ecology school
(Burton et al, 1993). Instead of focusing solely on the risk of exposure to physical phenomena, this
approach recognises that such physical phenomena are embedded in and mediated by the particular
human context (social, political, economic, institutional) in which they occur. Whilst physical
phenomena are necessary for the production of a natural hazard, their translation into risk and
potential for disaster is therefore contingent upon human exposure and a lack of capacity to cope
with the negative impacts that exposure might bring to individuals or human systems (Pelling, 2000).
This broader approach has thus highlighted the importance of assessing the complex reality of
vulnerability when predicting future impacts of environmental change as the most vulnerable people
may not be in the most vulnerable places: poor people can live in resilient biophysical environments
and be vulnerable, and wealthy people can be in fragile physical environments and live relatively
well (Liverman, 1994). Understanding the impacts of climate change is thus inextricably linked with
the human conditions that create a resilience or vulnerability to that event (Parry and Carter, 1998).
This recognition has consequences for vulnerability and impacts assessments, and there has been a
growth in theoretical and conceptual studies aimed at highlighting the nature of vulnerability to
climate change. Social vulnerability, in contrast to being seen as an outcome, is viewed more as a
potential state of human societies that can affect the way they experience natural hazards (Adger,
1999; Adger and Kelly, 1999; Blaikie et al, 1994). This potential state is in constant flux, reflecting
its dependence on the dynamic interaction of a range of economic and social processes which
influence the capacity of individuals, social groups, sectors, regions and ecosystems to response to
various socio-economic and biophysical shocks (Leichenko and O’Brien, 2002; Clark et al, 2000;
Comfort et al, 1999). The most vulnerable are considered those who are most exposed to
perturbations, who possess a limited coping capacity and who are least resilient to recovery (Bohle et
al, 1994). Other definitions of vulnerability focus on concepts of marginality, susceptibility,
adaptability, fragility and risk (Liverman, 1994).
Given evidence of differential social vulnerability in the face of hazards or broader environmental
risk exposure, a number of studies have tried to characterise the determinants that may give rise to
vulnerability, or its reciprocal state of resilience (e.g. Pelling, 2003; Smith, 2001; Blaikie et al,
1994). Vulnerability is therefore a function of economic, social, political, environmental and
technological assets. Who, where, and when vulnerability and disaster strike is determined by the
human and physical forces that shape the allocation of these assets in society (Pelling and Uitto,
2001). This is dependent on the scale of enquiry. On the large-scale macro processes will be most
important in determining the distribution and production of entitlements. In the well-developed food
security literature famines have been explained on the basis of entitlement theory (Sen, 1981), where
the distribution and reproduction of entitlements is dependent on the structural factors of political
economy that precipitate entitlement failure (Downing, 1996; Bohle et al, 1994; Watts and Bohle,
1993). In the face of exposure to climate change, some populations will be able to draw on their
entitlements to adapt to the risk, for example through awareness and preparation, insurance for
losses, and diversifying livelihoods. Adger (1999) shows how collective vulnerability (at
community level or higher) to extremes in coastal Vietnam is determined by institutional and market
structures. In contrast, on the local scale the role of human agency has a greater influence in access
to resources and household-level social status. In such cases entitlements are socially and spatially
differentiated according to such factors as gender, ethnicity, religion, class and age (Denton, 2002;
Enarson, 2000; Wisner, 1998; Cutter, 1995). The fact that vulnerability is embedded in wider
processes also creates the opportunity for reduction or increase through the social amplification of
risk (Kasperson et al, 1995).
Developing countries are particularly vulnerable to climate change impacts because of exposure and
sensitivity to climate change and because some elements of the capacity to adapt may be limited:
hence biophysical and social vulnerability (McCarthy et al, 2001). Africa is thought to be
particularly at risk due to having a higher share of economies in climate-sensitive environments than
other continents and a heavy reliance on the natural resource base (UNEP, 2001; Smith and Lenhert,
1996). Recurrent droughts and over dependence on rain fed agriculture mean that livelihoods are
closely related to resource availability, itself sensitive to climate change (Desanker and Magadza,
2001). A key threat of climate change to Africa is thus the projected changes in water availability
(Sokona and Denton, 2001; Parry et al, 2001). However, such generalisations disguise heterogeneity
that exists at the sub-continental scale due to variation in vulnerability (O’Brien and Leichenko,
2000; Downing et al, 1997). Some populations have shown resilience in the face of the climate
variability that characterises the vast swathes of semi-arid drylands (e.g. Hulme, 2001; Mortimore,
1998). Country-level analyses of vulnerability are therefore required.
2.3
Synthesising vulnerability research
Whilst vulnerability is an ever-emerging area of academic enquiry, the field is currently fragmented
and defined by competing paradigms, conflicting theory and terminology, incomparability of
empirical studies and a lack of comparative analysis and findings (Mitchell, 2001; Clark et al, 2000 :
3). Researchers originating out of the top-down positivist school of thought, who emphasise
biophysical vulnerability, have tended to focus on modelling of global impacts and place-based case
study approaches, often incomparable due to the lack of accepted methodology or conceptual
framework for study. Those who prefer the more bottom up constructivist approaches emphasise
social vulnerability and its role in mediating hazard exposure and determining whether or not it
results in an impact, and have tended to focus on developing theoretical insights into the processes
and interactions with emphasis on local level case studies. Moreover the divergent terminologies,
epistemologies and methodologies involved in vulnerability studies are thus reinforcing the
polarisation of top down biophysical risk exposure on the one hand and bottom up social
vulnerability on the other, to the overall detriment of holistic climate change impacts assessments.
With climate change it is vital to know not only the consequences for ecosystems (biophysical
vulnerability), but also if and how the social exposure unit will be able to respond to changing
exposures and the effects on their coping capacities (social vulnerability). There is therefore a need
to try and bridge the gap between the two diverging schools of thought. One possible way of doing
this is to apply the methodologies of the better-developed biophysical vulnerability school to the
social vulnerability school. This way more recent insights into social vulnerability can be
incorporated in vulnerability assessments that tend to currently favour the former approach (e.g.
IPCC). Upcoming research agendas for vulnerability science have called for the development of
comparative indicators of vulnerability in order to draw together emerging themes and enable more
systematic assessment of the nature of vulnerability (Cutter, 2003, 2001; UNEP, 2001; Clark et al,
2000).
Whilst there is no superior scale of analysis of vulnerability, country-level analyses have hitherto
been largely overlooked in favour of ecosystem-scale studies of biophysical vulnerability, and a
limited number of small-scale studies of social vulnerability. Due to the scale specificity it is not
methodologically possible to simply add small-scale studies, or apply theoretical frameworks across
scales of analysis (Clark et al, 2000; Adger and Kelly, 1999). As the state is the primary unit of
decision-making, it makes sense to conduct vulnerability assessments at this level. A lack of such
national level empirical assessments is currently impeding the effective allocation of financial aid
and technical assistance for climate change adaptation. The aim of this research is therefore to
develop a national-level aggregate index of social vulnerability (hereafter SVI) to one parameter of
climate change that is thought to bring particular risk to Africa, that of changing water availability.
This involves identifying the processes that give rise to national level vulnerability, finding
appropriate indicators of these variables, and selecting a means of aggregation. Encapsulating the
multiple dimensions of vulnerability in this manner will give each country an aggregate score.
These empirical results will allow cross-country comparison of vulnerability to climate changeinduced changes in water availability in Africa.
2.4
The conceptual framework for assessing vulnerability in this research
Given the variety of approaches to vulnerability it is worth being explicit about the conceptual
framework and terminology used. In this case an impact is a function of hazard exposure and both
the biophysical and social vulnerability, where biophysical vulnerability is the sensitivity of the
natural environment to the exposure and social vulnerability is the sensitivity of the human
environment to the exposure. Therefore exposure to a hazard is a necessary prerequisite for an
impact. Whether that exposure translates into a hazard depends on the nature of the vulnerability: if
the natural environment is particularly sensitive and the human population is of low economic status
with poor preparedness and few social institutions to facilitate coping then the impact will be high.
If the social vulnerability is lower due to a more appropriate coping capacity, then exposure of the
same nature may result in a lesser or even no impact.
This approach relies on the notion of each exposure unit having a coping capacity, or range of
exposure which can be coped with (figure 1). The nature of this range is dependent on the
determinants of vulnerability that render the unit vulnerable or resilient in the face of such exposure.
Africa is thought to already be near the edges of its coping ability (Sokona and Denton, 2001). The
timescale is an important consideration here, particularly in light of the projected incremental nature
of climate change. If the shock is temporally limited as in many natural hazards, the coping range
may diminish immediately following exposure as resources are diverted into coping mechanisms,
thus if there is further hazard exposure within a short time period the baseline vulnerability might be
higher than otherwise, making a population vulnerable to recurrent events. However, social learning
tends to rely on reactive adaptation, and thus if climate change is incremental it is possible that
changing exposure may promote adaptation amongst the exposure unit (Adger and Kelly, 1999).
Many African societies are already adapted to the climate variability to which they are exposed
(Mortimore, 1998), and this variability is a good proxy for risks associated with future climate
change, provided the rate of change is not too fast (Brooks and Adger, 2003; Adger and Brooks,
2003). Building adaptive capacity to climate change relies on expanding these coping ranges and
thus reducing vulnerability.
Figure 1: Diagrammatic representation of the conceptual framework of vulnerability, coping range
and adaptive capacity (source: adapted from Smith, 2001)
3
Measuring vulnerability-indicators and indices
3.1
What are indicators and why do we need them?
Indicators are quantifiable constructs that provide information either on matters of wider significance
than that which is actually measured, or on a process or trend that otherwise might not be apparent
(Hammond et al, 1995). Essentially they are a means of encapsulating a complex reality in a single
construct. Gross domestic product (GDP), for example, is created by summing the dollar output of
final goods and services in an economy over a given time period (usually a year), and is a general
proxy measure for the vitality of an economy. By summarising the totality of a number of complex
and intangible processes indicators are of use to decision-makers at all levels, particularly in
comparing across space and monitoring change over time. The UK sustainable development
strategy, for example, is monitored through the use of 15 headline indicators, 147 core national
indicators, and 29 local indicators all pertaining to aspects of economic growth, social progress and
environmental protection (DEFRA, 2002; DETR, 1999).
3.1.2 Aggregating indicators to form an index
In addition to being used in their own right, indicators can be aggregated to form indices. The
advantage is that a wider range of variables can be incorporated, ideally leading to a more
comprehensive model of reality. The World Economic Forum, for example, has created an
Environmental Sustainability Index based on 67 variables represented by 22 indicators within 5
broad dimensions (environmental systems, reducing environmental stresses, reducing social
vulnerability, social and institutional capacity, and global stewardship) (WEF, 2000). Likewise the
UNDP Human Development Index is an annually-updated composite index measuring three
dimensions of human development; a long and healthy life, knowledge, and a decent standard of
living (UNDP, 2002). It is arguably one of the most common benchmarks against which
development is measured, and can highlight non-progressing countries for multilateral aid
assistance.
3.2
Strengths and weaknesses of indicators and indices
Indicators and indices are thus useful for encapsulating a complex reality in simple terms and
permitting comparisons across space and/or time. However in providing useful summary
information there is a danger that indicators may not accurately represent the intended condition or
process – that is they may not be valid. The more complex the reality and more intangible the
processes that the indicator is trying to capture, the greater the danger of this occurring. For example
globalisation is a contested phenomenon, variably defined by people of different academic
backgrounds, and now accepted to incorporate a variety of processes operating and manifesting
themselves on a variety of scales. The multiple facets of globalisation are therefore difficult to
encapsulate in an indicator: instead the various theories of globalisation must be drawn upon in order
to choose proxies of the process from which indicators can be created using existing quantifiable
data. Thus on the economic side such indicators as trade dependence are deemed representative of
globalisation, whilst on the social side access to the information and knowledge economy might use
number of radios or telephone lines as a proxy.
Aggregating indicators creates even more opportunities for subjectivity and thus must be even more
critically appraised. Whilst the purpose of indices is to better encapsulate a complex reality such an
undertaking is limited in several ways. By their very nature, the role of indicators is to capture an
intangible process so it is not possible to “ground truth” them, and alternative means of validation
must be sought. Even with a comprehensive understanding of the conceptual and theoretical
underpinnings of the processes and conditions involved, indicators can necessarily only be snapshot
in time and thus are limited in their ability to represent dynamic processes. More often than not the
method of aggregating the indicator scores does not allow for the contribution of a variable to be
conditional on, or amplified by, another variable, thus there is no way of accounting for the
feedbacks, non-linearities and synergies that exist in real systems (Lohani and Todino, 1984). The
index is also very much contingent upon the choice of indicators at the lowest level, and there is a
real possibility that uninformed choices at this level filter through and can lead to an invalid index.
A critical evaluation of the appropriate use and limitations of indices is even more imperative given
the fact that they link science and policy. By summarising and simplifying reality they are
inherently useful to policy-makers, but the absolute certainties required are often incompatible with
the uncertainties of science. To ensure the most robust and durable results, indicators and indices are
never complete: rather they are in a process of evolution whereby a tentative theoretical proposition
is empirically tested and the results fed back into conceptual development after peer review through
expert judgement. The result is a continual process of refinement so that the indicators and index
have the greatest possible validity and thus utility. Hypothesising new valid and reliable indicators
and indices is thus critical for ensuring the ongoing development of policy-relevant science,
particularly in a theoretically diverse field such as vulnerability.
3.3
Review of vulnerability indicators and indices
The nature of vulnerability is fundamental in determining whether hazard exposure will translate
into impacts or be mediated by the biophysical and/or human systems. At the same time,
vulnerability as a potential state is difficult to assess due to the variety of determinants acting and
interacting on different scales. It is therefore necessary to rely on indicators that best represent the
complex underlying processes. These approaches have largely evolved over the last 10 years or so
in an attempt to build on existing case study based approaches developed primarily with regard to
biophysical vulnerability. The expansion of conceptual and theoretical debates surrounding social
vulnerability has also prompted recognition of the need to develop more systematic indicators to
contribute to more holistic impact studies (Adger, 1999).
There have been several attempts at developing national level indicators and indices for human
aspects of vulnerability, each varying in the nature of vulnerability addressed, the hazard involved,
and the geographical region (see Appendix A for a summary). There is a strong trend of each index
building on and attempting to refine its predecessors by adding to the complexity. This can occur
through a variety of means, for example increasing the number of variables considered, and/or using
more sophisticated techniques of econometric and statistical modelling to transform and aggregate
the indicators (see Appendix A). Briguglio (1995) was amongst the first, recognising the inadequacy
of GDP as an indicator of vulnerability for small island developing states (SIDS). Crowards (1999)
and Easter (1999) then attempted to include more indicators of economic vulnerability with specific
focus on the Caribbean and Commonwealth small states respectively. Kaly et al (1999a) expanded
on economic vulnerability for SIDS in the South Pacific by including elements of environmental
resilience and integrity (i.e. biophysical vulnerability). Others have taken more global approaches to
assessing vulnerability and resilience explicitly in regard to climate change (UNEP, 2001; Moss et
al, 2001). However, despite widespread acceptance of its state of vulnerability, as yet no national
level indicators have been created for Africa.
3.4
Good practice in creating an index of social vulnerability
Despite the weaknesses of indicators and some of the difficult methodological choices involved in
creating vulnerability indices, there is a need to develop existing work and in particular to quantify
social vulnerability. A single-value measure of vulnerability based on meaningful criteria has a
variety of practical applications, particularly at the national level. To ensure maximum validity and
utility of the index good practice ought to be followed. It should be intuitively comprehensible and
impartial. Indicator choice should be such that the index is able to differentiate among countries and
therefore be suitable for international comparisons. The method of construction should be
transparent, with results presented in breakdown and single figure formats (Andrews and Withey,
1976). Perhaps most importantly, the indicators and index should be refinable after testing so that
the model is in a continual process of improvement. The next section critically reviews the
methodological issues involved in the creation of an index, and justifies the chosen methods.
4
Constructing an index of social vulnerability
4.1
Methodological issues
A meta-analysis of major national level indices in the field of vulnerability and beyond highlights
the methodological issues embodied in the various stages of their creation (Appendix A). One of the
most fundamental choices is between a data-driven (inductive) or theory-driven (deductive)
approach (Niemeijer, 2002). In the former a large number of potential vulnerability indicators might
be chosen in what has been labelled a vacuum cleaner approach (UNEP, 2001). Final selection
might occur by means of expert judgement (Kaly and Pratt, 2000; Kaly et al, 1999a, 1999b), or
principle components analysis to determine those that account for the largest proportion of
vulnerability (e.g. Easter, 1999). However, the weakness in this is that a proxy variable for
vulnerability must be chosen as the benchmark against which indicators are tested, somewhat
paradoxically as the very need for vulnerability indicators is because there is no such tangible
element of vulnerability. In this research, therefore, the theory-driven approach is favoured,
whereby use is made of existing theoretical insights into the nature and causes of vulnerability to
select variables for inclusion, although in practice this necessarily occurs within the limits placed by
data availability (Briguglio, 1995). This inevitably leads to subjectivity in the choice of indicators,
but this can be addressed by ensuring all decisions are grounded in the existing literature and made
fully transparent.
4.2
Choice of indicators as determinants of vulnerability
Building on the human-ecological and political-economic approaches, the aim of this index is to
capture the operation and the dynamics of the processes that give rise to national level social
vulnerability to climate change-induced changes in water availability, as the chosen environmental
risk. Evaluating the existing vulnerability studies illustrates the need to consider not only economic
factors but also non-market social, cultural and institutional factors which mediate social
vulnerability (UNEP, 2001). In the context of the conceptual framework (figure 1), this means
identifying the factors that influence how narrow or wide the coping range, signifying vulnerability
and resilience respectively. Having made a theoretically informed decision on the determinants,
simple and easily comprehensible indicators or proxy indicators need to be chosen, within the
constraints of data availability. Making such choices is an inherently subjective process, and
therefore it is important to outline the theoretical arguments for inclusion and assumptions relating to
their functional relationship with vulnerability (i.e. whether it is a direct or inverse relationship).
4.2.1 Economic well-being and stability
The relationship between economic status, commonly measured as GDP, and vulnerability is a
complex and contested one. Economic factors inevitably play a key role in affecting a nation’s
vulnerability: there is a consensus that a strong economy acts as a safety net in the case of
environmental risk and hazard exposure, both pre-event through enabling anticipatory coping
strategies such as insurance and post-event in responding to a shock (e.g. Cannon, 1994; Burton et
al, 1993). However, experience shows that even the most economically developed nations may be
vulnerable in the case of hazard exposure (e.g. hurricane Andrew in Florida in 1992 or the Kobe
earthquake in 1995). Indices aimed at explaining economic vulnerability have highlighted how
vulnerability, fragility and lack of resilience in the face of external forces can be determined by a
wider range of forces, independent of the overall level of development of a country (e.g. Crowards,
1999; Briguglio, 1995).
Despite the relationship between economic development and vulnerability being complex, there is
still a need to include measures of economic well-being and stability in an index of social
vulnerability. Individuals with good access to resources arguably have a safety net in the case of
environmental risk and exposure, allowing them to draw on other resources to maintain their
livelihoods, and hence widening the range or intensity of hazards with which they can cope. Those
individuals with limited economic entitlements have a higher degree of dependence and are arguably
less resilient in the case of shocks to their livelihoods. There are a number of indicators that could
be used to reflect economic status. Simple GDP and human poverty indicators have been discounted
as they are average measures and therefore can distort the picture by failing to capture the subnational inequalities in wealth distribution that characterise many developing countries (Kates, 2000;
Adger, 1999). The UNDP Human Development Index uses an indicator of poverty/standard of
living that refers to the size of population below the poverty line. If a country has a large number
below the poverty line it can be assumed that they will have more limited resilience in the face of
risks and hazards, and in such cases exposure might be more likely to translate into an impact.
It is particularly difficult to capture such time-specific effects on vulnerability, so proxies are
necessary. When people are vulnerable to hazards and risks and have poor entitlements, migration
can occur in response to shocks (Meze-Hausken, 2000). In developing countries, where natural
resource-dependent rural livelihoods are predominant, high rates of rural-urban migration can be a
sign of lack of resilience and narrow coping ranges in rural populations (e.g. Adger, 2000a).
Furthermore, far from assuming that it is only rural populations that are vulnerable, if there is a high
rate of urbanisation caused by rural-urban migration it is highly likely that the new migrants to the
city will also be increasing their personal vulnerabilities by leaving behind the social networks and
collective institutions that might have facilitated adaptation (Adger, 2001; Moser, 1996).
Rate of urbanisation has not been collected per se, and so in this index the change in the percentage
of urban population (calculated as a proportion of mid-year population) between 1975 and 2000 will
be included, on the assumption that high rates of change are indicative of rural livelihoods being
vulnerable. If, however, the percentage of urban population has remained fairly static, or only
increased slightly (in degrees which may be explained by changing overall population sizes), it
might be assumed that rural populations have wider coping ranges, lower baseline vulnerabilities,
and are more resilient in the face of routine risks/hazards. Given that both these indicators capture
elements of economic well-being and stability, they will be aggregated to form a composite subindex of this name, which in turn will be a component of the SVI.
4.2.2 Demographic structure
In addition to economic well-being and stability being important in the resilience of populations to
environmental shocks, the structure and health of the population may also play a key role in
determining vulnerability. Age is an important consideration as the elderly and young tend to be
inherently more susceptible to environmental risk and hazard exposure (O’Brien and Mileti, 1992).
In general terms, populations with a low dependency ratio (high proportion of working age adults)
and in good health are likely to have the widest coping ranges and thus be least vulnerable in the
face of hazard exposure.
The majority of Africa comprises countries of low and middle development status, which by
definition have high birth rates and declining death rates. The result is often expanding populations
and high dependency ratios, formed largely out of the under 15 age group rather than the over 65 age
group. However with progressing development status, as death rates begin to fall and life
expectancy rises, this latter sector of the population is starting to increase.
In classifying under 15s as dependent, criticisms have arisen that varying proportions of this sector
do in fact contribute to household livelihoods and GDP (through the informal economy) in many
developing countries. On the aggregate national level, however, they act more as a burden on
country budgets, particularly with the growing moral commitments to education encapsulated in the
UNDP and World Bank Millennium Development Goals1. Education is an expensive provision that
is largely funded by the population of working age, and although it might help to reduce
vulnerability in future generations, the financial burden acts to increase current vulnerability by
diverting scarce resources. Likewise although low absolute numbers of the elderly means it is not
yet a major issue, as life expectancies continue to increase pressure will be exerted on the working
population from above as well as below. This will mean that not only is a larger proportion of the
population more inherently vulnerable, as children and the elderly tend to be, but that the increasing
demands they place on the working population will act to reduce their resilience through the sharing
of resources.
Whilst the degree of the dependency ratio is an important indicator of a country’s vulnerability in its
own right, in the last 50 years a new issue has emerged that further threatens the demographic
resilience of a population – that of HIV/AIDS. Many sub-Saharan African nations have now
reached epidemic levels of the disease, with Botswana, the most affected country, having a 38.8%
incidence among the working population (aged 15-492). As a debilitating illness, prevalence of
HIV/AIDS not only increases vulnerability to shocks amongst those affected, but also increases the
vulnerability of the aggregate population by further diverting scarce financial resources into health
care provision. As many of those are affected are the working age population, between 15 and 49,
this compounds the problems of dependency caused by the demographic structure. As a result, the
demographic structure sub-index will incorporate indicators of both elements of vulnerability:
dependent population being represented by the % of population aged under 15 and over 65; and
HIV/AIDS being represented by the % of population of working age (15-49) with the disease.
The two sub-indices described above refer largely to elements of population resilience in the
immediate face of exposure to environmental shocks or hazards. Theoretically speaking a
population that is economically secure, stable and healthy is more likely to be resilient. This is due
to their ability to draw on alternative entitlements in the face of a shock such that coping range
1
www.developmentgoals.org [accessed 7th July 2003]
Inconsistencies in data collection mean that the population aged 50-65 fall between the gaps and are unaccounted for in this
sub-index. HIV/AIDS is a relatively recent phenomenon, and arguably rates of incidence are higher among the younger
cohorts of the population. Therefore it was decided not to attempt to manipulate the existing data and risk error given the
negligible likely impact of this inconsistency.
2
thresholds are not exceeded. It may thus be said that such populations have a capacity for
anticipatory adaptation to reduce their vulnerability to impacts (Klein, 2002).
In other circumstances, risk exposure may lead to the occurrence of hazards and shocks to the
natural environment, particularly if it has high biophysical vulnerability. However, if the human
population in the unit of exposure has coping measures in place for reactive adaptation, that is if it
has a low level of vulnerability, such biophysical impacts may not necessarily translate into human
impacts. There are several determinants of this type of social vulnerability that also need to be
captured in a comprehensive index.
4.2.3 Institutional stability and strength of public infrastructure
In the wake of an environmental shock or hazard brought about by biophysical vulnerability, the
institutional stability and strength of public infrastructure are of paramount importance in
determining the coping range of a population, and therefore whether it is vulnerable or resilient. A
well-connected population with appropriate public infrastructure will be able to deal with a hazard
effectively and reduce, if not stop completely, the biophysical effects translating into human impacts
(Handmer et al, 1999). Such a society could be said to have low social vulnerability. Likewise in
reverse, if there is an absence of institutional capacity in terms of knowledge about the event and
ability to deal with it, then such high vulnerability is likely to ensure that biophysical risk turns into
an impact on the human population.
There are a variety of indicators that could be used to represent institutional stability. Given the
prevalence of civil and inter-country conflicts that have plagued many African nations in the postindependence period, a comprehensive attempt at capturing social vulnerability ideally needs to have
some measure of the strength and stability of government. However, such statistics are obviously
politically charged and thus there is no real robust measure constructed by the international
organisations. It might be possible to compile one based on recent historical analysis country by
country, but the durability of such an indicator might be suspect given the inherent subjectivities
required in defining “a conflict” and finding such evidence in documentary sources. In order to
maintain the quality of the index, theoretically-driven proxies available in the transparent and
respected data sets compiled by such international organisations as the World Bank and United
Nations Development Programme are preferred.
Fortunately a wide range of proxy data for strong governance is collected, of which one is the
amount of resources channelled into public service provision, such as healthcare. If the institutional
structure is weak, health expenditure tends to be low with high reliance on private provision. If
governmental organisation is more effective it is likely to be a more efficient provider of public
healthcare services (Smith, 2001). Secondarily, as seen above in the demographic structure subindex, healthy populations are likely to be more individually resilient in the face of environmental
hazards. For these reasons the public health expenditure as a % of GDP is included as a component
indicator.
Whilst public health expenditure is an important proxy of institutional stability it is by no means the
only possibility, and so to ensure robustness of the sub-index other indicators of the strength of
public infrastructure ought to be incorporated. At the scale of the country, social vulnerability is also
determined by the distribution of institutional strength which health expenditure as a proportion of
GDP may not accurately reflect: for example capital cities usually have far superior infrastructure
and facilities to other cities and the rural areas. As distribution is very difficult to capture a second
indicator in this sub-index helps to reinforce the durability of the index. Public infrastructure can act
as a “lifeline”, facilitating circulation of people, goods, services, and information (Platt, 1995).
Access to information and communications infrastructure is arguably important in influencing
vulnerability (Blaikie et al, 1994).
Data exists for several potential indicators here, for example number of radios and televisions, but
the number of telephones (excluding mobiles) standardised per 1000 population is chosen for several
reasons. First of all, telephones require physical infrastructure through wiring etc, and therefore the
number might give an indication of the penetration of such facilities throughout the country. Such
physical infrastructure requires constant maintenance to be effective, and would likely be amongst
the first casualties of budget cuts in the case of civil strife. In this case, therefore, a high number
might be indicative of internal political stability.
Second, telephones, unlike mobiles, are arguably in the latter stages of the product life cycle,
meaning that they are more widely accessible and available at a lower price, so there is no theoretical
barrier to their widespread use, even in the developing world. That they are still not omnipresent
suggests that their use is indicative of wider processes that are theoretical determinants of
vulnerability as hypothesised here, such as institutional strength and stability. In their own right
telephones play a key role in mediating social vulnerability by acting as an access point to
information, specifically that relating to hazard risk, and do not require high levels of literacy or
formal education to use. The greater the connectivity of the population to information services, then
it is easier to promote disaster preparedness and early warnings that could substantially reduce
vulnerability in the face of a hazard. Without such infrastructure early warnings are inhibited and
preparedness measures ignored, thus making populations more vulnerable. Whilst the indicator
gives no explicit consideration to distribution, it may be that although greater proportions may be
found in urban areas, the higher the number the greater the likelihood of overspill out to the rural
hinterland.
It is clear that there are many ways that institutional strength and stability of public infrastructure
may govern social vulnerability. A strong institutional setting can promote resilience in the face of
environmental risk exposure by ensuring appropriate monitoring of the hazard, information
dissemination to the public, and the facilitation of emergency preparedness and pre-disaster
planning, all of which reduce baseline vulnerability. Perhaps more importantly, a strong public
infrastructure and effective institutions can be used post-hazard to ensure it does not translate into an
impact. This might occur by facilitating collective coping mechanisms and perhaps redistributing
resources, for example ensuring food aid if rural livelihoods in one part of the country have been
destabilised in the wake of a drought. Even in the event of a hazard with a greater areal extent,
perhaps one which covers the majority of the country, a strong institutional setting is likely to have
been better prepared through social insurance mechanisms (e.g. food storage) and even through
forging reciprocal relations with neighbouring countries or other structures in the international
community. The institutional nature and strength of public infrastructure are often a function of the
stability of the ruling political regime. Unfortunately there is no direct indicator of political stability.
Public health expenditure and the number of telephones have thus been chosen as proxies, on the
assumption that a stable regime will be committed to equitable distribution and maintenance of
institutions and infrastructure.
However, whilst ideally a strong institutional structure ought to reduce social vulnerability, there
might be cases where political issues such as corruption act to impede equitable access to resources
and distribution of entitlements. In such cases, far from reducing social vulnerability the
inappropriate use of institutions may well increase it, at least for certain sectors of the population at
the expense of others (Robbins, 2000). Perhaps unsurprisingly, corruption is a complicated
phenomenon to quantify even if it can be observed in the first place. Transparency International
have been developing a Global Corruption Index over the past 5 years using a comprehensive and
transparent methodology (Hodess, 2003). As yet it only exists for 22 countries in Africa, which is
less than half and yields too many missing values for substitutes or predicted values to be quantified
through averaging or regression. However, given the importance of corruption as a determinant of
social vulnerability, a second index B has been created for the 22 countries. In this index corruption
appears as a third indicator in the institutional stability and strength of public infrastructure subindex.
4.2.4 Global interconnectivity
Whilst issues of internal structure and functioning play an important role in determining national
level vulnerability, many domestic issues are also increasingly interlinked with and dependent upon
processes operating on a global scale. Such trends towards globalisation are particularly evidenced
with the integration of domestic economies into a global market. However, the pattern of
globalisation tends to exploit and in turn reinforce existing inequalities in the global economy,
creating winners and losers at a variety of scales (O’Brien and Leichenko, 2000). Whilst the “triad”
of North America, Europe and East/Southeast Asia has benefited, the continent of Africa has become
marginalized as a result (Castells; 1998; Agnew and Grant, 1997; Castells, 1996). The upscaling of
comparative advantage to the global level has largely restricted Africa’s economic activities to the
supply of raw materials, locking countries into the global economy but on unfavourable terms,
subject to fluctuating demands and highly variable prices that characterise commodity markets.
On the continental scale such trends render Africa vulnerable, but at the national and even regional
scale globalisation processes are also uneven (Hirst and Thompson, 1996), providing opportunities
that allow for differentiation between countries. Whilst the ability of cities to exploit their
communications and industrial resources is well charted, studies have also shown there to be a
variable distribution of winners and losers in rural sectors such as agriculture (Leichenko and
O’Brien, 2002). It is therefore necessary to choose an indicator which captures the overall
continental state of global inter-connectivity whilst also being sensitive enough to differentiate
between country level differences when trying to determine national vulnerabilities. Possible
indicators include the following: flows of aid, private capital and debt, foreign direct investment,
other non-governmental development assistance and the structure of trade. However the trade
balance has been selected as the most appropriate for giving an indication of global
interconnectivity. Positive balances are assumed to indicate countries that are making the most of
the strategic opportunities afforded by globalisation, and hence have the networks and connections,
as well as the financial resources, to promote post-event coping. Those with negative trade balances,
in reverse, are likely to be more vulnerable through having a narrower coping range.
4.2.5 Natural resource dependence
As noted in the introduction, vulnerability is largely hazard-specific: it is perfectly possible for a
population to be vulnerable to one hazard yet resilient in the face of another. Such a status depends
on a variety of factors such as experience of past exposure and anticipatory coping mechanisms. For
example a dryland population with a long history of exposure to rainfall variability may have been
able to reduce their vulnerability to impacts by adapting their lifestyle (e.g. through migration) and
livelihoods (e.g. by adopting a diverse and flexible strategy). Such preparatory (anticipatory)
adaptation measures, in conjunction with institutional stability and strong public infrastructure which
may promote social insurance mechanisms, may reduce the vulnerability of this population to this
hazard. However, such measures have no impact on their vulnerability to other hazards, for example
pest infestations of their crops/herds. Perhaps more importantly in the context of climate change,
they may only be adapted to the frequency, temporal spacing, magnitude and areal extent of an
existing hazard/variability. Should these circumstances change, as for example water availability is
projected to do so with my climate change in many areas, their mechanisms may be incapable of
keeping the population within the appropriate coping range, and thus they may become vulnerable.
Whilst the indicators and sub-indices mentioned hitherto refer to fairly generic determinants of
vulnerability, which would apply universally in almost any case of unexpected hazard, the natural
resources dependence sub-index has been included to refer explicitly to vulnerability to changes in
water availability. This is because change in water availability is a major projected impact of
climate change on Africa (Arnell, 1999). Such a change may increase baseline vulnerability, due to
the high dependence on natural resource-dependent livelihoods relating to primary industries, for
example agriculture, fishing and forestry, the productivity of which is a function of water
availability. As a result the percentage of rural population has been included as an indicator.
4.3
Collecting the data and confidence in data
Whilst the aim of this SVI is to be theory-driven, as suggested above this necessarily has to take
place within the limits of the availability of robust and transparent comparable data. Once potential
determinants were identified, a range of appropriate indicators and proxies were considered, with the
most appropriate selected according to theoretical insights in the literature. Wherever possible, to
avoid errors arising out of comparing incomparable data, indicators were selected from indicators
routinely created by international organisations. Table 1 gives a summary of variables, indicators
and data sources used in the SVI.
The majority of indicators used in the index come from the World Bank. The World Bank compiles
approximately 800 World Development Indicators from data that are derived, either directly or
indirectly, from official statistical systems organised and financed by national governments. A
dedicated International Comparison Program is part of an ongoing process of improving the
transparency, reliability and comparability of data. One of the most commonly used measures for
ensuring comparability in financial statistics is an exchange rate conversion factor called purchasing
power parity (PPP), which takes account of price differences between countries.
The only non-international organisation statistic is the global corruption index. This composite
index gives each country a normalised score from 0-10 where 0 is bad, and is compiled using 15 data
sources from 9 institutions, amongst them the World Economic Forum, Economist Intelligence Unit,
and Columbia University. The use of multiple data sources increases reliability as erratic findings
can be balanced by the inclusion of at least two other sources. To ensure continuity and comparison
of like with like there are two criteria for the inclusion of data: it must include a rank and must
measure the overall level of corruption, itself a conservative measure chosen to ignore those that mix
corruption with political instability, nationalism etc. The index has been compiled over the last 5
years, with each annual revision building on the robustness: the 2003 refinement allowed for
checking correspondence of residents’ viewpoints on the nature and level of corruption with those
held by expatriates.
Table 1 – summary of variables, indicators and data sources used in the SVI
Determinant
of
vulnerability/s
ub-index
Economic
well-being and
stability
Demographic
structure
Component
indicators
What each indicator represents:
Hypothesised functional
relationship between
indicator and vulnerability
Data source
Standard of
living/poverty
population below income poverty line,
2000. The % of the population living below
the specified poverty line.
World Bank
(2002)
Change in % urban
population
change in % urban population between
1975 and 2000, based on midyear
population of areas defined as urban in a
country.
The greater the population
below the income poverty
line, the greater the
vulnerability.
The greater the change in
urban population the
greater the vulnerability.
Dependent
population
population under 15 and over 65 as % of
total, refers to de facto population, i.e. all
people actually present in a given area at a
given time.
Adults aged 15-49 living with HIV/AIDS
as a percentage of the population aged
between 15-49 in 2001.
Proportion of the
working population
with HIV/AIDS
Institutional
stability and
strength of
public
infrastructure
Health expenditure
as a proportion of
GDP
public health expenditure as % of GDP in
1998: recurrent and capital spending from
central and local government budgets
(including donations from international
agencies and NGOs) and social (or
compulsory) health insurance funds.
Telephones
number of mainland telephone lines per
thousand population in 2000.
Corruption
composite index using data from 15
sources from 9 institutions and perceptions
of well informed people with regard to
corruption, in 2002.
Net trade in goods and services (BoP,
current US$, 1999). Derived by offsetting
imports of goods and services against
exports of goods and services. Exports and
imports of goods and services comprise all
transactions involving a change of
ownership of goods and services between
residents of one country and the rest of the
world.
% of rural population, defined as the
difference between the total population and
urban population in 1999.
Global interconnectivity
Trade balance
Natural
resource
dependence
Rural population
UN (2002)
The higher the dependent
population, the greater the
vulnerability.
UN (2001)
The higher the proportion
of working population with
HIV/AIDS, the higher the
vulnerability.
The higher the health
expenditure as a proportion
of GDP, the lower the
vulnerability (inverse).
UNAIDS and
WHO (2002)
The higher the number of
telephones, the lower the
vulnerability (inverse).
The lower the score (i.e. the
higher the corruption), the
higher the vulnerability
(inverse).
The more negative the trade
balance, the higher the
degree of vulnerability
(inverse).
ITU (2002)
The higher the rural
population, the greater the
vulnerability.
World Bank
(2002)
Transparency
International
(Hodess, 2003)
World Bank
(2001)
World Bank
(2002)
4.3.1 Missing value analysis
In addition to ensuring appropriate data quality, attempting to rely on established and well-reputed
sources that are routinely gathered for all countries worldwide also helps to reduce the occurrence of
incomplete data sets. However, with the theory-driven nature of the index, there are inevitably
occasions when a component indicator has missing values, and therefore some form of missing value
analysis needs to occur. Of the 52 sovereign states in Africa, the SVI is calculated for 49, with
Liberia, Sao Tome e Principe and Somalia the only exclusions on the basis of lacking data in
substantial counts (at least 3 indicators). As mentioned above, the 22 African values in the
corruption index were deemed too few to construct meaningful substitutes for the missing values,
and hence this is included only for the countries with an actual value as an alternative B index.
Where incompleteness was restricted to less than half the countries, alternative measures were taken.
This was applicable to 2 data series; standard of living/poverty (19 missing values) and number of
telephones per 1000 population (1 missing value). In order to select an appropriate method of
missing value analysis each data series was eyeballed and basic descriptive statistics obtained. For
standard of living/poverty the mean of the series was applied. For telephones, where the missing
value was the Democratic Republic of Congo, data for neighbouring countries was assessed and a
value of 7 ascribed, as this corresponds to the figure for the Republic of Congo. Whilst substitution
for missing values may be contentious on the basis of subjectivity, on occasion it is unavoidable
given the variable data availability reflecting such factors as changing political boundaries and civil
conflict. The only solution is to make such choices transparent in order to enable effective critical
evaluation of the robustness of the index. Democratic Republic of Congo has had two missing
values calculated, for all other countries the maximum is one (for standard of living/poverty). The
results table clearly marks those countries where missing values existed (Appendix C).
4.4
Methods of aggregation to form a composite index
Having considered the theoretical determinants of national level social vulnerability and selected
appropriate indicators to capture this, further methodological choices need to be made relating to the
standardisation of indicators, and their means of combination into a single index of social
vulnerability. As vulnerability indices have developed more sophisticated means of econometric
modelling and mathematical transformation have been attempted (Appendix A). The variety of
approaches will be reviewed and evaluated and justifications given for choices in the context of a
theoretically-driven index.
4.4.1 Standardisation of indicators
Having selected indicators based on their theoretical role in determining social vulnerability, it is
necessary to carry out some form of standardisation to ensure that they are comparable. There are
several means by which this may occur. Most simply, standardisation fits variables to relative
positions between 0 and 1. Some indices applied a normalisation procedure so that rather than
refitting the actual range of values across the 0-1 scale, they are fitted to a normative scale of what is
deemed high and what is deemed low. In the UNDP Human Development Index, for example, the
GDP component index is calculated using goalposts of $40,000 as high (1 on the index) and $100 as
low (0 on the index) (UNDP, 2002). Likewise in the Environmental Vulnerability Index actual
values are normalised onto a categorical scale such that each indicator is ascribed a value of 0-7
(Kaly et al, 1999); and the Caribbean Vulnerability Index experiments with condensed decile
normalisation aimed at emphasising sensitivity to extreme values in the indicator ranges (Crowards,
1999). However, normalisation adds an extra element of subjectivity, and may disguise any
interactions between indicators. Whilst that may be useful in attempting to quantify actual
vulnerability, as the purpose of this study is to assess relative levels the simple standardisation
method will be used. All indicators are standardised so that the highest value in the range equates to
1, and the lowest value in the range equates to 0. In some cases, for example inverse relationships,
this involves a transformation so that the highest figure always equates to the greatest vulnerability.
4.4.2 Creating the sub-indices
Having standardised the indicators an appropriate means of creating the sub-indices needs to be
selected. In a data-driven index this would require that the most appropriate indicators of
vulnerability be selected from the shortlist. In the theoretically-driven approach, however, the
importance of each of the variables in affecting national level social vulnerability has already been
determined. It is clear that the variables fall into several distinct groups, and thus it makes sense to
use these categories as the sub-indices in the SVI: namely economic stability and well-being,
demographic structure, institutional stability and strength of public infrastructure, global
interconnectivity and natural resource dependence. The first three of these contain multiple
indicators, and therefore appropriate means of aggregation need to be employed.
There are a variety of examples of aggregation choices made in existing indices. The simplest is to
maintain the status quo and aggregate indicators on an equal basis, as is done in the Environmental
Sustainability Index (WEF, 2000). The problem with this is that it can overcredit one factor
(Niemeijer, 2002), so often a means of weighting is employed. Briguglio (1995), for example, used
two methods to create his index, the first employing equal weighting and the second using non-equal
weighting, to reflect the perceived importance of the various indicators in promoting vulnerability.
Crowards (1999) built on this by using both non-weighted and PCA-derived weights in his
Caribbean Vulnerability Index. The aim of the SVI is to be comprehensive, requiring a wideranging set of indicators within each sub-index category. Whilst there is a strong basis for their
theoretical involvement of each indicator, there is no reason to suggest that their roles are equal. The
application of some sort of weighting is therefore appropriate.
Following the theoretical underpinnings of the index, both the conceptual aggregation of indicators
in the sub-indices and the various weightings have been derived based on existing literature
combined with discussions held with various Tyndall Centre colleagues who can be considered
experts in various elements of vulnerability and adaptation. On this basis the economic well-being
and stability sub-index comprises the standard of living/poverty indicator accounting for 80% and
the growth in percentage urban population for 20%. The demographic structure sub-index comprises
dependent population and the proportion of working population with HIV/AIDS equally weighted.
The institutional stability and strength of public infrastructure index appears in two versions, A and
B. In version A health expenditure is weighted to account for 80% whilst the number of telephones
per 1000 population accounts for 20%. In version B, the corruption index is included for those
countries for which data is available, with a weighting of 60% for health expenditure, 20% for the
corruption index and 20% for the number of telephones. The global interconnectivity and natural
resource dependence sub-indices comprise only one indicator each, and therefore each indicator
counts for 100%. Details of the method of standardising and aggregating indicators for a case study
example can be found in Appendix B.
4.4.3 Combining the sub-indices to form the overall aggregate index
Having derived the five sub-indices a similar range of methodological concerns need to be addressed
when deciding how to aggregate these into the final composite index of social vulnerability.
Jollands and Paterson (2003) make the distinction between aggregate indices, where the constituent
parts are not recognisable, and composite indices, where they are. Typically aggregate indices
involve a scalar function, whilst composite indices involve merely presenting a matrix of component
indicators. This research aims to combine both approaches, by using an explicit scalar function
within the conceptual framework to create a single aggregate overall score, but also with a
commitment to transparency in the composite make-up of that score. Therefore the overall index
will be formed from weighted average of the sub-indices, with weights derived from theoretical
understanding. The aggregate figure will therefore be a number between 1 and 0, with 1
representing the highest level of vulnerability. Based on evidence in the literature and expert input, a
decision has been made which applies the following weights: 20% to economic well-being and
stability, 20% to demographic structure, 40% to institutional stability and strength of public
infrastructure, and 10% each to global interconnectivity and natural resource dependence. The
overall equation summarising the model employed for the SVI for each country is thus:
SVI = Σ (Ii*Wi)(Iii*Wii)(Iiii*Wiii)(Iiv*Wiv)(Iv*Wv)
where Ii =
economic well-being and stability sub-index
Iii =
demographic structure sub-index
Iiii =
institutional stability and strength of public infrastructure sub-index (version A/B)
Iiv =
global interconnectivity sub-index
Iv =
natural resource dependence sub-index
Wi =
0.2
Wii = 0.2
Wiii = 0.4
Wiv = 0.1
Wv = 0.1
5
Results and Discussion
This section summarises the results of both versions of the SVI. Although actual scores are
presented it is worth reinforcing that these have been created by standardising indicators across the
range of data for Africa, not across a normative range with theoretical high and low values.
Therefore those countries at the top end of the range with “high” scores nearing one have the highest
relative vulnerability. The countries at the bottom of the range with “low” scores nearer to 0 do not
necessarily have low absolute human vulnerabilities, rather they are slightly better off compared to
other countries in Africa. Africa, and in particular sub-Saharan Africa, has already been identified as
relatively vulnerable on the international scale. The aim of this index has therefore been to add an
extra element of resolution by refining this evidence and highlighting areas of particularly high
vulnerability. For that reason, wherever possible the results will be examined as part of a complete
series rather than in arbitrary groups. In addition to overall scores, the sub-indices and their
component indicators will be analysed in order to illustrate the variation in composition of the
overall scores.
5.1
The social vulnerability indices
The results of the SVI are presented in table 2 and figures 2 and 3. The country with the highest
level of social vulnerability is Niger, followed by other sub-Saharan countries such as Sierra Leone,
Burundi, Madagascar, Burkina Faso, Uganda, Ethiopia and Mauritania. Figure 2 shows that there is
no discernible geographical trend to high social vulnerability, with certain nations in western and
eastern Africa having the highest vulnerabilities. The countries at the bottom end of the range rather
unsurprisingly are the north African states of Egypt, Morocco, Libya, Tunisia and Algeria, along
with the relatively developed southern African countries of Namibia and South Africa, the Indian
Ocean island of Mauritius and Senegal in the west. Perhaps most surprising is that Djibouti scores
relatively well. Further analysis of its vulnerability profile below will illuminate reasons for this.
Taking corruption into account in index B does not have much effect on the overall ranking, with
Madagascar, Uganda and Tanzania, Cameroon and Ethiopia exhibiting the highest levels of social
vulnerability, and the north African states, Mauritius, Senegal, South Africa and Namibia better off
(figure 3). It is difficult to draw too many conclusions about changing ranks given the different sizes
of the samples. In terms of actual scores, the largest changes between indices A and B are Zambia
(+0.034), Kenya (+0.025), Namibia (+0.024), Angola (+0.022) and Zimbabwe (+0.022),
highlighting the relative importance of corruption in their vulnerability profiles.
Table 2 – results of the SVI
COUNTRY
Niger
Sierra Leone
Burundi
Madagascar
Burkina
Faso*
Uganda
Ethiopia*
Mauritania
Lesotho
Tanzania
Cameroon
Togo*
Rwanda
Ghana
Nigeria
Chad
Angola
Eritrea
Swaziland
Zambia
Guinea
Bissau
Vulnerability Index
A
COUNTRY
Vulnerability Index
B
score
Rank
Madagascar
Uganda
Tanzania
Cameroon
0.697
0.670
0.640
0.637
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Ethiopia*
Angola
Zambia
Nigeria
Malawi
Ghana
Kenya
Ivory Coast
Zimbabwe
Botswana*
Morocco*
Namibia*
Senegal
Egypt
South Africa*
Tunisia
0.635
0.634
0.631
0.624
0.606
0.604
0.603
0.576
0.567
0.537
0.525
0.498
0.489
0.487
0.381
0.341
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
22
Mauritius
0.300
21
Score
rank
0.725
0.705
0.703
0.691
1
2
3
4
0.658
0.657
0.655
0.654
0.649
0.646
0.640
0.633
0.627
0.624
0.621
0.618
0.612
0.601
0.599
0.597
0.591
Dem. Rep.
Congo
Malawi
Botswana*
Mali*
Ivory Coast
Cent. Afr.
Rep*
Benin
Comoros*
Kenya
Rep Congo*
The Gambia
Guinea
Equat.
Guinea
Mozambique*
Sudan*
Morocco*
Gabon
Zimbabwe
Cape Verde*
Namibia*
Egypt
Senegal
Libya*
South Africa*
Tunisia
Algeria
Mauritius
Djibouti
0.591
0.591
0.586
0.585
0.584
22
22
24
25
27
0.584
0.584
0.581
0.578
0.576
0.567
0.562
27
27
29
30
31
32
33
0.561
0.557
0.556
0.550
0.547
0.545
0.543
0.522
0.493
0.481
0.405
0.390
0.368
0.360
0.329
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
0.303
49
*represents countries where missing
value analysis has been applied for
an indicator.
Figure 2
Figure 3
5.2
Individual country vulnerability profiles
The social vulnerability indices are composites formed through weighted averages of the subindices. Whilst there are many advantages to having one overall figure for ease of comparison and
accessibility of the data to non-specialist (policy-making) audiences, there is necessarily a trade-off
between the component sub-indices when they are viewed in aggregated form (Hicks and Streeten,
1979). To add some depth to the overall assessment, therefore, it is important to also look at the subindices and indicators for each country to understand about the composition of vulnerability. A
comprehensive summary of the results for each country, including actual scores for each of the
indicators, actual scores and ranks for the sub-indices, as well as the overall social vulnerability
indices ranks and scores can be found in Appendix C.
Figure 4 shows the distribution of scores by country for each sub-index. In general these are
characterised by fairly smooth distributions, which reinforces how similar many African nations are
in terms of the macro level determinants of vulnerability. Frequently it is only a handful of values at
either extreme that buck the trend and show higher than average or lower than average scores. Many
of the countries that do well in the overall index also have low scores in the sub-indices: this tends to
apply for the north African states and Mauritius.
In terms of economic well-being and stability Zambia appears most vulnerable, whilst Mauritius and
Tunisia score particularly well. An important observation from the demographic structure sub-index
graph is the poor scores of many of the south African countries, namely Botswana, Swaziland,
Lesotho and Namibia, resulting particularly from their high incidences of HIV/AIDS (Botswana is
the highest score in this individual indicator).
The institutional stability and strength of public infrastructure sub-index has 14 countries scoring
between 0.8 and 1 (the highest level of vulnerability), highlighting the importance of this particular
dimension of vulnerability. Djibouti scores best in this regard due to the highest level of health
expenditure (score of 0 in that indicator). The north African states perform slightly less well, giving
way to the southern African states, with South Africa, Namibia and Zambia all showing the lowest
levels of vulnerability in both sub-indices. Many of the highest scoring (most vulnerable) countries
are in West Africa, for example Nigeria, Sierra Leone, Cameroon, Togo, Ivory Coast, Burkina Faso
and Benin, and these tend to score badly in both health expenditure and telephones.
In terms of global interconnectivity, Egypt, Angola, Morocco, Ghana, Uganda and Mozambique
have the highest trade deficits and thus are deemed most vulnerable in this regard. The rest of the
countries score very similarly, with only a small group scoring better: Algeria, South Africa and
Libya, but also Nigeria, which obviously benefits in trade with its oil industry. The results for
natural resource dependence show a more differentiated distribution, with the highest levels of
dependence in Rwanda and Burundi. Libya and Djibouti have the lowest, the latter perhaps
explained by increasing dependence on its strategic location for shipping and trade.
5.3
Evaluation of results
As discussed, aggregate indices play an important role in simplifying multiple processes into a single
figure. However, in doing so there is a danger of overlooking the subjectivity and using the figures
uncritically. The best way of dealing with this is to develop a clear conceptual framework, identify
the assumptions and sources of data, and maintain transparency in the choices of indicators, subindices, and aggregation functions (Jollands and Paterson, 2003; Hammond et al, 1995). Therefore
an evaluation of the validity and reliability of the results depends as much on the critical analysis of
methodological choices in the creation of the index as the figures and rankings themselves.
5.3.1 Data quality and availability
An index is only as good as the quality of the data sources it uses. Despite increases in the quantity
of national-level data collected, limitations of data availability played an important role in the
construction of the SVI. It is difficult to assess the durability and robustness of data sources, and for
that reason the majority of data was selected from international organisations with a long history and
solid reputation. Nevertheless, questions arise over the use of this: for example there are very few
missing values in data series despite such factors as civil war and regime changes having a likely
impact on the gathering of information.
5.3.2 Construct validity
Regardless of the quality of the data, the results are dependent upon how well the various indicators
capture the identified determinants of vulnerability. The most tenuous indicator in the SVI is that
relating to natural resource dependence. Theoretical insights suggested that a measure of the
dependence on water resources be included. Taking as read the physiological needs of water for the
human population, a means of capturing this is to examine the proportion of the population
dependent on water for their productive livelihoods. Data constraints mean that percentage rural
population is the most suitable proxy for this, assuming rural populations largely rely on primary
industries themselves dependent on natural resources, whose productivity is directly linked to water
availability. However, this is making the assumption that rural populations are dependent on
activities such as agriculture, which may not always be the case.
Likewise an evaluation of the index depends on scrutiny of how well assumptions hold about the
functional relationship between the indicators and vulnerability. For example with the global
interconnectivity sub-index, data for trade balance has been used as an indicator. Theoretical
insights suggest that those national economies with a negative trade balance are locked into external
market forces on unfavourable terms, and thus are more vulnerable. However, the US has the largest
negative balance, but it is unlikely to be considered the most vulnerable country in the world. It
might be possible that the better integrated a country is into the global economy, the more
opportunities it might have to diversify and thus in fact is increasing its resilience, a phenomenon
that has been investigated at the sub-national level for some sectors (Leichenko and O’Brien, 2002).
Ideally it would therefore be important to ground-truth and validate the precise role of various
indicators in contributing to or reducing vulnerability.
5.3.3 Validation
The SVI essentially comprises predictive indicators of vulnerability based on existing theory.
However it is extremely difficult to validate the effectiveness of the indicators in representing
determinants of vulnerability as indeed the whole objective of the indicators is to capture intangible
processes. A common method for assessing the validity of vulnerability and risk measures involves
looking at correlations with past disasters data (Brooks and Adger; 2003; Pelling and Uitto, 2001;
Crowards, 1999; Easter 1999). Whilst that may determine whether high levels of vulnerability
contributed to hazard exposure translating into an impact, it gives less insight into the situations
where low social vulnerability (high resilience) impeded the occurrence of a disaster. However,
using historical occurrences of disasters and applying the model index to temporally-specific data
might at least act as a means of validation for the structure of the index in explaining social
vulnerability.
5.3.4 Limitations of capturing vulnerability in an index
In addition to the specifics relating to the SVI, a critical evaluation needs to take account of the
limitations of indices in general when assessing vulnerability. Vulnerability is multi-dimensional in
nature and a potential state that is time and scale specific. As a result, an index of social
vulnerability is only a snapshot in time and may disguise ongoing evolutions of certain dimensions.
Similarly it is impossible to represent the inter-relationships between different determinants or
driving processes that interact in different ways according to the temporal and spatial scales of
analysis (Wilbanks and Kates, 1999; Dow, 1992). The majority of data used in the index refers to
annual figures from 1998-2002. The only exception is the growth in urban populations between
1975 and 2000, which actively tries to capture a temporal element that may add depth to the nature
of vulnerability at a given time. The result is an index of current social vulnerability. These
conditions are unlikely to remain constant into the future when climate changes are projected to
occur. However, although some indices have embraced the use of socio-economic scenarios (e.g.
Moss et al, 2001), others suggest that current vulnerability is the best possible proxy (e.g. Adger and
Kelly, 1999), and is appropriate for identifying the means of increasing resilience, coping ranges and
adaptive capacity (Adger et al, 2003).
Using current vulnerability and being unable to capture temporal shifts and assess their potential
effect on the overall social vulnerability must be borne in mind when using the results. Djibouti, for
example, has the highest rate of health expenditure as a percentage of GDP, data which causes the
country not only to score highest (least vulnerable) in the institutional stability and strength of public
infrastructure sub-index, but also in the overall composite index. This result is fairly unexpected
given Djibouti is not amongst the most developed countries in the human development index, nor
particularly reputed for its political commitment to public health. A potential explanation, therefore,
is that this figure is not representative of normal health expenditure. In such a case, social
vulnerability indices would need to be calculated on an annual basis in order to chart change over
time.
As the timescale can be a limitation in the SVI, so can the scale of analysis. A national level index is
important for policy purposes as the state is frequently the unit of analysis, and it is also appropriate
for one of the first systematic assessments of the social vulnerability of Africa. On the national
level, therefore, institutional stability and strength of public infrastructure are the most appropriate
structural determinants of vulnerability. However, higher resolution studies at the sub-national level
might place more emphasis on local institutions and human agency, and how their role in mediating
vulnerability is affected by the changing nature of that state. Social capital, or norms, networks and
reciprocity, may act to increase social resilience at the community level (Adger, 2001, Putnam,
1993). Recent research has highlighted the interaction of the national policy framework and such
institutions, showing how traditional local mechanisms may be eroded by the changing nature of the
state (Adger, 2000b; Scott, 1985). Use of the SVI must therefore bear in mind its use of countries as
the scale of analysis, and that it is not necessarily appropriate for analysing vulnerability at higher
resolutions.
Essentially the subjectivity involved in such an index will always be a problem, but the only solution
is to use theoretical insights to ensure appropriate variables are selected, and then be transparent with
the assumptions and subsequent methods of transformation from indicator to index. By doing this,
the index is as durable as it can be in explaining relative levels of social vulnerability to climate
change-induced changes in water availability across countries in Africa. However as with all indices
it should be subject to a process of continual testing and refinement. The final section identifies
some areas for further research in this regard, and elaborates the potential applications of such an
index.
6
Conclusion and further research directions
This research has derived a theory-driven aggregate index of social vulnerability formed through the
weighted average of five composite sub-indices: economic well-being and stability (20%),
demographic structure (20%), institutional stability and strength of public infrastructure (40%),
global interconnectivity (10%) and dependence on natural resources (10%). The outcome, which
shows current vulnerability to climate change-induced changes in water availability, puts Niger,
Sierra Leone, Burundi, Madagascar, Burkina Faso and Uganda as the most vulnerable countries in
Africa, whilst Djibouti, Mauritius, Algeria, Tunisia, South Africa and Libya are the least vulnerable,
although it is important to remember that this is a relative scale and should not imply that the latter
countries are entirely resilient. By virtue of their role in encapsulating complex and often intangible
processes, the creation of indicators and indices can be a double-edged sword. If used uncritically
they may seriously distort the reality they attempt to represent. As their construction necessarily
involves a level of subjectivity, the most appropriate means of ensuring the durability and robustness
is to maintain transparency in all decisions. This index is theoretically grounded in existing
literature on vulnerability and uses the most durable national-level data sets. If it is used with
appropriate regard to the limitations, it therefore marks the first robust and systematic assessment of
relative levels of social vulnerability in Africa.
The creation of an index of social vulnerability has several applications. The first is a contribution to
the growing academic field of vulnerability. It adds to the ongoing debates about notions of
vulnerability, and helps to operationalise a conceptual framework. It has also made important
methodological advances. The use of indicators is suited to the more developed approach of the top
down, biophysical vulnerability-focused natural hazards school of thought, which arises out of a
positivist epistemology which posits that there is an objective reality that can be quantified.
Applying this methodology to the political ecology approach translates their relativist ontologies into
a language that can be understood by the natural hazards researchers, and thus should help bridge the
gap between the two approaches. This step is a prerequisite for further interdisciplinary studies that
embrace both the biophysical and human aspects of vulnerability, and thus in turn should promote
more integrated climate change impact assessments.
A contribution to conceptual clarity of vulnerability also meets an important policy demand within
the UNFCCC commitments to climate change adaptation3 (Bodansky, 1993; Sands, 1992). With the
understanding that climate change will continue even with the implementation of mitigation policy
instruments, adaptation policy is currently experiencing rapid development (Burton et al, 2002). For
3
The UNFCCC is available online at http://unfccc.int/resource/docs/convkp/conveng.pdf
policy purposes, deciding where adaptation efforts are most required depends on robust vulnerability
and impacts assessments. There have been a wide variety of studies linking projected climate
change to biophysical vulnerability, for example with regard to water availability (Arnell, 2004;
Arnell 1999), malaria incidence (Martens et al, 1999) and food security (Parry et al, 1999), but more
integrated studies that also consider aspects of social vulnerability have largely been limited to small
scale research. This index marks the first attempt at an empirical assessment of relative levels of
social vulnerability for the continent of Africa at a scale appropriate for international decisionmaking. However, it must be remembered that social vulnerability is only one part of the equation –
and the results here hold only if risk exposure and biophysical vulnerability are held constant. In
reality, however, different countries have differing levels of risk exposure and biophysical
vulnerability, and therefore these results need to be used in conjunction with appropriate elements of
biophysical vulnerability to add depth to impact assessments. In this way the results can contribute
to the identification of areas where high biophysical vulnerability and high social vulnerability
coincide. This is important in the immediate term for prioritisation of aid, and in the longer term to
the development of adaptation policy within the UNFCCC.
The most evident policy applications of the SVI relate to the UNFCCC. In its unique capacity of
explicitly addressing equity, the UNFCCC has an article that commits developed countries to
contribute financial and technical resources to developing country adaptation to climate change
(article 4) (Verheyen, 2002). In addition several funds have been earmarked for adaptation purposes
under the Marrakech Accords; the Least Developed Countries Fund (of the UNFCCC) and the
Adaptation Fund (of the Kyoto Protocol) (Dessai, 2003; Dessai and Schipper, 2003). The
administrators of these funds, the Global Environment Fund, acknowledge the need for more
information on vulnerability assessments (GEF, 2000), and the results of this study will assist in the
effective targeting of aid for adaptation and capacity building. They may also provide an input into
decisions regarding the operation of the Clean Development Mechanism, a flexible mechanism of
the Kyoto Protocol whereby Annex I countries can receive certified emissions credits for projects
involving transfer of environmentally-friendly technologies to non-Annex I countries (Yamin, 1998).
Whilst the index therefore has important academic and policy applications, its construction has also
raised a number of potential directions for further research. The need to try and validate the index
by applying the model (with appropriate data) to past historical hazards to explain why they did or
did not translate into impacts (disasters) has already been mentioned. Likewise adapting the hazardspecific sub-index is important to allow examination of vulnerability to other climate-related facets
of global environmental change beyond water availability. In addition to a continual process of
testing and modification required to ensure the robustness of the index, the results highlight areas for
related study. Identifying countries with high social vulnerability is a crucial precursor to higher
resolution studies at the sub-national level, which can be more comprehensive and identify
appropriate contexts in which to investigate local processes of adaptation to climate change and their
synergies with other issues (Leichenko and O’Brien, 2002). This is particularly important as a rank
showing relative resilience on the index does not automatically mean that communities within that
country are not vulnerable. As vulnerability is scale-dependent, and the national level is an arbitrary
scale chosen on the basis that it is the primary scale of analysis in political decision-making, such
local level studies will add depth to the complex realities of vulnerability.
References
Adger, WN. 2001. Social capital and adaptation to climate change. Tyndall Centre Working Paper 8.
Norwich: Tyndall Centre for Climate Change Research. 19pp.
Adger, WN. 2000a. Social and ecological resilience: are they related? Progress in Human
Geography 24 : 347-364.
Adger, WN. 2000b. Institutional adaptation to environmental risk under the transition in Vietnam.
Annals of the Association of American Geographers 90 (4) : 738-758.
Adger, WN. 1999. Social Vulnerability to Climate Change and Extremes in Coastal Vietnam. World
Development 27 : 249-269.
Adger, W Neil and Nick Brooks. 2003. Does global environmental change cause vulnerability to
disaster? In Pelling (ed). Natural disasters in a globalising world. pp 19-42.
Adger, WN; Huq, S; Brown, K; Conway, D and M Hulme. 2003. Adaptation to climate change in
the developing world. Progress in Development Studies 3 (3) : 179-195.
Adger, WN and PM Kelly. 1999. Social vulnerability and the architecture of entitlements. Mitigation
and Adaptation Strategies for Global Change 4 : 253-266.
Agnew, J and R Grant. 1997. Falling out of the world economy? Theorising “Africa” in world trade.
In Lee and Wills (eds) Geographies of Economies.
Andrews, FM and SB Withey. 1976. Social indicators of well-being: Americans’ perceptions of life
quality. New York: Plenum Press.
Arnell, NW; Livermore, MJL; Kovats, S; Levy, PE; Nicholls, R; Parry, ML; and SR Gaffin. 2004.
Climate and socio-economic scenarios for global-scale climate impacts assessments: characterising the
SRES storylines. Global Environmental Change 14 (1) : 3-20.
Arnell, NW. 2004. Climate change and global water resources: SRES emissions and socio-economic
scenarios. Global Environmental Change 14 (1) : 31-52.
Arnell, N. 1999. Climate change and global water resources. Global Environmental Change 9 : S31S49
Blaikie, P; Cannon, T; Davis, I; and B Wisner. 1994. At Risk: Natural Hazards, People’s
Vulnerability, and Disasters. London: Routledge.
Bodansky, Daniel. 1993. The UN Framework Convention on Climate Change: a commentary. Yale
Journal of International Law 18 : 451.
Bohle, HG; Downing, TE; and MJ Watts. 1994. Climate change and social vulnerability. Global
Environmental Change 4 (1) : 37-48.
Briguglio, L. 1995. Small Island Developing States and Their Economic Vulnerabilities. World
Development 23 (9) : 1615-1632.
Brooks, Nick and W. Neil Adger. 2003. Country level risk measures of climate-related natural
disasters and implications for adaptation to climate change. Tyndall Centre Working Paper 26.
Norwich: Tyndall Centre for Climate Change Research. 26pp.
Burton, Ian; Huq, Saleemul; Lim, Bo; Pilifosova, Olga; and Emma Lisa Schipper. 2002. From
impacts assessment to adaptation priorities: the shaping of adaptation policy. Climate Policy 2 : 145159.
Burton, Ian; Kates, Robert W and Gilbert F White. 1993. The Environment as Hazard. Second
Edition. Oxford: Oxford University Press.
Cannon, Terry. 1994. Vulnerability analysis and the explanation of natural disasters. In Varley, Ann
(ed). Disasters Development and Environment. pp 13-30.
Capobianco, M; DeVriend, JJ; Nicholls, RJ; and MJF Stive. 1999. Coastal area impact and
vulnerability assessment: the point of view of a morphodynamic modeller. Journal of Coastal
Research 15 (3) : 701-716.
Castells, Manuel. 1996. The rise of the network society. Oxford: Blackwell.
Castells, Manuel. 1998. End of Millennium. Oxford: Blackwell.
Castree, N and B Braun (eds) 2000. Social Nature: Theory, Practice and Politics. Oxford: Blackwell.
Chipanshi, AC; Chanda, R; and O Totolo. 2003. Vulnerability assessment of the maize and sorghum
crops to climate change in Botswana. Climatic Change 61 (3) : 339-360.
Clark, William ; Jaeger, J; Corell, R; Kasperson, R; McCarthy, JJ; Cash, D; Cohen, SJ; Desanker, P;
Dickson, NM; Epstein, P; Guston, DH; Hall, JM; Jaeger, C; Janetos, A; Leary, N; Levy, MA; Luers,
A; MacCracken, M; Melillo, J; Moss, R; Nigg, JM; Parry, ML; Parson, EA; Ribot, JC; Schellnhuber,
HJ; Seielstad, GA; Shea, E; Vogel, C; and TJ Wilbanks. 2000. Assessing vulnerability to global
environmental risks. Report of the Workshop on Vulnerability to Global Environmental Change:
Challenges for Research, Assessment and Decision Making. May 22-25, 2000, Airlie House,
Warrenton, Virginia. Cambridge, MA: Belfer Center for Science and International Affairs (BCSIA)
Discussion Paper 2000-12, Environment and Natural Resources Program, Kennedy School of
Government, Harvard University, 2000. Available online at www.ksg.harvard.edu/sust
Comfort, L; Wisner, B; Cutter, S; Pulwarty, R; Hewitt, K; Oliver-Smith, A; Weiner, J; Fordham, M;
Peacock, W; and F. Kringold. 1999. Reframing disaster policy: the global evolution of vulnerable
communities. Environmental Hazards 1 : 39-44.
Crowards, T. 1999. An Economic Vulnerability Index for developing countries, with special
reference to the Caribbean: alternative methodologies and provisional results (draft). Barbados:
Caribbean Development Bank.
Cutter, Susan L. 2003. The vulnerability of science and the science of vulnerability. Annals of the
Association of American Geographers 93 (1) : 1-12.
Cutter, Susan L. (ed) 2001. A research agenda for vulnerability science and environmental hazards.
IHDP Update. Newsletter of the International Human Dimensions Programme on Global
Environmental Change 2 (1) : 8-9.
Cutter, Susan L. 1996a. Vulnerability to Environmental Hazards. Progress in Human Geography 20
: 529-539.
Cutter, Susan L. 1996b. Societal responses to environmental hazards. International Social Science
Journal 48 : 525-536.
Cutter, Susan L. 1995. The forgotten casualties: women, children and environmental change. Global
Environmental Change 5 (1) : 181-194.
DEFRA. 2002. Achieving a better quality of life: review of progress towards sustainable
development. Government Annual Report 2001. London: HMSO. 155pp.
Denton, F. 2002. Climate change vulnerability, impacts and adaptation: why does gender matter?
Gender and Development 10 (2) : 10-21.
DETR. 1999. A better quality of life. London: HMSO.
Desanker, Paul and Christopher Magadza. 2001. Africa. In McCarthy, JJ et al. Climate Change
2001: Impacts, Adaptation and Vulnerability (Contribution of Working Group II to the Third
Assessment Report of the Intergovernmental Panel on Climate Change). p 487-531.
Dessai, S. 2003. The Special Climate Change Fund: origins and prioritization assessment. Climate Policy,
3, 295-302.
Dessai, S and EL Schipper (2003). The Marrakesh Accords to the Kyoto Protocol: Analysis and
Future Prospects. Global Environmental Change (in press).
Dixon, RK; Perry, JA; Vanderklein, EL and FH Hiol. 1996. Vulnerability of forest resources to
global climate change: case study of Cameroon and Ghana. Climate Research 6 (2) : 127-133.
Dow, K. 1992. Exploring differences in our common future(s): the meaning of vulnerability to
global environmental change. Geoforum 23 : 417-436.
Downing, TE. (ed). 1996. Climate change and world food security. Heidelberg: Springer.
Downing, TE; Ringius, L; Hulme, M and D Waughray. 1997. Adapting to climate change in Africa.
Mitigation and Adaptation Strategies for Global Change 2 : 19-44.
Easter, Christopher. 1999. Small states development: a Commonwealth Vulnerability Index. The
Round Table 351 : 403-422.
Enarson, F. 2000. Gender issues in natural disasters: talking points and research needs, ILO
InFocus Programme on Crisis Response and Reconstruction Workshop, Geneva, 3-5 May 2000.
GEF (Global Environment Facility). 2000. Review of climate change enabling activities: evaluation
summary. Washington DC: Global Environment Facility. Report no 2-00, November 2000.
Giupponi, C and M Schechter (eds). 2002. Climate Change and the Mediterranean: Socioeconomics of Impacts, Vulnerability and Adaptation. Cheltenham: Edward Elgar.
Hammond, A; Adrianse, A; Rodenburg, E; Bryant, D. and R. Woodward. 1995. Environmental
indicators: a systematic approach to measuring and reporting on environmental policy performance
in the context of sustainable development. Washington DC: World Resources Institute.
Handmer, JW; Dovers S; and TE Downing. 1999. Societal vulnerability to climate change and
variability. Mitigation and Adaptation Strategies for Global Change 4 : 267-281.
Hartig, EK; Grozev, O and C Rosenzweig. 1997. Climate change, agriculture and wetlands in
Eastern Europe: vulnerability, adaptation and policy. Climatic Change 36 (1-2) : 107-121.
Hewitt, K. 1997. Regions of Risk: A Geographical Introduction to Disasters. Essex: Longman.
Hicks, N. and P. Streeten. 1979. Indicators of development. World Development 7 : 567-580.
Hirst, P. and G. Thompson. 1996. Globalisation in question. Cambridge: Polity Press.
Hodess, Robin (ed). 2003. Global Corruption Report 2003. Profile Books, 329pp. Available online
at www.transparencyinternational.org [accessed 15th May 2003]
Houghton, J; Ding, Y; Griggs, D; Noguer, M; van der Linden, P; Da, X; Maskell, K. and C. Johnson
(eds). 2001. Climate Change 2001: The Scientific Basis. Cambridge: Cambridge University Press.
Hulme, M. 2001. Climate perspectives on Sahelian desiccation: 1973-1998. Global Environmental
Change 11 : 19-29.
Hulme, M; Mitchell, J; Ingram, W; Lowe, J; Johns, T; New, M; and D. Viner. 1999. Climate change
scenarios for global impacts studies. Global Environmental Change 9 : S3-S19.
Hurd, B; Smith, J; Jones, R and K Spiecker. 1999. Water and climate change: a national assessment
of regional vulnerability. Boulder: Stratus Consulting Inc.
ITU. 2002. World Telecommunications Indicators. Geneva: International Telecommunications
Union.
Jollands, Nigel and Murray Paterson. 2003. The holy grail of sustainable development indicators: an
approach to aggregating indicators with applications. Paper presented to the USSEE conference, 22
May 2003.
Jones, R and R Boer. 2003. Assessing current climate risks. Adaptation Policy Framework: A Guide
for Policies to Facilitate Adaptation to Climate Change, UNDP in review. Available online at
http://www.undp.org/cc/apf-outline.htm [accessed 25th June 2003]
Kaly, U and C Pratt. 2000. Environmental Vulnerability Index: Development and provisional indices
and profiles for Fiji, Samoa, Tuvalu and Vanuatu. Phase II report for NZODA. SOPAC Technical
Report 306. 89pp.
Kaly, U; Briguglio, L; McLeod, H; Schmall, S; Pratt, C; and R Pal. 1999a. Environmental
Vulnerability Index (EVI) to summarise national environmental vulnerability profiles. SOPAC
Technical Report 275. 66pp.
Kaly, U; Briguglio, L; McLeod, H; Schmall, S; Pratt, C; and R Pal. 1999b. Report on the
Environmental Vulnerability Index (EVI) Think Tank, 7-10 September 1999, Pacific Harbour, Fiji.
SOPAC Technical Report 299. 78pp.
Karim, Z; Hussain, SG and M Ahmed. 1996. Assessing impacts of climate variations on foodgrain
production in Bangladesh. Water Air and Soil Pollution 92 (1-2) : 53-62.
Kasperson, RE; Kasperson, JEX and BL Turner. 1995. Regions at Risk: comparisons of threatened
environments. Tokyo: United Nations University Press.
Kates, Robert W. 2000. Cautionary tales: adaptation and the global poor. Climatic Change 45 : 5-17.
Klein, Richard JT 2002. Adaptation to climate variability and change: what is optimal and
appropriate? In C Giupponi and M Schechter (eds). Climate Change and the Mediterranean: Socioeconomics of Impacts, Vulnerability and Adaptation. Cheltenham: Edward Elgar.
Klein, Richard JT and RJ Nicholls. 1999. Assessment of coastal vulnerability to climate change.
Ambio 28 (2) : 182-187.
Lal, M; Singh, KK; Rathore, LS; Srinivasan, G and SA Saseendran. 1997. Vulnerability of rice and
wheat yields in NW India to future changes in climate. Agricultural and Forest Meteorology 89 (2) :
101-114.
Lee, R and J Wills (eds). 1997. Geographies of Economies. London: Arnold.
Leichenko, R and K O’Brien. 2002. The dynamics of rural vulnerability to global change: the case of
Southern Africa. Mitigation and Adaptation Strategies for Global Change 7 : 1-18.
Levy, PE; Cannell, MGR; and AD Friend. 2004. Modelling the impact of future changes in climate,
CO2 concentration and landuse on natural ecosystems and the terrestrial carbon sink. Global
Environmental Change 14 (1) : 21-30.
Liverman, DM. 1994. Vulnerability to global environmental change. In Cutter, SL. (ed).
Environmental Risks and Hazards. New Jersey: Prentice Hall.
Lohani, B and G Todino. 1984. Water quality index for the Chao Phraya River. Journal of
Environmental Engineering 110 : 1163-1176.
Magrin, GO; Travasso, MI; Diaz, RA and RO Rodriguez. 1997. Vulnerability of the agricultural
systems of Argentina to climate change. Climate Research 9 (1-2) : 31-36.
Martens, P; Kovats, RS; Nijhof, S; de Vries, P; Livermore, MTJ; Bradley, DJ; Cox, J; and AJ
McMichael. 1999. Climate change and future populations at risk of malaria. Global Environmental
Change 9 : S89-S107.
McCarthy, JJ; Canziani, OF; Leary NA; Dokken DJ; and KS White. 2001 Climate Change 2001:
Impacts, Adaptation and Vulnerability (Contribution of Working Group II to the Third Assessment
Report of the Intergovernmental Panel on Climate Change). Cambridge: Cambridge University
Press.
Mendoza, VM; Villanueva, EE and J Adem. 1997. Vulnerability of basins and watersheds in Mexico
to global climate change. Climate Research 9 (1-2) : 139-145.
Meze-Hausken, Elisabeth. 2000. Migration caused by climate change: how vulnerable are people in
dryland areas? Mitigation and Adaptation Strategies for Global Environmental Change 5 : 379-406.
Mitchell, James K. 2001. What’s in a name?: issues of terminology and language in hazards
research. Editorial. Environmental Hazards 2 : 87-88.
Mortimore, M. 1998. Roots in the African dust: sustaining the sub-Saharan drylands. Cambridge:
Cambridge University Press.
Moser, CON. 1996. Confronting crisis: a comparative study of urban household responses to poverty
and vulnerability in four poor urban communities. Environmental Sustainable Studies and
Monograph Series 8. Washington DC: World Bank.
Moss, RH; Brenkert, AL and EL Malone. 2001. Vulnerability to climate change: a quantitative
approach. Report prepared by the Pacific Northwest National Laboratory, operated by the Battelle
Memorial Institute, for the US Department of Energy. 70pp.
Niemeijer, David. 2002. Developing indicators for environmental policy: data-driven and theorydriven approaches examined by data. Environmental Science and Policy 5 (2) : 91-103.
Nicholls, Robert J. 2004. Coastal flooding and wetland loss in the 21st century: changes under the SRES
climate and socio economic scenarios. Global Environmental Change 14 (1) : 69-86.
Nicholls, Robert J; Hoozemans, Frank MJ; and Marcel Marchand. 1999. Increasing flood risk and
wetland losses due to global sea-level rise: regional and global analyses. Global Environmental
Change 9 : S69-S87.
O’Brien, K and R Leichenko. 2000. Double exposure: assessing the impacts of climate change
within the context of economic globalization. Global Environmental Change 10 : 221-232.
O’Brien, P and D Mileti. 1992.Citizen participation in emergency response following the Loma
Prieta earthquake. International Journal of Mass Emergencies and Disasters 10 : 71-89.
Olivio, MD. 1997. Assessment of the vulnerability of Venzuela to sea-level rise. Climate Research 9
(1-2) : 57-65.
Parry, ML; Rosenzweig, C; Iglesias, A; Livermore, M and G Fischer. 2004. Effects of climate change on
global food production under SRES emissions and socio-economic scenarios. Global Environmental
Change 14 (1) : 53-67.
Parry, M; Arnell, N; McMichael, T; Nicholls, R; Martens, P; Kovats, S; Livermore, M; Rosenzweig,
C; Iglesias, A and G. Fischer. 2001. Millions at risk: Defining critical climate change threats and
targets. Global Environmental Change 11 (3) : 1-3.
Parry, Martin; Rosenzweig, Cynthia; Iglesias, Ana; Fischer, Guenther and Matthew Livermore.
1999. Climate change and world food security: a new assessment. Global Environmental Change 9 :
S51-S67.
Parry, M and T Carter. 1998. Climate Impact and Adaptation Assessment. London: Earthscan.
Pelling, Mark (ed) 2003. Natural disasters in a globalising world. London: Routledge.
Pelling, Mark. 2000. Natural Disasters? In Castree and Braun (eds). Social Nature: Theory, Practice
and Politics. Oxford: Blackwell.
Pelling, M and J Uitto. 2001. Small island developing states: natural disaster vulnerability and global
change. Environmental Hazards 3 : 49-62
Platt, R. 1995. Lifelines: an emergency management priority for the United States in the 1990s.
Disasters 15 : 172-176.
Putnam, D. 1993. Making democracy work: civic traditions in modern Italy. Princeton: Princeton
University Press.
Robbins, Paul. 2000. The rotten institution: corruption in natural resource management. Political
Geography 19 : 423-443.
Sands, Philippe. 1992. The United Nations Framework Convention on Climate Change. Review of
European Community and International Environmental Law 1 (3) : 270.
Scott, JC 1985. Weapons of the weak: everyday forms of peasant resistance. New Haven: Yale
University Press.
Sen, Amartya. 1981. Poverty and famines: an essay on entitlement and deprivation. Oxford:
Clarendon Press.
Smith, K. 2001. Environmental Hazards: assessing risk and reducing disaster. (3rd edition). London:
Routledge.
Smith, JB and SS Lenhert. 1996. Climate change adaptation policy options. Climate Research 6 :
193-201.
Sokona, Youba and Fatma Denton. 2001. Climate change impacts: can Africa cope with the
challenges? Climate Policy 1 : 117-123.
Stonich, S. 2000. The human dimensions of climate change: the political ecology of vulnerability.
Department of Anthropology and Environmental Studies, University of California.
UN. 2001. World Population Prospects 1950-2020: 2000 review. New York: UN Department of
Economic and Social Affairs.
UN. 2002. World Urbanisation Prospects: the 2001 Revisions. New York: UN Department of
Economic and Social Affairs.
UNAIDS and WHO. 2002. Report on the global HIV/AIDS epidemic 2002. Geneva: UNAIDS.
UNDP. 2002. Human Development Report 2002. New York: United Nations Development
Programme. Available online at http://hdr.undp.org/reports/global/2002/en [accessed 20 May 2003]
UNEP. 2001. Vulnerability Indices: Climate Change Impacts and Adaptation. UNEP Policy Series 3.
Nairobi: UNEP. 91pp.
van Liesert, M; Kovats, RS; Livermore, MTJ; and P Maartens. 2004. Climate change and malaria:
analysis of the SRES climate and socio-economic scenarios. Global Environmental Change 14 (1) : 8799.
Varley, Ann. (ed). 1994. Disasters Development and Environment. Chichester: John Wiley.
Verheyen, Roda. 2002. Adaptation to the impacts of anthropogenic climate change – the
international legal framework. Review of European Community and International Environmental
Law 11 (2) : 129-143.
Watts, MJ and HG Bohle. 1993. The space of vulnerability: the causal structure of hunger and
famine. Progress in Human Geography 17 : 43-67.
Wilbanks, T and RW Kates. 1999. Global change in local places: how scale matters. Climatic
Change 43 : 601-628.
Wisner, Ben. 1998. World views, belief systems, and disasters: implications for preparedness,
mitigation
and
recovery.
Available
online
at
http://online.northumbria.ac.uk/geography_research/radix/resources/wordview98.doc [accessed 15
June 2003]
World Bank. 2002. World Bank Development Indicators. Washington DC: World Bank.
World Bank. 2001. World Bank Development Indicators CD ROM. Washington DC: World Bank.
WEF (World Economic Forum). 2000. Pilot Environmental Sustainability Index: An initiative of the
Global Leaders for Tomorrow Environment Task Force, World Economic Forum. Davos, 39pp.
Yamin, Farhana. 1998. The Clean Development Mechanism and Adaptation. Paper for the UNFCCC
Secretariat Capacity Building Workshop for Project-based mechanisms, FIELD Working Paper no
19. London: Foundation for International Environmental Law and Development.
Appendix A – Summary of national level vulnerability indices
Name of index
and author
Scale and
area focus
Variables included
Indicators chosen
Briguglio
(1995)
Applies to
SIDS
Exposure to foreign
economic
conditions
remoteness and insularity
Ratio of exports and imports to GDP
disaster proneness
UNDRO index of disaster proneness
Small island
developing
state
vulnerability
index
Crowards
(1999)
Applies to
Caribbean
Economic
Vulnerability
Index for
developing
countries with
special
reference to
the Caribbean
Easter (1999)
Commonwealt
h Vulnerability
Index
Small
states in
the
Common
wealth
Ratio of transport and export costs to exports
proceeds
Peripherality/accessibility
Freight and insurance costs for imports as a % of
total import costs
Export concentration
% of total export receipts accounted for by the major
export and the top 3 exports (goods and services),
combined with total export earnings as a % of GDP
Convergence of export
Amount of export receipts deriving from the single
destination
most important destination and the 3 most important
destinations (goods and services), combined with
total export earnings as a % of GDP
Dependence upon
Net energy imports as a % of total energy
imported energy
consumption
Reliance upon external
Measured as a combination of annual disbursement
finance/capital
of concessionary overseas development assistance
and annual foreign direct investment, both as a
proportion of annual gross fixed capital formation
Chose the 3 most significant variables using income volatility as a proxy for
vulnerability) from 30 on elements of remoteness and susceptibility to natural
disasters
Lack of diversification
UNCTAD’s diversification index
(vulnerability)
Export dependence
The proportion of exports in GDP
(vulnerability)
Impact of natural disasters
(vulnerability)
Proportion of population affected by natural disasters
Resilience
Average GDP
Means of mathematical
transformation (at
indicator/sub-index level)
Standardisation formula:
Vij= (Xij-MinXi)
(Max Xi-MinXi)
where Vij= degree of
vulnerability arising from the
ith variable for country j
Xij=value of the ith variable in
the index for country j
Max and Min represent the
extremes of the data range
Standardisation/normalisation
Modified normalisation with
maximum and minimum
deciles
Method(s) of aggregation to form
composite index
Equal weighting
Nonequal weighting (50% to
economic exposure, 40% to transport
index, 10% to disaster proneness
index)
Averaging across the selected series
for each country with the variables
grouped into 4 main parameters
(peripherality and energy, export
concentration, export destination and
external finance) varying the
transformed components
Borda rule; use rank of component
variables to assign aggregate rank
Equal weighting
Condensed decile normalisation
Principle components analysis
Index comprises two elements:
vulnerability impact index and
resilience index. For the VII the
three indicators were weighted using
PCA; and then in combining with
resilience PCA was also used
Kaly et al (1999)
for the South
Pacific Applied
Geoscience
Commission
SIDS with
emphasis
on the
South
Pacific
The VulnerabilityResilience
Indicator
Prototype Model
(VRIP)
Three respective sub-indices (risk exposure,
intrinsic resilience and environmental
degradation) created on the basis of 57
indicators
Normalisation onto a
categorical scale so that each
indicator is ascribed a value of
0-7.
Weighted average of scores allocated
to the 3 sub-indices.
Cereals production/area
Animal protein consumption/capita
% land managed
fertiliser use
Population at flood risk from sea level rise
Population without access to clean water
GDP per capita
Gini index
Completed fertility
Life expectancy
Dependency ratio
Literacy
Renewable supply and inflow
Water use
Population density
SO2 emissions per area
Percentage of unmanaged land
Proxy indicators were indexed
(scaled) firstly against the
1990 world data (set to 100)
and
secondly against the 1990
USA data (set to 100).
Following side-by-side and
statistical analyses of results
for each method, world
baseline values were chosen to
facilitate country-to-country
comparisons.
Hierarchical aggregation of
geometric means, such that the
geometric means of indexed proxies
determine the values of
sectoral indicators, and the geometric
means of sectoral indicators become
indicators for sensitivity to climate
impact or coping and adaptive
capacity.
Ecosystem integrity (the
health or condition of the
environment as a result of past
impacts)
Environmental
Vulnerability
Index
Moss et al (2001),
Pacific Northwest
National
Laboratory
operated by
Battelle
Risks to the environment
(natural and anthropogenic)
The ability of the environment
to cope with risk
Selection
of
developed
and
developin
g
countries
Food sensitivity
Ecosystems sensitivity
Settlements sensitivity
Economic coping capacity
Human health sensitivity
Human and civic resources
Water resource sensitivity
Environmental coping
capacity
Appendix B: Calculating the Social Vulnerability Index: a worked example (Botswana)
The social vulnerability index is a composite index created by weighted aggregations of 5
sub-indices, some of which are themselves composites of indicators.
Economic well-being and stability
This comprises 2 indicators: (i) standard of
living/poverty and (ii) % change in urban
population
(i)
(ii)
36.20- (-2.1) =
44.1-10.6 = 0.448
0.880
86-10.6 41.4-(-2.1)
80:20 weighting so (0.448x4)+0.880 =
0.535
5
Demographic structure
This comprises 2 indicators: (i)
dependency ratio and (ii) proportion of
working population with HIV/AIDS
(i)
(ii)
44.90-31.8 = 0.652
38.8-0.08 = 1
51.90-31.8
38.8-0.08
equal weighting so 0.0652+1 = 0.826
2
Institutional stability and strength of public infrastructure
This comprises 2 indicators in version A: (i) health expenditure and (ii) number of telephones;
and an additional third (iii) corruption in version B.
(i)
(ii)
(iii)
2.5-0.6 = 0.396 (1-0.396=0.604) 93-1 = 0.393 (1-0.393=0.607) 6.40-1.6 = 1 (1-1=0)
5.4-0.6
235-1
6.40-1.6
(All need to be transformed 1-result so that the higher numbers refer to conditions of higher
vulnerability)
Index A is weighted 80:20
Index B is weighted
60:20:20
(0.604x4)+0.606 = 0.605
((0.604x3)+0.607+0)
= 0.484
5
5
Global interconnectivity
This comprises 1 indicator for trade
balance
531,288 – (-7,572,300) = 0.743 (10.743=0.257) 3,340,613-(-7,572,300)
Insert: example of standardisation process for
global interconnectivity (after transformation)
-$7,572,300-8
1.0
0.8
-$4,000,000
0.6
Natural resource dependence
This comprises 1 indicator –
rural population
50.2 - 12.9 = 0.460
93.9 – 12.9
0.4
0
Botswana - $531,248
$3,340,613
0.2
0.0
$4,000,000
Calculating index A: SVI = Σ (Ii*Wi)(Iii*Wii)(Iiii*Wiii)(Iiv*Wiv)(Iv*Wv)
=(0.535*0.20)+(0.826*0.20)+(0.605*0.40)+(0.257*0.10)+(0.460*0.10) = 0.586
Calculating index B :
=(0.535*0.20)+(0.826*0.20)+(0.484*0.40)+(0.257*0.10)+(0.460*0.10) = 0.537
Appendix C – Full results table (vulnerability profiles)
14
0.637
4
% with
HIV/AIDS
0.537
rank
Zimbabwe
6
actual
score
0.584
0.618
0.581
0.634
rank
47
17
27
24
5
3
11
40
actual
score
0.360
0.612
0.584
0.586
0.658
0.703
0.640
0.543
rank
rank
Algeria
Angola*
Benin
Botswana*
Burkina Faso*
Burundi
Cameroon
Cape Verde*
Cent. Afric.
Rep*
Chad
Comoros*
Dem Rep.
Congo*
Djibouti
Egypt
Equat. Guinea*
Eritrea
Ethiopia*
Gabon*
Ghana
Guinea
Guinea Bissau
Ivory Coast
Kenya
Lesotho
Libya*
Madagascar
Malawi
Mali*
Mauritania
Mauritius
Morocco*
Mozambique*
Namibia*
Niger
Nigeria
Rep. of Congo*
Rwanda
Senegal
Sierra Leone
South Africa*
Sudan*
Swaziland
Tanzania
The Gambia
Togo*
Tunisia
Uganda
Zambia
actual
score
actual
score
COUNTRY
Demographic structure subindex
(20%)
dependent
population
Economic wellbeing and stability
subindex (20%)
growth in
urban
popns
Vulnerability
Index B
standard
of
living/pove
rty
Vulnerability
Index A
0.214
0.444
0.341
0.535
0.415
0.308
0.423
0.556
46
25
40.5
10
32
43
30
9
0.159
0.448
0.297
0.448
0.448
0.340
0.390
0.448
0.434
0.425
0.517
0.880
0.283
0.182
0.554
0.986
0.177
0.548
0.476
0.826
0.583
0.571
0.525
0.444
45
16
30
2
10
12
21
35
0.353
0.955
0.861
0.652
1.000
0.930
0.746
0.602
0.001
0.140
0.091
1.000
0.166
0.212
0.303
0.287
27
16
29
0.403
0.614
0.424
34
6
29
0.448
0.708
0.448
0.221
0.237
0.324
0.544
0.488
0.487
18
26.5
28
0.756
0.886
0.687
0.331
0.091
0.287
0.591
0.303
0.493
0.561
0.601
0.655
0.547
0.624
0.562
0.591
0.584
0.578
0.649
0.405
0.691
0.591
0.585
0.654
0.329
0.550
0.557
0.522
0.725
0.621
0.576
0.627
0.481
0.705
0.390
0.556
0.599
0.646
0.567
0.633
0.368
0.657
0.597
22
49
42
34
18
7
38
14
33
22
27
30
9
44
4
22
25
8
48
37
35
41
1
15
31
13
43
2
45
36
19
10
32
12
46
6
20
38
24
47
19
16
35
8
44
37
17
40.5
27
14
15
2
13
28
4
49
21
18
31
7
39
11.5
20
42
3
33
22.5
36
26
5
22.5
48
11.5
1
0.448
0.458
0.163
0.448
0.562
0.448
0.448
0.276
0.390
0.505
0.347
0.416
0.512
0.448
0.788
0.576
0.484
0.615
0.000
0.448
0.448
0.448
0.695
0.312
0.448
0.538
0.302
0.761
0.448
0.448
0.390
0.411
0.708
0.448
0.046
0.589
1.000
0.067
0.395
0.030
0.533
0.186
0.186
1.000
0.186
0.306
0.405
0.313
0.520
0.444
0.662
0.352
0.209
0.370
0.908
0.000
0.455
0.586
0.285
0.278
0.524
0.747
0.099
0.352
0.398
0.253
0.444
0.333
0.559
0.363
0.441
0.407
0.184
0.159
0.557
0.507
0.192
0.436
0.408
0.490
0.572
0.731
0.519
0.416
0.457
0.553
0.690
0.137
0.398
0.626
0.475
0.529
0.000
0.174
0.547
0.680
0.643
0.479
0.534
0.489
0.379
0.470
0.403
0.324
0.761
0.488
0.309
0.465
0.095
0.559
0.715
14
23
44
36
38
24
11
4
22
37
34
15
6
47
40
9
31
20
49
46
17
7
8
29
19
25
41
32
39
42
3
26.5
43
33
48
13
5
0.990
0.726
0.383
0.786
0.746
0.816
0.856
0.612
0.751
0.761
0.667
0.721
0.582
0.274
0.791
0.866
0.910
0.771
0.000
0.348
0.761
0.781
1.000
0.811
0.886
0.751
0.746
0.761
0.289
0.582
0.662
0.776
0.577
0.776
0.189
0.990
0.976
0.124
0.287
0.000
0.085
0.070
0.163
0.287
0.850
0.287
0.071
0.247
0.386
0.799
0.000
0.005
0.385
0.041
0.287
0.001
0.000
0.334
0.579
0.287
0.148
0.183
0.227
0.011
0.179
0.517
0.065
0.862
0.200
0.041
0.153
0.000
0.127
0.554
0.545
39
45
0.198
0.409
0.847
1
0.826
0.869
0.487
18
0.635
5
0.604
10
0.576
0.603
12
11
0.697
0.606
1
9
0.300
0.525
21
15
0.498
16
0.624
8
0.489
17
0.381
19
0.640
3
0.341
0.670
0.631
20
2
7
0.372
0.445
0.136
0.465
0.487
0.396
0.559
0.258
0.373
0.485
0.341
0.437
0.498
0.491
0.701
0.502
0.433
0.674
0.000
0.450
0.476
0.416
0.612
0.354
0.508
0.451
0.312
0.689
0.409
0.447
0.379
0.441
0.639
0.447
0.119
0.508
0.832
0.567
13
0.240
43
Institutional strength and public infrastructure subindex
(40%)
Global
interconnectivity
subindex (10%)
3
corruption
health
expenditur
e
rank A
rank B
19
0.583
0.700
0.792
0.604
0.813
1.000
0.917
0.750
0.979
0.000
0.875
rank
0.921
8
% rural
population
0.585
0.484
42
20
14
43
13
1
4
33
rank
Zimbabwe
0.812
trade
balance
0.619
0.757
0.827
0.605
0.847
0.998
0.929
0.693
Telephone
s
Algeria
Angola*
Benin
Botswana*
Burkina Faso*
Burundi
Cameroon
Cape Verde*
Cent. Afric.
Rep*
Chad
Comoros*
Dem Rep.
Congo*
Djibouti
Egypt
Equat. Guinea*
Eritrea
Ethiopia*
Gabon*
Ghana
Guinea
Guinea Bissau
Ivory Coast
Kenya
Lesotho
Libya*
Madagascar
Malawi
Mali*
Mauritania
Mauritius
Morocco*
Mozambique*
Namibia*
Niger
Nigeria
Rep. of Congo*
Rwanda
Senegal
Sierra Leone
South Africa*
Sudan*
Swaziland
Tanzania
The Gambia
Togo*
Tunisia
Uganda
Zambia
actual
score B
actual
score A
COUNTRY
Natural
resource
dependence
subindex
(10%)
0.761
0.983
0.970
0.607
0.987
0.991
0.979
0.466
0.000
0.451
0.329
0.257
0.345
0.312
0.310
0.325
49
2
20
43
14
35.5
38.5
24
0.340
0.660
0.563
0.460
0.853
0.968
0.481
0.327
43
23
32
39
5
2
38
44
0.765
0.717
0.752
18
30
24.5
0.708
0.646
0.700
0.991
1.000
0.962
0.314
0.323
0.309
33
27
40
0.572
0.785
0.672
30
9
21
0.755
0.188
0.687
0.750
0.754
0.881
0.724
0.807
0.711
0.753
0.885
0.692
0.752
0.669
0.915
0.631
0.748
0.862
0.600
0.858
0.631
0.497
0.899
0.964
0.762
0.766
0.649
0.947
0.453
0.751
0.657
0.880
0.695
0.876
0.657
0.782
0.494
0.591
0.815
0.577
21
49
35
27
22
8
29
15
31
23
7
34
24.5
36
5
40.5
28
11
44
12
40.5
46
6
2
19
17
39
3
48
26
37.5
9
32
10
37.5
16
47
16
7
17
0.700
0.000
0.700
0.700
0.700
0.854
0.688
0.771
0.646
0.700
0.875
0.625
0.700
0.700
0.896
0.542
0.688
0.833
0.750
0.875
0.572
0.438
0.875
0.958
0.708
0.708
0.583
0.938
0.438
0.700
0.604
0.854
0.646
0.854
0.667
0.729
0.375
0.333
0.896
0.792
0.974
0.940
0.637
0.949
0.970
0.987
0.868
0.953
0.970
0.966
0.927
0.962
0.962
0.543
0.991
0.987
0.991
0.974
0.000
0.791
0.987
0.735
0.995
0.987
0.974
0.996
0.910
0.987
0.517
0.953
0.868
0.983
0.893
0.966
0.620
0.991
0.970
0.325
0.325
1.000
0.310
0.355
0.394
0.229
0.421
0.318
0.308
0.195
0.352
0.362
0.119
0.333
0.343
0.333
0.311
0.319
0.429
0.402
0.336
0.319
0.142
0.317
0.324
0.334
0.313
0.002
0.372
0.326
0.402
0.312
0.326
0.348
0.408
0.318
24
24
1
38.5
11
8
44
4
30.5
41
45
12
10
47
18.5
15
18.5
37
28.5
3
6.5
16
28.5
46
32
26
17
34
48
9
21.5
6.5
35.5
21.5
13
5
30.5
0.705
0.049
0.520
0.495
0.848
0.863
0.084
0.607
0.679
0.788
0.510
0.678
0.740
0.000
0.718
0.784
0.712
0.379
0.568
0.393
0.595
0.700
0.827
0.543
0.314
1.000
0.499
0.632
0.456
0.641
0.753
0.684
0.683
0.671
0.276
0.904
0.587
15
48
34
37
6
4
47
27
19
8
45
20
12
49
13
10
14
42
31
41
28
16
7
33
45
1
36
26
40
25
11
17
18
22
46
3
29
0.640
45
15
0.500
0.771
0.927
0.287
42
0.648
24
0.672
0.831
0.757
0.865
0.755
0.932
0.668
0.529
0.796
0.439
0.972
0.670
0.433
0.863
12
6
10
4
11
2
14
18
9
20
1
13
21
5
0.625
0.604
0.521
0.771
0.938
0.979
0.729
0.396
0.563
0.146
1.000
0.688
0.333
0.771
44
Acknowledgements
This research was undertaken for my MRes dissertation. I would like to thank the following persons
for their time and help: Neil Adger, Nick Brooks, Suraje Dessai, Marisa Goulden and Elisabeth
Meze-Hausken. Many members of the Tyndall Centre’s adaptation research theme have also
contributed to the development of ideas presented here, albeit perhaps unknowingly, through general
discussions, particularly at the 2003 Tyndall Assembly.
45
The trans-disciplinary Tyndall Centre for Climate Change Research undertakes integrated research into the
long-term consequences of climate change for society and into the development of sustainable responses
that governments, business-leaders and decision-makers can evaluate and implement. Achieving these
objectives brings together UK climate scientists, social scientists, engineers and economists in a unique
collaborative research effort.
Research at the Tyndall Centre is organised into four research themes that collectively contribute to all
aspects of the climate change issue: Integrating Frameworks; Decarbonising Modern Societies; Adapting to
Climate Change; and Sustaining the Coastal Zone. All thematic fields address a clear problem posed to
society by climate change, and will generate results to guide the strategic development of climate change
mitigation and adaptation policies at local, national and global scales.
The Tyndall Centre is named after the 19th century UK scientist John Tyndall, who was the first to prove the
Earth’s natural greenhouse effect and suggested that slight changes in atmospheric composition could bring
about climate variations. In addition, he was committed to improving the quality of science education and
knowledge.
The Tyndall Centre is a partnership of the following institutions:
University of East Anglia
UMIST
Southampton Oceanography Centre
University of Southampton
University of Cambridge
Centre for Ecology and Hydrology
SPRU – Science and Technology Policy Research (University of Sussex)
Institute for Transport Studies (University of Leeds)
Complex Systems Management Centre (Cranfield University)
Energy Research Unit (CLRC Rutherford Appleton Laboratory)
The Centre is core funded by the following organisations:
Natural Environmental Research Council (NERC)
Economic and Social Research Council (ESRC)
Engineering and Physical Sciences Research Council (EPSRC)
UK Government Department of Trade and Industry (DTI)
For more information, visit the Tyndall Centre Web site (www.tyndall.ac.uk) or contact:
External Communications Manager
Tyndall Centre for Climate Change Research
University of East Anglia, Norwich NR4 7TJ, UK
Phone: +44 (0) 1603 59 3906; Fax: +44 (0) 1603 59 3901
Email: tyndall@uea.ac.uk
Recent Working Papers
Tyndall Working Papers are available online at
http://www.tyndall.ac.uk/publications/working_papers/working_papers.shtml
Mitchell, T. and Hulme, M. (2000). A Country-byCountry Analysis of Past and Future Warming
Rates, Tyndall Centre Working Paper 1.
Hulme, M. (2001). Integrated Assessment
Models, Tyndall Centre Working Paper 2.
Berkhout, F, Hertin, J. and Jordan, A. J. (2001).
Socio-economic futures in climate change
impact assessment: using scenarios as
'learning machines', Tyndall Centre Working
Paper 3.
Barker, T. and Ekins, P. (2001). How High are
the Costs of Kyoto for the US Economy?,
Tyndall Centre Working Paper 4.
Barnett, J. (2001). The issue of 'Adverse Effects
and the Impacts of Response Measures' in the
UNFCCC, Tyndall Centre Working Paper 5.
Goodess, C.M., Hulme, M. and Osborn, T. (2001).
The identification and evaluation of suitable
scenario development methods for the
estimation of future probabilities of extreme
weather events, Tyndall Centre Working Paper 6.
Barnett, J. (2001). Security and Climate
Change, Tyndall Centre Working Paper 7.
Adger, W. N. (2001). Social Capital and Climate
Change, Tyndall Centre Working Paper 8.
Barnett, J. and Adger, W. N. (2001). Climate
Dangers and Atoll Countries, Tyndall Centre
Working Paper 9.
Gough, C., Taylor, I. and Shackley, S. (2001).
Burying Carbon under the Sea: An Initial
Exploration of Public Opinions, Tyndall Centre
Working Paper 10.
Barker, T. (2001). Representing the Integrated
Assessment of Climate Change, Adaptation
and Mitigation, Tyndall Centre Working Paper 11.
Dessai, S., (2001). The climate regime from
The Hague to Marrakech: Saving or sinking
the Kyoto Protocol?, Tyndall Centre Working
Paper 12.
Dewick, P., Green K., Miozzo, M., (2002).
Technological Change, Industry Structure and
the Environment, Tyndall Centre Working Paper
13.
Shackley, S. and Gough, C., (2002). The Use of
Integrated Assessment: An Institutional
Analysis Perspective, Tyndall Centre Working
Paper 14.
Köhler, J.H., (2002). Long run technical change
in an energy-environment-economy (E3)
model for an IA system: A model of
Kondratiev waves, Tyndall Centre Working Paper
15.
Adger, W.N., Huq, S., Brown, K., Conway, D. and
Hulme, M. (2002). Adaptation to climate
change: Setting the Agenda for Development
Policy and Research, Tyndall Centre Working
Paper 16.
Dutton, G., (2002). Hydrogen Energy
Technology, Tyndall Centre Working Paper 17.
Watson, J. (2002). The development of large
technical systems: implications for hydrogen,
Tyndall Centre Working Paper 18.
Pridmore, A. and Bristow, A., (2002). The role of
hydrogen in powering road transport, Tyndall
Centre Working Paper 19.
Turnpenny, J. (2002). Reviewing organisational
use of scenarios: Case study - evaluating UK
energy policy options, Tyndall Centre Working
Paper 20.
Watson, W. J. (2002). Renewables and CHP
Deployment in the UK to 2020, Tyndall Centre
Working Paper 21.
Watson, W.J., Hertin, J., Randall, T., Gough, C.
(2002). Renewable Energy and Combined Heat
and Power Resources in the UK, Tyndall Centre
Working Paper 22.
Paavola, J. and Adger, W.N. (2002). Justice and
adaptation to climate change, Tyndall Centre
Working Paper 23.
Xueguang Wu, Jenkins, N. and Strbac, G. (2002).
Impact of Integrating Renewables and CHP
into the UK Transmission Network, Tyndall
Centre Working Paper 24
Xueguang Wu, Mutale, J., Jenkins, N. and Strbac,
G. (2003). An investigation of Network
Splitting for Fault Level Reduction, Tyndall
Centre Working Paper 25
Brooks, N. and Adger W.N. (2003). Country level
risk measures of climate-related natural
disasters and implications for adaptation to
climate change, Tyndall Centre Working Paper 26
Brooks, N. (2003). Vulnerability, risk and
adaptation: a conceptual framework, Tyndall
Centre Working Paper 38
Tompkins, E.L. and Adger, W.N. (2003). Building
resilience to climate change through adaptive
management of natural resources, Tyndall
Centre Working Paper 27
Tompkins, E.L. and Adger, W.N. (2003).
Defining response capacity to enhance climate
change policy, Tyndall Centre Working Paper 39
Dessai, S., Adger, W.N., Hulme, M., Köhler, J.H.,
Turnpenny, J. and Warren, R. (2003). Defining
and experiencing dangerous climate change,
Tyndall Centre Working Paper 28
Klein, R.J.T., Lisa Schipper, E. and Dessai, S.
(2003), Integrating mitigation and adaptation
into climate and development policy: three
research questions, Tyndall Centre Working Paper
40
Brown, K. and Corbera, E. (2003). A MultiCriteria Assessment Framework for CarbonMitigation Projects: Putting “development” in
the centre of decision-making, Tyndall Centre
Working Paper 29
Hulme, M. (2003). Abrupt climate change: can
society cope?, Tyndall Centre Working Paper 30
Turnpenny, J., Haxeltine A. and O’Riordan, T.
(2003). A scoping study of UK user needs for
managing climate futures. Part 1 of the pilotphase interactive integrated assessment
process (Aurion Project), Tyndall Centre
Working Paper 31
Xueguang Wu, Jenkins, N. and Strbac, G. (2003).
Integrating Renewables and CHP into the UK
Electricity System: Investigation of the impact
of network faults on the stability of large
offshore wind farms, Tyndall Centre Working
Paper 32
Watson, J. (2003), UK Electricity Scenarios for
2050, Tyndall Centre Working Paper 41
Kim, J. A. (2003), Sustainable Development and
the CDM: A South African Case Study, Tyndall
Centre Working Paper 42
Anderson, D. and Winne, S. (2003),
Innovation and Threshold Effects in
Technology Responses to Climate Change,
Tyndall Centre Working Paper 43
Shackley, S., McLachlan, C. and Gough, C. (2004)
The Public Perceptions of Carbon Capture and
Storage, Tyndall Centre Working Paper 44
Purdy, R. and Macrory, R. (2004) Geological
carbon sequestration: critical legal issues,
Tyndall Centre Working Paper 45
Pridmore, A., Bristow, A.L., May, A. D. and Tight,
M.R. (2003). Climate Change, Impacts, Future
Scenarios and the Role of Transport, Tyndall
Centre Working Paper 33
Watson, J., Tetteh, A., Dutton, G., Bristow, A.,
Kelly, C., Page, M. and Pridmore, A., (2004) UK
Hydrogen Futures to 2050, Tyndall Centre
Working Paper 46
Dessai, S., Hulme, M (2003). Does climate policy
need probabilities?, Tyndall Centre Working Paper
34
Berkhout, F., Hertin, J. and Gann, D. M., (2004)
Learning to adapt: Organisational adaptation
to climate change impacts, Tyndall Centre
Working Paper 47
Tompkins, E. L. and Hurlston, L. (2003). Report to
the Cayman Islands’ Government. Adaptation
lessons learned from responding to tropical
cyclones by the Cayman Islands’ Government,
1988 – 2002, Tyndall Centre Working Paper 35
Pan, H. (2004) The evolution of economic
structure under technological development,
Tyndall Centre Working Paper 48
Kröger, K. Fergusson, M. and Skinner, I. (2003).
Critical Issues in Decarbonising Transport: The
Role of Technologies, Tyndall Centre Working
Paper 36
Ingham, A. and Ulph, A. (2003) Uncertainty,
Irreversibility, Precaution and the Social Cost
of Carbon, Tyndall Centre Working Paper 37
Awerbuch, S. (2004) Restructuring our
electricity networks to promote
decarbonisation, Tyndall Centre Working Paper 49
Powell, J.C., Peters, M.D., Ruddell, A. & Halliday, J.
(2004) Fuel Cells for a Sustainable Future?
Tyndall Centre Working Paper 50
Agnolucci, P., Barker, T. & Ekins, P. (2004)
Hysteresis and energy demand: the
Announcement Effects and the effects of
the UK climate change levy Tyndall Centre
Working Paper 51
Agnolucci, P. (2004) Ex post evaluations of
CO2 –Based Taxes: A Survey Tyndall Centre
Working Paper 52
Agnolucci, P. & Ekins, P. (2004) The
Announcement Effect and environmental
taxation Tyndall Centre Working Paper 53
Turnpenny, J., Carney, S., Haxeltine, A., &
O’Riordan, T. (2004) Developing regional
and local scenarios for climate change
mitigation and adaptation, Part 1: A
framing of the East of England Tyndall
Centre Working Paper 54
Mitchell, T.D. Carter, T.R., Jones, .P.D, Hulme,
M. and New, M. (2004) A comprehensive set
of high-resolution grids of monthly
climate for Europe and the globe: the
observed record (1901-2000) and 16
scenarios (2001-2100), Tyndall Centre
Working Paper 55
Vincent, K. (2004) Creating an index of
social vulnerability to climate change for
Africa, Tyndall Centre Working Paper 56