Social Justice and Adaptation in the UK

Adaptation priorities are usually identified based on the results of top-down, impacts-based assessments that may fail to represent the true social nature of vulnerability to climate change. This paper reviews recent studies to highlight the social nature of vulnerability and to consider the ways in which adaptation may or may not protect the most vulnerable. It identifies a ‘quadruple injustice’ of climate change: that certain social groups emit the least, may be negatively affected by mitigation policies, are most vulnerable to climate impacts, and may be negatively affected by adaptation measures. The paper considers whether public attitudes towards fairness might be invoked to address this injustice by supporting new governance arrangements that transfer resources from high emitters to the most vulnerable. Intro In recent years research and policy have begun to focus on the need to adapt to inevitable climate change. This requires decision-makers at a range of scales (global to local, sectoral, etc.) to assess the likely impacts of climate change in order to identify priorities for adaptation (Brown et al, 2011). To date, few adaptation assessments have considered the uneven distribution of climate impacts and vulnerability across groups and individuals within society. This is partly a result of the tools and methods that are used to inform adaptation policy. Another reason is that the task of assessing individuals’ and groups’ vulnerability to future change is highly complex, given large uncertainties about the direction and pace of future socio-economic and climatic trends and events. This makes it difficult to say with certainty which groups or individuals are ‘most’ vulnerable. Here vulnerability to climate change, climate variability and extreme events is defined as a function of exposure to climate impacts, sensitivity to those impacts and the adaptive capacity of the people or systems impacted (following Blaikie et al, 1994; IPCC, 2007). This paper takes as a premise that society should and does care about social justice as a core value. It also takes the perspective that when assessing the impacts of climate change, the focus should be on who suffers, how much, when and how, and that adaptation should aim to be equitable as well as effective, efficient and legitimate (after Adger et al, 2005). Adaptation should therefore strive to be socially just and to protect those who are most vulnerable to climate impacts. 1 Research Fellow, Stockholm Environment Institute, Sweden magnus.benzie@sei-international.org 2 For example, the current coalition government in the UK chose “fairness” as one of their three core values in the coalition agreement: http://www.general-election-2010.co.uk/2010-general-election-manifestos/election2010-coalition-government-programme.pdf. The Scottish Government declared: “The Scottish Executive has put delivering social justice at the heart of our Programme for Government” in its report ‘Social Justice A Scotland where everyone matters’: http://www.scotland.gov.uk/Publications/1999/11/SocialJustice This paper begins by briefly considering the types of research and methodologies that have been used to support adaptation decision-making. It then focuses on how social justice issues relate to climate change by drawing on recent studies on the distribution of household emissions, the social nature of vulnerability to climate impacts, and the potential for adaptation to reduce or even increase the vulnerability of the most vulnerable. The paper concludes that there is a strong injustice surrounding climate change and adaptation, even within developed countries such as the UK; the paper identifies what it calls the ‘quadruple injustice of climate change’ and briefly discusses some of the implications for the governance of adaptation. Adaptation research and methods Adaptation outcomes will only be as good as the methods and evidence used to inform adaptation decisions. Adaptation decision support methods include climate science, risk assessment, economic analyses and vulnerability assessment. A very brief overview of these methods and their potential for capturing social issues is given below. Climate science Decision-makers take adaptation seriously because of the messages provided by climate science. Various general circulation models are used to project scenarios of future climate, which can be downscaled to model local-level impacts (e.g. precipitation, run-off, flooding, etc.). Information of this type can be used to assess the likelihood and severity of future changes in climate and thereby identify priorities for adaptation. Methodologies based on this kind of top-down approach can loosely be termed ‘impacts-based’ (see Brown et al, 2011). Impacts-based approaches tend to focus on physical and natural systems rather than social systems. This is because the drivers and mechanics of these systems are better known and already modelled to some extent in most cases. Models of the impacts of climate change on hydrology (including flooding) and coastal change are therefore fairly common (HR Wallingford, 2012); new models of physical systems (e.g. urban heat islands) are emerging (e.g. Hoffman et al, 2011). Social systems are highly complex and are rarely modelled as such. Human actors respond to a range of stimuli in often irrational and highly context-specific ways, which makes it difficult to simulate human decisions at the societal scale in models (Goldspink, 2000). For example, it is more difficult to model how an individual will respond to a flood than it is to model how a river system will respond to heavy rain. The result is that science-led, impacts-based approaches fail to represent the social nature of climate impacts. Risk assessment Climate science and impacts model results can be used in a variety of ways to support actual decisions. One of the most commonly advocated methods to support adaptation is to follow a risk management framework (Willows & Connell, 2003; Jones & Preston, 2010). Risk management is particularly appropriate because of the pervasive uncertainty involved with adaptation decisions. One way to undertake a climate risk assessment is to identify a long list of potential impacts and then to scientifically assess their likelihood and magnitude in order to identify a level of risk (Brown et al, 2011). This sort of risk assessment can be seen as a continuation of an impacts-based approach to adaptation decision-making. It implies that there is a value-neutral or scientific measurement of risk. An alternative approach is objective-based risk identification (Institute of Risk Management, 2002). This involves the assessment of a number of possible ‘risk drivers’ on the cost or potential for achieving an explicit set of objectives. This sort of approach is more common in project or corporate risk assessment. It is equally applicable to policy-based organisations or governments, however. It requires decision-makers to be explicit about their objectives, which often involves making normative choices of about what is most important, or what ‘should’ be achieved. The civil servants and advisors that often undertake climate risk assessments aim to appear neutral in terms of future policy choices and are therefore often reluctant to specific strategic objectives. It is sometimes considered ‘safer’ not to state normative preferences and base adaptation decision-making on some value-free, ‘scientific’ (impacts-based) approach. Bradbury (1989) argues that risk management is better when it is based on openly subjective preferences about what is important to society. If policymakers can be explicit about their objective to, for example, improve the quality of life for all citizens, or to reduce social inequalities, then objective-based risk assessments hold significant potential for capturing the social nature of climate impacts, risks and vulnerability. Economic analyses Adaptation decisions can also be supported by economic analyses, including social or project cost benefit analysis (CBA) and, at the global scale, integrated assessment modelling (IAM). The objective of these tools is to identify efficient or optimal policy choices, not to consider equity as a priority criterion. Information on the costs of climate impacts and the benefits of adaptation are limited for most impacts and in most sectors (Watkiss, 2011). However, investment decisions need to be informed by analysis of available options and in some instances, where investment costs and the value of avoided damages can be relatively well understood (e.g. for physical flood defences), CBA is an important and effective decision support tool. Social CBA seeks to maximise welfare from a utilitarian perspective, meaning that there might be winners and losers from an investment, but it will remain attractive as long as the winners are able to compensate the losers and still be better off. However, the distribution of the costs and (dis)benefits from social CBA for adaptation rarely receive much attention, and some argue that the treatment of time preference in CBA via discount rates also raises questions of intergenerational justice (e.g. Ackerman, 2009). Similarly, the use of IAMs to inform decisions on adaptation policy design fails to shed light on social inequality or justice issues and may overestimate society’s ability to adapt because of the crude representation of adaptation decisions in such models (see Stanton et al, 2009; Patt et al, 2009). Economic analyses, whilst important in many respects, therefore fail to adequately account for the distribution of climate impacts across society. The use of impacts-based approaches can be generally characterised as ‘top-down’ (Dessai & Hulme, 2004). Top-down approaches, because they are based on climate scenarios, focus on exposure to harm and tend to see vulnerability as an ‘end-point’ (Kelly & Adger, 2000) or an ‘outcome’ (O’Brien et al, 2007) and therefore static,

This paper begins by briefly considering the types of research and methodologies that have been used to support adaptation decision-making.It then focuses on how social justice issues relate to climate change by drawing on recent studies on the distribution of household emissions, the social nature of vulnerability to climate impacts, and the potential for adaptation to reduce or even increase the vulnerability of the most vulnerable.The paper concludes that there is a strong injustice surrounding climate change and adaptation, even within developed countries such as the UK; the paper identifies what it calls the 'quadruple injustice of climate change' and briefly discusses some of the implications for the governance of adaptation.

Adaptation research and methods
Adaptation outcomes will only be as good as the methods and evidence used to inform adaptation decisions.Adaptation decision support methods include climate science, risk assessment, economic analyses and vulnerability assessment.A very brief overview of these methods and their potential for capturing social issues is given below.

Climate science
Decision-makers take adaptation seriously because of the messages provided by climate science.Various general circulation models are used to project scenarios of future climate, which can be downscaled to model local-level impacts (e.g. precipitation, run-off, flooding, etc.).Information of this type can be used to assess the likelihood and severity of future changes in climate and thereby identify priorities for adaptation.Methodologies based on this kind of top-down approach can loosely be termed 'impacts-based' (see Brown et al, 2011).
Impacts-based approaches tend to focus on physical and natural systems rather than social systems.This is because the drivers and mechanics of these systems are better known and already modelled to some extent in most cases.Models of the impacts of climate change on hydrology (including flooding) and coastal change are therefore fairly common (HR Wallingford, 2012); new models of physical systems (e.g.urban heat islands) are emerging (e.g.Hoffman et al, 2011).Social systems are highly complex and are rarely modelled as such.Human actors respond to a range of stimuli in often irrational and highly context-specific ways, which makes it difficult to simulate human decisions at the societal scale in models (Goldspink, 2000).For example, it is more difficult to model how an individual will respond to a flood than it is to model how a river system will respond to heavy rain.The result is that science-led, impacts-based approaches fail to represent the social nature of climate impacts.

Risk assessment
Climate science and impacts model results can be used in a variety of ways to support actual decisions.One of the most commonly advocated methods to support adaptation is to follow a risk management framework (Willows & Connell, 2003;Jones & Preston, 2010).Risk management is particularly appropriate because of the pervasive uncertainty involved with adaptation decisions.
One way to undertake a climate risk assessment is to identify a long list of potential impacts and then to scientifically assess their likelihood and magnitude in order to identify a level of risk (Brown et al, 2011).This sort of risk assessment can be seen as a continuation of an impacts-based approach to adaptation decision-making.It implies that there is a value-neutral or scientific measurement of risk.
An alternative approach is objective-based risk identification (Institute of Risk Management, 2002).This involves the assessment of a number of possible 'risk drivers' on the cost or potential for achieving an explicit set of objectives.This sort of approach is more common in project or corporate risk assessment.It is equally applicable to policy-based organisations or governments, however.It requires decision-makers to be explicit about their objectives, which often involves making normative choices of about what is most important, or what 'should' be achieved.The civil servants and advisors that often undertake climate risk assessments aim to appear neutral in terms of future policy choices and are therefore often reluctant to specific strategic objectives.It is sometimes considered 'safer' not to state normative preferences and base adaptation decision-making on some value-free, 'scientific' (impacts-based) approach.Bradbury (1989) argues that risk management is better when it is based on openly subjective preferences about what is important to society.If policymakers can be explicit about their objective to, for example, improve the quality of life for all citizens, or to reduce social inequalities, then objective-based risk assessments hold significant potential for capturing the social nature of climate impacts, risks and vulnerability.

Economic analyses
Adaptation decisions can also be supported by economic analyses, including social or project cost benefit analysis (CBA) and, at the global scale, integrated assessment modelling (IAM).The objective of these tools is to identify efficient or optimal policy choices, not to consider equity as a priority criterion.
Information on the costs of climate impacts and the benefits of adaptation are limited for most impacts and in most sectors (Watkiss, 2011).However, investment decisions need to be informed by analysis of available options and in some instances, where investment costs and the value of avoided damages can be relatively well understood (e.g. for physical flood defences), CBA is an important and effective decision support tool.Social CBA seeks to maximise welfare from a utilitarian perspective, meaning that there might be winners and losers from an investment, but it will remain attractive as long as the winners are able to compensate the losers and still be better off.However, the distribution of the costs and (dis)benefits from social CBA for adaptation rarely receive much attention, and some argue that the treatment of time preference in CBA via discount rates also raises questions of intergenerational justice (e.g.Ackerman, 2009).Similarly, the use of IAMs to inform decisions on adaptation policy design fails to shed light on social inequality or justice issues and may overestimate society's ability to adapt because of the crude representation of adaptation decisions in such models (see Stanton et al, 2009;Patt et al, 2009).Economic analyses, whilst important in many respects, therefore fail to adequately account for the distribution of climate impacts across society.
The use of impacts-based approaches can be generally characterised as 'top-down' (Dessai & Hulme, 2004).Top-down approaches, because they are based on climate scenarios, focus on exposure to harm and tend to see vulnerability as an 'end-point' (Kelly & Adger, 2000) or an 'outcome' (O'Brien et al, 2007) and therefore static, as opposed to part of a social process.Top-down approaches, including relatively high-resolution maps or indicators, can also imply that vulnerability is heterogeneous across groups or places, which may be inaccurate and stigmatise certain people or places as being 'high risk' (Benzie et al, 2010).An advantage of top-down assessments, however, is that they can generally be carried out for a large geographical area, for example across a country, or indeed globally.

Vulnerability assessment
'Bottom-up' approaches, also known as vulnerability-based assessments, on the other hand, tend to focus on the impacts of current climate variability and the underlying causes and processes that cause some people to be more vulnerable than others to those impacts (an example is Brown & Walker, 2008).In this way, they place a greater emphasis on adaptive capacity rather than exposure in assessing vulnerability and try to avoid seeing vulnerability as an inevitable effect of certain socioeconomic characteristics (Spiers, 2000).Bottom-up approaches are more likely to incorporate people's own perception of vulnerability and attitudes towards risk, which may ultimately be more important than exposure in determining who adapts and who suffers during climate events.A disadvantage of bottom-up assessments is that they are resource-intensive and generally only apply to specific local areas.A number of qualitative, mostly bottom-up, studies on the social nature of vulnerability to climate change are referred to below.

The UK Climate Change Risk Assessment
The UK has generally been considered a forerunner in adaptation. 3In 2008 the UK parliament passed the Climate Change Act which, among other things, created a duty to conduct a Climate Change Risk Assessment (CCRA) every 5 years.The first CCRA was laid before Parliament in January 2012, with an accompanying report on the economics of climate resilience (ECR) due to be finished in 2013.The reports will play an important role in influencing the UK National Adaptation Programme, due to be published in 2013.
The CCRA includes features of both a top-down, impacts-based assessment and an objective-based one.It takes as its starting point a list of over 700 impacts, identified after considering climate projections, reviewing existing evidence and consulting with stakeholders.This long list of impacts was reduced to around 100 key risks, using a methodology that considered the magnitude of the impact and the level of confidence associated with the evidence (HR Wallingford, 2010).The CCRA aims to identify all climate risks to the UK, but also considers risks in light of key government objectives, not least as the result of risk identification processes that consider Departmental Adaptation Plans, which themselves look at key departmental policies when identifying climate risks (HR Wallingford, 2010).
The CCRA is based on a series of 12 sectoral assessments,4 each of which is led by a separate sector expert and team.This has implications for the way in which social issues are captured -or missedin the assessment.Many if not all of these 12 sectors can be relevant to social justice and inequality in some way (built environment, floods and coastal erosion, health and water are perhaps most relevant).However, by taking a sector approach, there is a danger that impacts are considered in terms of their effect on the functioning of the sectoral system and not on the lives of people affected by that system (e.g.focus on the number of people suffering from heat stress, rather than which people in society are likely to suffer harm as a result of heat stress).
Another key feature of the CCRA -and the ECR report -is that they are both based on existing evidence.The majority of climate change research, particularly quantitative analyses, has taken place within physical and natural sciences.CCRA and ECR results are therefore likely to be skewed by the available evidence 5 and may under-estimate the social nature of vulnerability and risk.
As well as under-estimating the social nature of risk, the scope of the CCRA was limited to impacts within the UK, although as recent evidence suggests, indirect impacts on the UK resulting from climate change elsewhere in the world may be as significant, if not more so, than direct impacts at home (Foresight, 2011).In particular, of the indirect impacts identified in the Foresight report, health, security and migration impacts may affect some groups in society more than others.
The CCRA is the first national assessment of its kind and embodies the proactive approach to adaptation policy being taken in the UK.However, for various methodological reasons, not least the reliance on existing evidence, the first iteration of the CCRA may under-value the social nature of climate impacts, risks and vulnerability.The resulting policies may thus also fail to protect the most vulnerable members of UK society.

Social justice and climate change
Recognising this possibility, the Joseph Rowntree Foundation (JRF), a social policy research and development charity in the UK, initiated a research programme on Climate Change and Social Justice. 6The JRF programme has funded various projects to improve the evidence base on the links between social justice and climate change mitigation and adaptation.This section briefly reviews the results of a selection of projects from the first phase of that programme.Below is a paraphrase of the research questions posed by these projects:  Who emits the most? Who is most vulnerable to climate impacts? Does adaptation protect the most vulnerable?

Who emits the most?
There is a strong correlation between household income and household emissions.A quantitative study by Fahmy et al (2011) explored the nature of this link by compiling a new dataset that combines information on household income, consumption of household fuels, private road travel, public transport use and domestic and international aviation.The report provides new insights into who is responsible for emitting how much carbon dioxide and identifies the relative contributions of different aspects of consumption to household carbon emissions (Fahmy et al, 2011).The graph below shows the breakdown of different emissions types and total emissions for different household income deciles in Great Britain.The relationship between emissions and income is clear.Higher-income households generally emit more; lower income groups emit the least7 .This observation may imply a level of injustice, if it is the case that lower-emitting groups are also more vulnerable to the climate impacts caused by greenhouse gas emissions, as argued below.Fahmy et al go on to use this database of household emissions to analyse the social impacts of mitigation policies, which in some cases have important negative implications for social justice (see also Speck, 1999).

Who is most vulnerable?
Climate impacts will vary between climatic zones and local areas in the UK (CCRA, 2012).However, the harm (or opportunities) that will result from these changes in climate will also vary between social groups within and across different localities.Furthermore, different climate impacts will affect groups differently.
The current evidence base on the social distribution of climate impacts and vulnerability is generally poor.Studies have looked at the relationship between impacts such as heat and mortality (e.g.Basu & Samet, 2002;Hajat et al, 2007;Mirabelli, 2005), sometimes with a social analysis of the results.Literature from the US, drawing on the rich tradition of environmental justice research in that country, has analysed the impact of extreme weather such as heatwaves from a social perspective, for example to analyse the distribution of harm (again, usually mortality) across different ethnic groups (e.g.Klinenburg, 2002;O'Neill et al, 2005;Morello-Forsch et al 2009), as well as other social effects associated with heatwaves, including stress, social disruption, violence and increased crime levels (Simister & Cooper, 2004).Recently, similarly social-based analyses of heat have begun to emerge in the UK (e.g. Brown & Walker, 2008;Wolf et al, 2009).
Following the environmental justice angle, in the UK, where flooding has been a more traditional impact, a small number of studies have looked explicitly at flood vulnerability and environmental inequalities (e.g.Tapsell et al, 2002;Thrush et al, 2005;Fielding et al, 2005;Walker et al, 2006) with a specific Environment Agency research programme focussing on the social aspects of flood management published in 2005 (EA, 2005).This literature did not explicitly focus on flooding in the context of climate change or changes in future risk, however.
More specific research on the social processes that drive vulnerability is not yet widespread.A SNIFFER 8 report in 2009 gave an overview of many of the social issues that are relevant to climate change and identified gaps in research (CAG Consultants, 2009).The JRF Climate Change and Social Justice programme represents the first effort since this report to strategically improve the evidence base and communicate issues directly to policy makers and wider stakeholders.
Below, three recent studies from the JRF programme are reviewed, each looking at different but overlapping aspects of social vulnerability to climate change, namely: Zsamboky et al (2011) on the impacts of climate change on disadvantaged coastal communities, Benzie et al (2011)

Coastal vulnerability
Zsamboky et al (2011) looked at the impacts of climate change on disadvantaged coastal communities.They found a strong social dimension to the exposure of people to coastal flood risk, but also a number of links between deprivation and the ability of households and communities to respond to flood events and adapt to future threats.
Coastal communities in the UK tend to be characterised by high levels of youth out-migration and a corresponding in-migration and concentration of older people, as well as transient groups (including tourists), who are identified as being most sensitive to climate impacts.Coastal communities are particularly vulnerable because of their high reliance on coastal infrastructure, ecosystems and communications, which are especially susceptible to damage and disruption from climate impacts, namely sea-level rise, storm-related damage, coastal erosion and flooding.
A high economic reliance on seasonal employment related to tourism also heightens the sensitivity of coastal communities to extreme weather events and possibly long term climate change (although climate change may present opportunities for tourism-related employment as well).Local health, social and emergency services are also put under pressure during the tourism season, reducing the capacity of these services to support local populations in the event of extreme events such as heatwaves or floods (and rendering tourists themselves at higher risk).
The physical isolation of some coastal communities, particularly those on islands and remote coastlines, adds to their vulnerability, as do the generally old and poor-quality housing stock and below-average income levels associated with deprived coastal communities.The report highlights the risk that certain coastal towns and villages become 'blighted' by the perception of high risks from coastal change and therefore lose property values, job opportunities and investment (and possibly even flood defence protection from central government) leading to further social deterioration and increased vulnerability.Benzie et al (2011) examine the concepts of vulnerability and resilience as they are used by the climate change community and relate these to the concept of social justice.They find that there is likely to be a strong social dimension to climate change vulnerability in the UK.The report focuses on the complexity of social vulnerability to high temperatures.

Vulnerability to high temperatures
In the UK, vulnerability to high temperatures has previously been understood as a function of health and a matter for the health service (DoH, 2010).Whilst accurate, this physiological focus on sensitivity fails to account for the social processes and social context that determine who is able to anticipate, cope with and adapt in order to avoid harm during heatwaves.
There are spatial patterns to the distribution of vulnerability to high temperatures; for example, urban residents are more exposed (Hajat et al, 2007) and within towns and cities it is often deprived areas that are most likely to be located within the centre of urban heat islands (UHIs).Deprived inner-city communities are also less likely to have close access to cool spaces such as green parks, gardens or woodlands.Social factors are also important determinants of vulnerability, although research in this area is generally lacking.
Social cohesion at the community level, including ethnic or faith-based communities that span different urban neighbourhoods, can help to identify vulnerable people and offer support during heatwaves; such informal networks often replace official medical or social services, particularly where population turnover is high and trust between social groups and official services is low, as in some deprived urban neighbourhoods (Brent Council, 2009).In this respect, the presence of strong communities and high social capital may sometimes be higher in some deprived neighbourhoods than in some higher income neighbourhoods (e.g.suburban commuter districts).
Some studies indicate a link between social isolation and mortality during heatwaves (e.g.Klinenberg, 2002), whilst others identify stronger links between mortality and residency in care homes (e.g. Brown & Walker, 2008).Welfare losses from heat, as opposed to mortality, including discomfort, distress, morbidity, violence and social unrest, have received much less attention in the literature, but may also be concentrated in deprived inner-city areas (Benzie et al, 2011).Studies have found links between ethnicity and vulnerability to heat (e.g.Basu & Samet, 2002;Morello-Frosch et al, 2009), depending on non-physiological factors such as employment (Mirabelli, 2005), education and air conditioning (O'Neill et al, 2005).
Many studies find a link between age and vulnerability to high temperatures (e.g.Fouillet et al, 2006), but it is also important to note the role of risk perception among all groups (Grothman & Patt, 2005).Various studies show that people who do not perceive themselves to be at risk are less likely to take measures to prevent the effects of heat stress, even when they are in fact at high risk (Abrahamson et al, 2008;Wolf et al, 2009).
One study has looked at the links between employment and climate change vulnerability, uncovering an additional social layer of vulnerability whereby people's occupation exposes them to a greater level of climate risk, for example lower-paid, unskilled jobs (e.g.outdoor manual labourers, train drivers or factory workers) are higher-risk than higher-paid, high-skill jobs (e.g.those in airconditioned offices) (TUC, 2009).
One key finding of this literature is that vulnerability to climate events is highly dependent on local context, and that a better account of individuals' and groups' adaptive capacity is needed to understand the social nature of vulnerability in each case (e.g. Brown & Walker, 2008).Lindley et al (2011) identify the determinants of vulnerability to heatwaves and floods and use these to develop a spatially explicit index of vulnerability that accounts for the social nature of exposure and adaptive capacity.The results constitute the first quantitative representation of the potential for future losses in well-being that take explicit account of the social nature of vulnerability. Figure 2 below shows the results of the index for flood vulnerability in England.In total 8% of English neighbourhoods are estimated to have extremely high flood-related social vulnerability, with a strong concentration of vulnerability in deprived urban and coastal areas (38% of the areas are within 2km of the coast).These areas are represented as having low capacities to prepare, respond and recover from flood events, according to the index.The specific variables that influence the distribution of social flood vulnerability shown in this map include: lack of gardens and green space (which help to regulate run-off flows); a proxy measure of insurance availability -based on current flood probabilities; low income; poor knowledge; and low mobility.A similar index has been produced for social vulnerability to heat (see Figure 3 below).A similar proportion of neighbourhoods are estimated to have extremely high social vulnerability to heat as to flooding: 9% in this case.Again, there is a strong urban dimension to the results, particularly in London, which has three times the mean level of heat vulnerability compared to other regions and 40% of all of the extremely vulnerable neighbourhoods in England.Interesting detail emerges in the disaggregated results, which show a low capacity to respond to high temperature events in remote rural areas that are isolated from medical services, for example; even though exposure is not particularly high in rural areas.There is also a strong coastal component, which is influenced by the adaptive capacity indicators rather than the signal of increased exposure from higher temperatures, underlying the dominantly social nature of vulnerability to climate change.Taken together, the two aggregate indices of social vulnerability show the significant overlap between vulnerable neighbourhoods to multiple climate impacts; fully 64% of the extremely socially vulnerable neighbourhoods to flooding are also classified as extremely vulnerable to heat.This has significant implications for the design of adaptation strategies: If the same social groups tend to be vulnerable to multiple climate impacts, then a key element of adaptation strategy should be to protect and build resilience among these most vulnerable groups.

Does adaptation protect the most vulnerable?
Adaptation implementation is in its early phases in the UK.Despite a growing body of research and an improved level of awareness among decision-makers of the need to adapt, specific adaptation actions are few and difficult to identify (ASC, 2011).Nevertheless, the JRF programme has undertaken some case studies to assess emerging lessons on whether adaptation offers protection to the most vulnerable.

Planned vs. autonomous adaptation
Adaptation can either be planned (usually by a national or local government) or autonomous (i.e.undertaken by private actors in response to their own calculation of costs and benefits) (Smit & Pilifosova, 2001).Planned adaptation should be carefully considered; the scope for considering social justice issues should be high.Autonomous adaptation may occur within narrower contexts, and the potential for mal-adaptation may be higher.Mal-adaptation is generally understood as those actions taken ostensibly to avoid or reduce vulnerability to climate change that impact adversely on, or even increase the vulnerability of, other systems, sectors or social groups (Barnett & O'Neill, 2009).
Below, two examples of autonomous adaptation (water pricing and flood insurance) and one example of planned adaptation (heatwave planning) are considered through a social justice lens.

Improving water efficiency in response to drought risk
The south west of England is projected to become much drier as a result of climate change, particularly in the summer, with extreme scenarios suggesting up to 50% or 70% less rainfall (UKCP09).This region also happens to have the fastest-growing population in the UK (ONS, 2003) with the number of households projected to increase by 36% by 2030 (DCLG, 2009).The south west is also one of the most popular tourism destinations in England, with over 21 million visits per yeara figure that is also projected to rise due to changing patterns in the tourism industry and also due to climate change impacts elsewhere in Europe (EEA, 2008).Improving water efficiency is therefore a priority for South West Water, the utility company that manages and delivers water services in the region.
However, water affordability is already a serious issue in the south west, where prices are around 40% higher than in other regions of the UK, due to a mixture of legacy, infrastructure and tourism factors.Delivering affordable water efficiency is therefore the additional challenge.
Benzie et al (2011) examined a pilot project by South West Water to introduce a Rising Block Tariff for water customers.In theory, this tariff structure incentivises water efficiency whilst delivering affordable water to all.It offers three differently priced 'blocks' of water use:  Block 1: 'essential use' at 73% of the standard unit price  Block 2: standard price (a buffer)  Block 3: 'premium use' at 181% of the standard unit price The theory is that users who reduce their use are rewarded with cheaper water, and those who chose to use more pay a premium for doing so.The system relies on there being a free and equal choice between households on how much water they use.Yet the Benzie et al case study showed that water use requirements differ between households; some are less able to reduce their water use, as a result of household size, certain medical requirements or even as a result of tenure and inflexibility to fit water-saving devices or inability to purchase water-efficient technologies (e.g.new washing machines).Such households, if on low incomes, may be unfairly penalised by the introduction of a Rising Block Tariff system, raising the prospect that water efficiency schemes could push more households into a situation of 'water poverty' (defined as spending more than 3% of disposable income on water bills, after Fitch & Price, 2002).
The case study also revealed the role of support schemes that protect low-income households from water poverty where metering and new tariff structures are in place.In the south west, the WaterSure scheme caps bills for qualifying households (those on low incomes or with defined medical requirements) and the WaterCare scheme aims to improve water efficiency and provide support to households in debt with water companies.Thus, efficiency incentive schemes, including water metering and new tariff structures, are not inherently regressive and do provide the potential to address climate risks (i.e.drought) in socially just ways, provided that they are always implemented in tandem with support schemes for vulnerable households.Benzie et al (2011) and the independent Walker Review of affordability and water charging (Walker, 2009) identify various features of the water sector that are important for maintaining affordability and make various recommendations to ensure that water poverty is avoided as a consequence of mal-adaptation to climate change.
In this case, autonomous adaptation (i.e. using pricing mechanisms to address resource scarcity and reduce risks) only avoids being unjust because consumer rights are well represented in the water sector, which is heavily regulated by the UK government.

Future flood insurance
Flood insurance in the UK is currently governed by an agreement between the state and the insurance sector (known, in its most recent incarnation, as the 'Statement of Principles').In short, the state commits to provide flood defences and prevent development in very high-risk areas, in return for a commitment from insurers to provide insurance cover to all households and most small businesses (see Crichton, 2002).This agreement is due to expire in 2013, stimulating a lively debate between insurers, the government and various stakeholders on what should replace it.The JRF recently published a 'viewpoint' report that addresses the social justice aspects of this debate (O'Neill & O'Neill, 2012).
One in six homes in England is currently at risk from flooding (EA, 2009), and low-income households are the likeliest to be uninsured and the least able to recover from the financial impacts of flooding (Pitt, 2008).Flood risk is increasing in the UK as a result of increased development (more and higher value homes) and climate change (including changing precipitation patterns and sea level rise).
Looking into the future, there is a tension between creating disincentives to live and develop in flood zones and penalising people who already live there.If the insurance market were left to adapt autonomously by simply pricing the actual risk for each household, insurance rates would drive demand for housing in low-risk zones (and therefore raise property values), whilst high-risk areas would become cheap (attracting low-income households), 'blighted' and potentially uninsured or uninsurable, creating significant inequalities and social injustice (O'Neill & O'Neill, 2012).
Alternative (non-market) insurance models exist and are common in other European countries, where the state often plays a larger role.O'Neill & O'Neill (2012) explain how different concepts of fairness imply different insurance models and argue strongly for a more solidaristic flood insurance regime in the UK.The case of flood insurance, as an example of autonomous adaptation where risk is priced, shows the inability of some forms of adaptation to protect the most vulnerable and a much more worrying possibility that mal-adaptation may significantly increase the vulnerability of some groups to climate change.

The Heatwave Plan
Temperatures in the south of England are projected to increase significantly over coming decades, with extreme heatwave events in particular likely to cause significant harm and economic damage (Stott et al, 2004, Metroeconomica, 2006).Demographic trends in parts of England, particularly the south west, are likely to increase the population's vulnerability to high temperatures, largely as a result of increases in the number of older people (the proportion of whom is projected to rise from 22.5% -already above average -to 29% by 2031) and increased numbers of tourists.
The only current strategy for explicitly managing risks relating to high temperatures in the UK is the Heatwave Plan for England and Wales, led by the Department of Health.The existence of the strategy is in part a reaction to the unforeseen impacts of the 2003 heatwave in Europe, which led to the premature deaths of around 2,000 people in the UK (Metroeconomica, 2006) in addition to 30,000 deaths and billions of Euros' worth of damage across mainland Europe (COPA-COGECA, 2003).
The Heatwave Plan relies on various tiers of governance to 'cascade' down responsibilities during heatwave events.Different Levels of alert are defined within the plan and are triggered once weather forecasts exceed certain thresholds (see DoH, 2010; DoH 2011, for details).
Benzie et al (2011) interviewed various 'responsible authorities' under the Heatwave Plan and other relevant stakeholders in order to examine their understanding of 'vulnerability to high temperatures' and to see how the concept of vulnerability was operationalised in practice.Unsurprisingly, given the complexity and lack of an evidence base identifying vulnerable people (alluded to above), there is some confusion and an oversimplification of vulnerability in practice (which tends to revert to healthbased definitions of vulnerability).This tends to ignore the social processes that determine vulnerability and therefore misses opportunities to build resilience to high temperatures, rather than merely to respond during emergency situations.In many cases it is unlikely that the most vulnerable people will receive support during heatwave events, largely because health services are not able to identify who is most vulnerable.
The case study found that resources for implementing the Heatwave Plan were severely limited, and responsibility to implement the plan fell mostly on emergency planning departments within local government and health care professionals.This precludes the potentially crucial role of local stakeholders who benefit from a more detailed understanding of vulnerability and climate change, including social services, climate change partnerships, researchers and, in particular, community groups.The role of stakeholders who have the ability to help build resilience to high temperatures, for example spatial planners and educators, is not yet a feature of Heatwave Plan implementation, according to Benzie et al.However, a key recommendation in the report is to go beyond the Heatwave Plan (which is understandably focussed mostly on health and emergency response, given its remit) to take a much more proactive and integrated approach to build resilience to high temperatures in the UK in ways that explicitly promote social justice and equity.

Quadruple injustice
The cases presented above highlight the social nature of vulnerability and the uneven distribution of vulnerability to climate change.They also highlight the potential for mal-adaptation to increase certain groups' vulnerability (especially autonomous mal-adaptation) and the uneven and uncertain benefits that adaptation might bring to certain social groups in the UK.
In conclusion, it is possible to identify a 'quadruple injustice' to climate change within the UK.Certain disadvantaged groups9 , including those on low incomes, the socially marginalised and older people:  Emit the least.
 May be negatively impacted by mitigation policies. Are most vulnerable to climate impacts.
 May be negatively impacted by adaptation policies.
This presents a particular set of challenges to adaptation decision-makers.
First, it raises questions of procedural justice.How can the voices of the most vulnerable be heard in the design and implementation of adaptation policies?It has been shown that the implementation of the Heatwave Plan in England and Wales has so far failed to involve vulnerable groups, or organisations that represent vulnerable groups, in the process of identifying who is vulnerable or in delivering emergency response services during heatwave events.However, the involvement of the Consumer Council for Water in the RBT trial in the south west of England is evidence of more participatory decision-making process, which also happens to result in a more 'just' outcome (Benzie et al, 2011).Given the complexities involved in adaptation policy design, including the treatment of complex and uncertain science and the normative choices relating to risks that characterise adaptation, it is both difficult and yet essential that vulnerable groups be brought into adaptation decision-making.
Second, questions of substantive justice have been raised, particularly with respect to autonomous adaptation and, most of all, where market forces are employed to deliver efficient adaptation outcomes, whether these are in the pricing of risk (via free market insurance models) or via scarcity and price signals (via water metering).
Third, there is the potential for adaptation to enhance social justice, although this has not been the focus of this paper.Adaptation measures, for example the creation of sustainable urban landscapes that offer free, cool public spaces and reduce flood risks, could improve the quality of life for residents and facilitate more cohesive community living spaces.

Methods
This paper began with a brief review of methodologies used to inform adaptation decision making.This analysis has implications for the governance of adaptation.On one level, over-reliance on topdown assessment techniques may tend to 'hide' the social nature of vulnerability and lead to adaptation strategies that fail to protect the most vulnerable.This would be the consequence of focussing on the size of a risk, or its aggregate costs, rather than on the social nature of the risk: who will suffer harm as a result of that risk.
As has been seen from the case studies, vulnerability to multiple climate impacts tends to overlap for certain social groups, namely those with low adaptive capacity, who now tend to be marginalised and disadvantaged in society.The implications of this for adaptation governance are that more bottom-up analyses should be used to inform adaptation policy, incorporating procedural elements, such as more consultation with vulnerable groups, as well as methodological elements that base climate risk assessment more on current climate vulnerability and that focus more on identifying cross-cutting issues from different sectors.
The UK Climate Change Risk Assessment provides an interesting case.Whilst initially designed as an impacts-based risk assessment, the CCRA methodology was modified to better account for the social aspects of risks, perhaps in response to the evidence generated and put forward by the Joseph Rowntree Foundation.The CCRA methodology now includes a stage (step 5) to explicitly consider equity issues and social vulnerability.Within this stage, broad clusters of risks have been assessed using a Social Vulnerability Checklist (HR Wallingford, 2010).An evidence review report on social vulnerability to climate change impacts (CEP, 2011) was also prepared as part of the CCRA process to inform decision-makers (though this has not yet been made public).
Step 7 of the CCRA methodology, which develops risk metrics for each risk, also provides the opportunity to develop social metrics that can be used to measure changes in risks relevant to social justice (e.g.changes in the number of deprived households at risk from flooding, see HR Wallingford, 2010).In these ways, the social nature of vulnerability is recognised in the CCRA, and attempts have been made to update the methodology to better account for social vulnerability.Top-down assessments can therefore be carried out ways that do draw attention to the social nature of vulnerability and risk, although current evidence gaps make this difficult in practice.

Governance
The quadruple injustice of climate change challenges adaptation governance to become more just and to deliver more just outcomes.Below, the concept of a new policy concept to transfer funds between high emitters and the vulnerable is introduced, based on the logic of the quadruple injustice.
Public attitudes to fairness represent a barrier but also an opportunity for designing new, socially accepted climate change policies.Another report from the JRF Climate Change and Social Justice programme (Horton & Doran,211) used focus groups to look at people's sense of fairness in relation to behaviours and rules governing climate change.For example, they looked at collective antipathy towards 'freeriding' behaviour and public support for rules to prevent excessive consumption (specifically in situations of 'resource scarcity', such as climate change10 ).The report concludes that climate change regulations do not need to appeal solely to self-interest and to cost-saving opportunities -as many regulations and policy initiatives currently do.Instead, regulations may be more successful if they appeal to people's sense of fairness, based on the link between excessive consumption (i.e.emissions) and climate change (Horton & Doran, 2011).Whilst the focus groups did not explicitly ask people about their sense of fairness in relation to adaptation, the results present an interesting question: could people's sense of fairness be harnessed to address the quadruple injustice of climate change by transferring resources from high emitters to the most vulnerable?
The logic behind this question is that a scientifically robust causal link can be established between excessive consumption (e.g.driving a high-emitting SUV, or frequent flying), physical climate change and impacts on vulnerable people (e.g.flooding a family living on low income).Whilst it is unlikely that an acceptable scheme could be designed for direct payments between emitters and vulnerable groups, there may be public support for policies that use revenues from taxes on high-emitting behaviour to protect compensate or protect vulnerable groups, or preferably to invest in building resilience among vulnerable people.This, after all, is the logic applied to global negotiations on climate change, where developed countries have agreed to provide significant financing to help the most vulnerable countries adapt, via the Green Climate Fund and the Adaptation Fund, in recognition of their historical emissions and greater financial capacity. 11.Public awareness of climate change, vulnerability and adaptation would need to improve significantly before the fairness aspects of such a policy could be widely understood and accepted, but perhaps one day, similar mechanisms could be applied within countries to fund investment in socially just adaptation.

Conclusion
This paper has briefly reviewed the dominant methods used to inform adaptation policy, which are generally top-down, impacts-based approaches.Adaptation based on these assessments may fail to protect the most vulnerable groups in society.Mal-adaptation, particularly where market forces are relied upon to increase resource efficiency or to price risk, may have particularly unjust outcomes.Top-down assessments can, however, be carried out in ways that draw attention to the social nature of vulnerability, as the UK CCRA demonstrates.However, better evidence on the social distribution of vulnerability to climate change and on the social impact of planned and autonomous adaptation is needed, including more bottom-up vulnerability analysis at the local level.Policy-makers should consider appealing to the public's sense of fairness to support new policies that transfer funds from high emitters to invest in resilience-building for vulnerable groups.Greater efforts should also be made to boost public awareness of the need for climate change adaptation in the UK.
11 The historical nature of these emissions are pertinent to the global debate, but problematic when applied to transfers between emitters and the vulnerable within current generations (e.g. an individual may question the credibility of claims for compensation between members of the same generation when current climate impacts are the result of emissions more than 20 or 30 years in the past).This paper does not set out to explore this concept in detail, though a discussion on these issues would be welcome.

Figure 1
Figure 1 Mean annual total CO2 emissions from all sources by equivalised household income decile (metric tonnes of CO 2 on y-axis, with income deciles on x-axis) (Source: Fahmy et al, 2011:7).

Zsamboky et al (2011) Benzie et al (2011) Lindley et al (2011)
Lindley et al (2011)heatwaves and drought, andLindley et al (2011)on climate change, justice and vulnerability.Table1below provides an overview of the socially contextual factors that determine vulnerability to climate change, drawing on these three projects.