Measures of vulnerability (V) typically include three
components: exposure to climate change (E), sensitivity to its effects
(S) and adaptive capacity for coping with the effects (AC), such that:
V = f (E, S, AC)
(1)
It
is important to note that in the literature there are different ways of
thinking about vulnerability - and that has implications for
adaptation. Adaptation to outcome vulnerability involves reducing
exposure to climate change either through efforts to mitigate
anthropogenic climate change (Burton et al., 2002) or policies that
limit negative outcomes; for example, through the development of early
warning systems, risk-sharing mechanisms (e.g. insurance), and
adjustments in design standards and capital investments in
infrastructure. Social vulnerability assumes that political,
institutional, economic and social structures interact dynamically to
influence exposure. From this perspective, adaptation involves
“altering the context in which climate change occurs, so that
individuals and groups can better respond to changing
conditions” (O’Brien et al., 2007, p 76). Here we
focus more on social vulnerability, as we believe outcome vulnerability
is the same as sectoral impacts (see above). However the distinction is
important - and when appraising any literature referring to
vulnerability it is critical to understand which definition of
vulnerability the authors are using.
Box 3 6: Overview of
Vulnerability Indication
Theoretical assumption
Individual
or social capacities and external climate drivers are
responsible
for impacts, but their interactions cannot be reliably simulated using
computational models.
Question addressed
Which combinations of variables
give an indication of how
climate change may impact the
study unit?
Data
requirements
Data
on potential indicating variables.
Typical result
A function that maps the
current state of the study unit to a
measure of possible
future impacts.
Generic steps
- Select
potential indicating variables based on literature
- Aggregate
indicating variables based on theoretical and normative arguments
|
Indicators
and indices are a popular option for prioritising adaptation
interventions (Klein, 2009). Some of the indicators (primarily of
exposure and sensitivity) are from the biophysical realm; others
(mainly describing adaptive capacity) are drawn from socioeconomic
statistical sources. Indicators can then be combined to form indices:
either as a composite (where the make-up of the component indicators is
apparent), or an aggregate (where it is not) (Jollands and Paterson,
2003). In particular many indices have focused on adaptive capacity
(social vulnerability), for use in conjunction with exposure and
sensitivity (biophysical vulnerability) data.
There
have been several attempts at developing national level indicators and
indices for aspects of adaptive capacity (social vulnerability), each
varying in the nature of vulnerability addressed, the hazard involved,
and the geographical region. 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. Initial development of indices
took place with reference to the small island developing state context
(e.g. Briguglio, 1995; Crowards, 1999; Easter, 1999; Kaly et al, 1999a,
Turvey, 2007). An index of social vulnerability to climate
change-induced changes in water availability has been created for
Africa (Vincent, 2004). Assessments of vulnerability to climate change
have also taken place at sub-national level. For instance, Figure 3-4
is a depiction of district-level vulnerability to climate change of the
agricultural sector in India, based on a set of composite indicators
(O’Brien et al., 2004).
Figure 3-4: Vulnerability of the agricultural sector to climate change in India by
district.
Vulnerability is computed as a composite of indices of adaptive capacity and climate
sensitivity
under
exposure to climate change (O’Brien et al., 2004).
Whilst
many indices have focused on specific regions, others have taken more
global approaches to assessing vulnerability and resilience, explicitly
in regard to climate change (UNEP, 2001; Moss et al, 2001). Within the
last year various explicit indices have been released, including the
Global Adaptation Index
1,
World Risk Index
2,
and Climate Vulnerability Monitor
3.
Clearly there is a policy appeal for such global indices, particularly
given the need to transparently allocate the growing pool of adaptation
funding. However, a recent study showed the sector-specific or
hazard-specific criteria give a more robust assessment of
vulnerability, since the patterns of vulnerability factors for
different sectors vary geographically (Fuessel, 2010).
The
methodological debates on the use and construction of indicators have
grown, commensurate with the range of indicators and indices (for a
review, see Fuessel, 2009). One of the most fundamental distinctions is
between an inductive (data-driven) and a deductive (theory-driven)
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. The alternative is the theory-driven approach, whereby
use is made of existing theoretical insights into the nature and causes
of vulnerability to select variables for inclusion (Adger, 2006),
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.
Although a number of
indicators and indices have been devised for assessing social
vulnerability to climate change, there is no “one size fits
all” blueprint that can be used regardless of the context.
Indicators are context specific and typically cannot be transferred to
different scales of analysis. Whilst the driving forces of social
vulnerability might be similar, the appropriate indicator to capture
that at a national level will likely be different from that at a
sub-national level (Vincent, 2007; Eriksen and Kelly, 2007). A recent
paper reviewed the use of indices in a variety of circumstances,
concluding that they are most appropriate for identifying vulnerable
populations at the sub-national level (Hinkel, 2011). Various indices
have been created for assessing social vulnerability at community level
(e.g. Vincent, 2007; Bell, 2011), including in Mozambique (Hahn et al,
2009). These community indices are based on household level data.
The
value of vulnerability indices is disputed in the literature (Patt et
al., 2009). Some of these criticisms relate to indices in general; and
others relate to the nature of vulnerability. 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. It is
impossible to verify vulnerability at this point in time, and thus
indicators can generally only portray a measure of relative
vulnerability (e.g. between places, or between time periods). 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). Given these uncertainties, many of the indices
presented above use current data to show current social vulnerability,
on the grounds that if the vulnerability exists now, it will likely be
magnified when exposure changes in the future.
However,
current conditions are unlikely to remain constant into the future when
climate changes are projected to occur. 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). Ideally this index should be annually updated with
new data in order to capture temporal shifts. As with all indices, the
assumptions and subsequent methods of transformation used should be
evident, and the index should be subject to a process of continual
testing and refinement. If decision-makers require more specific
information, then estimates of impacts might be more appropriate.
Indeed, Metzger and Schröter (2006) suggest that potential
impacts can be estimated by combining the exposure and sensitivity
terms in expression (1), and this formulation was applied in several of
the studies of ecosystem service vulnerability to climate change in
Europe in the A-TEAM project (Schröter et al., 2005).
Box
3 7: UNFCCC Compendium on methods and tools to evaluate
impacts of, and vulnerability,
and adaptation to climate change
As part of the
Nairobi Work Programme on impacts, vulnerability and
adaptation
to climate change, the UNFCCC maintains various resources
on
its website. The Compendium on methods and tools to evaluate impacts
of,
and vulnerability, and adaptation to climate change is a knowledge
resources
of the Nairobi Work Programme
(http://unfccc.int/adaptation/nairobi_work_programme/
knowledge_resources_and_publications/items/5457.php).
|
1 http://gain.globalai.org/
2 http://www.ehs.unu.edu/article/read/risk-index-maps-world-s-disaster-hot-spots-dw-world-de
3 http://daraint.org/climate-vulnerability-monitor/climate-vulnerability-monitor-2010/
