2.1. Systems

Cities require high levels of infrastructure to deliver essential services. Cities are also linked across multiple scales to other systems, such as regional food production that relies on ecosystems to deliver provisioning services. At the global scale, cities are connected through international trade and investment patterns, which can have direct effects on local employment and livelihoods as well as on supplies ranging from pharmaceuticals to imported staple foods. For example, in the Bangkok floods of October–November 2011, the flooding of local manufacturing facilities affected global supply chains for computer and automobile components and led to temporary factory closures and layoffs in many cities outside Thailand (Chachavalpongpun, 2011).

The underlying support systems that enable networks of provisioning and exchange for urban populations are therefore an essential element of urban resilience. They include physical infrastructure and ecosystems, either within the city, immediately adjacent or remote ecosystems that provide key services such as food production, runoff management or flood control. While local managers may have limited influence beyond city boundaries, their systems can be strongly affected by factors at multiple scales and at long distances. City electricity distribution depends on the performance of distant reservoirs, generators and grids. Crop losses on the other side of the world can affect local supplies and prices for widely traded food commodities.

Core or ‘critical’ systems are essential to urban function (Little, 2002). Their failure jeopardizes human well-being in all affected areas, and precludes higher order economic activity until their function is restored. These systems include water and food supply, and the ecosystems that support these, as well as energy, transport, shelter and communications. In assessing the potential for these systems to fail under climate-induced stress, it is crucial to recognize the interdependencies of complex linked systems  because failures of one system often lead to cascading failures in linked systems (Kirshen, Ruth, & Anderson, 2008).

Resilient systems differ from an engineering approach to robust systems, which rely primarily on hard protective structures (e.g. sea walls) or are designed in ways that emphasize the strength of specific individual components to ensure functionality. Resilient systems, in contrast, ensure that functionality is retained and can be rapidly re-instated through system linkages despite some failures or operational disruptions (Bruneau et al., 2003; McBain, Wilkes, & Retter, 2010; O'Rourke, 2007). Rather than relying on the strength of individual components, resilient systems retain functionality through flexibility and diversifying functional dependence.

From the study of complex engineering and ecological systems, characteristics that are widely cited as contributing to the resilience of systems include the following (Alberti et al., 2003; Andersson, 2006; Bruneau et al., 2003; Ernstson et al., 2010; Folke et al., 2002; Gunderson & Holling, 2002; Leichenko, 2011; Liu et al., 2007; Meadows, 1999; Resilience Alliance, 2007):

These characteristics of resilient systems should be seen as guidelines for thinking about complex urban systems in new ways, rather than as technical prescriptions. Each context and system will be different, and it is impossible to provide specific prescriptions for all conditions, although there is a growing body of research looking at the resilience of specific systems and the wider consequences and costs of disruption or failure due to climate change (Bruneau et al., 2003; Brunner, Steelman, Coe-Juell, Edwards, & Tucker, 2005; FEMA, 2012; McBain et al., 2010). Table 1 presents examples of typical performance specifications and system configuration in relation to water supply, so that readers can better grasp how the general descriptions above can be interpreted in light of hypothetical local conditions. From these examples, it can be seen that system characteristics should not be considered as mutually exclusive categories. In any given system, a particular desired performance factor might be ascribed to more than one category (in some systems, modularity may be similar to diversity, e.g. multiple water pumping stations in various locations). Users of the framework could construct their own descriptions in sectors of interest based on their specific local conditions.

2.2. Agents

The recent literature on resilience in complex adaptive systems emphasises the integration of social agents and institutions along with biophysical elements as components of socio-ecological systems (Folke, 2006; Folke et al., 2002; Gunderson & Holling, 2002). It is also argued that the adaptive capacity of social organizations and individuals is a concept closely related to resilience (Berkes, 2007; Folke et al., 2002; Gallopin, 2006). Key aspects of resilience addressed in relation to hazard assessment and disaster risk reduction in cities have included flexibility and diversity (addressed under the Systems section) and capacity for learning and innovation (Leichenko, 2011). The capacities of social agents therefore comprise an important part of any urban climate resilience framework.

Social agency differs from system function in that outcomes arise not only from interaction between elements but also from purposive decisions. Agents, unlike systems, are capable of deliberation, independent analysis, voluntary interaction and strategic choice in the face of new information. Agents are actors in the sense that they introduce volition and intent into choice; they behave in ways that reflect their location and structure within society (i.e. as government entities, businesses, community advocates, households and individuals), their preferences, and the opportunities and constraints they perceive. Techniques for analysing agent behaviour and capacity are different from those required for systems.

Agents, or actors in urban systems, comprise the second key element in the resilience framework. They include individuals (e.g. farmers, consumers); households (as units for consumption, social reproduction, education, capital accumulation); and private and public sector organizations (government departments or bureaus, private firms, civil society organizations). They have identifiable but differentiated interests and are able to change behaviour based on strategy, experience and learning. In order to work effectively with agents it is important to recognize the opportunities and constraints they face and the incentives to which they respond. Agent behaviour can be changed, but depending on the circumstances this may not be any easier than modifying complex technical infrastructure systems.

Many agents (e.g. households) depend on urban systems and demand services but are not proactively involved in the creation, management or operation of those systems. Other agents are directly concerned with management of critical urban systems. In the case of water supply, for example, these might include the municipal water utility, key water quality or regulatory agencies, private water market suppliers and civil society organizations involved in water-related advocacy.

Resilience is not a characteristic that is evenly spread through the urban population. It depends crucially on the socially differentiated capacities of different groups and individuals. Poverty, gender, ethnicity and age have all been documented as contributing to differential vulnerability of social groups in cities to climate hazards, through features such as the quality of housing, location and access to services or social networks (Moser & Satterthwaite, 2010; Pelling, 2003; Satterthwaite, Dodman, & Bicknell, 2009).

For individuals and households in particular, key capacities can be considered as assets, that is, a stock of resources upon which individuals and social groups can call in order to advance their well-being. The main asset categories that comprise these resources include financial assets (wealth or access to credit), physical assets (house, possessions), natural assets (land or rights to streams of ecosystem services such as fish or water), social assets such as family or clan networks, and human assets (health and skills) (Moser, 2006). Even the very poor have assets such as knowledge or social contacts. These serve as the basis for an agent's power to act. For these categories of agents, the social differentiation of climate vulnerability is closely linked to differential asset profiles. Losses caused by climate hazards typically erode multiple types of assets and further impoverish such groups (Moser & Satterthwaite, 2010; Pelling, 2003).

But adaptive capacities of other social organizations are also important to resilience. The role of local governments and of community organizations is crucial here, as these are the primary sources of organization and delivery of key planning, prevention and response services (e.g. land use, building controls, emergency services) that are essential for ensuring urban resilience and climate adaptation (Satterthwaite et al., 2009). The capacity to access a range of assets or resources in order to take adaptation action includes the collective effort to use a wide range of knowledge to assess risks (Berkes, 2007).

Positive examples of asset-based adaptation demonstrate the opportunities for poor urban communities to organise and plan for preventive risk reduction strategies, either autonomously through community groups such as savings or credit groups, or in collaboration with local governments (Moser & Satterthwaite, 2010). This includes the ability to organise, plan and coordinate for disaster preparedness and emergency response as contributing to resilience (UNISDR, 2012).

The capacity of individuals and organizations to learn is a crucial aspect of resilience approaches across a range of disciplines (Berkes, 2007; Diduck, 2010; Folke, 2006; Gunderson & Holling, 2002). Learning includes not only the mobilization and sharing of knowledge but also such factors as basic literacy and access to education. These kinds of factors have been identified empirically as contributing to community resilience to disasters (Twigg, 2007).

We summarize and label these agent capacities consistent with the terminology used in resilience thinking and in disaster risk reduction as follows (Gunderson & Holling, 2002; Twigg, 2007):

High-capacity agents have the ability to anticipate and act in order to adjust to external changes and stresses. Organizations have the authority and mandate to take action, as well as the financing to do so. Agents’ ability to act is facilitated by adequate resources and by access to supporting systems, including the ability to access resources provided by other agents. Agents may develop these capacities through experience, gradually acquiring a repertoire of responses to stresses and shocks. The awareness of hazards, the ability to learn new responses and the ability to acquire information needed to assess hazards and outcomes are, therefore, all important elements in strengthening the capacity of agents.

The illustration in Table 2 presents examples of agent capacities that are consistent with the concepts derived from the literature summarized above, using the urban water sector as an example.

2.3. Institutions

The concept of institutions in social sciences refers to the social rules or conventions that structure human behaviour and exchange in social and economic interactions (Hodgson, 2006). Institutions may be formal or informal, overt or implicit, and are created to reduce uncertainty, to maintain continuity of social patterns and social order, and to stabilize forms of human interaction in more predictable ways (Campbell, 1998; North, 1990; Ostrom, 1990). Institutions condition the way that agents and systems interact to respond to climate stress, so this is the third element of the resilience framework.

Institutions of property and tenure, of social inclusion or marginalization and of collective action influence the vulnerability of particular social groups (Adger et al., 2005). Institutions that enable or constrain individuals to organize or to engage in decision making (i.e. who is seen as a legitimate ‘stakeholder’) determine whose interests are considered in political decision making. Similarly, the standards to which systems are designed and managed, as with building and engineering codes, have a strong influence on whether those systems will reliably meet the needs of users. Finally, the pricing structure for urban services is an institution that influences access to infrastructure systems and the resilience they offer, particularly for the urban poor (McGranahan, 2002).

Institutions may enable and support, or constrain and inhibit, the capacities of vulnerable urban groups (Moser & Satterthwaite, 2010). For example, the migration of farm labourers in times of drought depends on formal and informal institutional structures that organize the labour market through communication, recruitment, transport and remittances. Those institutions facilitate migration as an adaptive response and enable households to diversify their sources of income (Moench & Dixit, 2004). Urban planning decisions such as slum clearance and resettlement may increase or decrease climate vulnerability depending on the institutions governing rights, compensation, participatory planning and decision making associated with the resettlement process (Satterthwaite et al., 2009). With inadequate consultation or participation, minimal rights and only token compensation, resettlement could increase impoverishment and vulnerability. However, under different institutional conditions, the outcomes could be opposite. Institutions for collective action and governance can also be designed to strengthen resilience by supporting ecosystem restoration and sustainability (Adger et al., 2005; Folke, Hahn, Olsson, & Norberg, 2005; Ostrom, 1990; Tompkins & Adger, 2004).

Social marginalization – the process whereby some social groups are excluded either formally or informally from access to critical services or resources – is closely linked to institutions (Beck & Fajber, 2006). The concept of entitlements has been introduced to explain the way that institutions shape how individuals are able to claim access to resources, food supplies or other services (Leach, Mearns, & Scoones, 1999; Sen, 1981). An example of an institution that fosters marginalization is the requirement for legal registration of residence in Vietnam or China, which can make it difficult for migrants to gain access to services, social benefits or to government compensation for disaster impacts. Those individuals and groups who are systematically marginalized through institutions that delegitimize their claims to the services provided by urban systems (i.e. have fewer entitlements) are likely to be more vulnerable to similar climate impacts (Moser & Satterthwaite, 2010; Pelling, 2003).

Governance (i.e. the process of decision making) is an important factor affecting resilience. Decision-making processes that build resilience for vulnerable groups are likely to be participatory and inclusive, allowing those individuals and groups most affected by climate hazards to play an active role in determining how best to avoid them (Lebel et al., 2006; Satterthwaite et al., 2009). Many authors argue that adaptation and resilience require local governments to be accountable to marginalized populations, which is different from current practices that often actively discriminate against them (Moser & Satterthwaite, 2010). And because proactive attempts to build climate resilience require coordinated actions by many different local government departments, new mechanisms for collaboration between these departments, and with civil society organizations, are typically needed.

Public information is an important component of a positive institutional environment. Communities who have access to timely hazard information are better able to respond to climate threats, even in vulnerable sites, especially when this is matched with credible and supportive advice on appropriate response such as evacuation routes and transport support (Moser & Satterthwaite, 2010).

Institutional structures that foster learning and change are important tools to build agent capacity. Public and private support for applied research, for publication and presentation of new evidence, and for facilitating critical assessment of new knowledge and its implications all speed the introduction of effective innovation. In the absence of these types of institutions, professional norms and legislated codes or standards may act as barriers to innovative practices. Institutions capable of fostering evolutionary change, and of adapting to new information, lend themselves to building resilience (Berkes, 2007).

From studies of economic behaviour, collective action, social marginalization and decision making, the key aspects of institutions linking agents and systems that should be considered in assessing whether they enhance or constrain resilience appear to be those outlined below2:

An important right that can enhance resilience is that of groups to self-organize and to engage in collective choice mechanisms (Huntjens et al., 2012) in order to respond to climate hazards (e.g. to improve local drainage, or to deliver disaster preparedness training). The ability of self-organizing groups to assess fees or levies on members in order to undertake investments in service delivery, maintenance or resilience investments is central to their ability to act but sometimes may be constrained by legal limitations on the organization of civil society (Table 3).

2.4. Resilience, exposure and vulnerability: using the framework to understand vulnerability

The terminology of resilience, exposure and vulnerability is widely used across several related fields, but with little consistency or consensus on definitions (Berkes, 2007; Gallopin, 2006; Klein et al., 2003). Low resilience systems are intrinsically vulnerable to stress and shock, so in this sense increasing resilience reduces vulnerability (Folke, 2006). However, these two concepts have different origins and published studies: resilience has emerged from a positivist biophysical scientific perspective, while vulnerability has been described mainly from a constructivist social science and political ecology framework (Miller et al., 2010).

Most of the literature reviewed for this paper presents vulnerability as a lack of resilience. An exception is Gallopin (2006) who describes vulnerability as a different concept, broader than resilience, which he links more closely to the notion of ‘adaptive capacity’. An important distinction is that most authors (excepting Gallopin again) include hazard exposure as a component of vulnerability. In other words, vulnerability is argued to have meaning only in relation to a particular hazard (Klein et al., 2003). Mountain dwellers are not vulnerable to sea level rise. Vulnerability to flooding involves assessing different factors than those assessed for vulnerability to earthquake. By comparison, resilience is seen as an intrinsic or emergent feature of complex social–ecological systems (Folke, 2006). Resilience is only manifest through exposure to stress or shocks, and in the subsequent recovery or reorganization period, but its latent character exists within a system independent of that exposure.

We pointed out above that marginalization imposes both capacity and institutional barriers to adaptation, but it is important to recognize that depending on the local context, other social groups may be more vulnerable than the poor due to their exposure (e.g. coastal residences, mountain homes subject to wildfire) or due to their greater dependence on critical systems (e.g. electricity for borehole pumps).

The question of exposure is also more complex than it may at first appear. Some of the greatest stresses on urban areas from climate change are likely to be indirect, incremental or both. They will emerge as a consequence of distant changes that are translated to urban areas through interlinked systems as a result of global markets, supply chains and dependence on remote ecosystems or wider infrastructure networks. This suggests again that while such indirect linkages may be missed in vulnerability assessment, they can be managed through resilience building.

While many climate-driven shocks and stresses in urban areas can be projected on the basis of experience and climate science, others will be unanticipated. Vulnerability to such surprises may not be apparent in advance, so building resilience is a useful strategy to prepare. Increasing the inherent capacity of complex systems to manage a range of stresses and shocks, through experience or strategic preparation, better enables such systems to deal with surprise (Tompkins & Adger, 2004; Walker et al., 2002).

The conceptual framework introduced above defines the key constituent elements of urban resilience. Following this framework, vulnerability to climate change occurs when fragile, inflexible systems and/or marginalized or low-capacity agents are exposed to increased climate hazards, and their ability to respond or shift strategies is limited by constraining institutions. Resilience is high where robust and flexible systems can be accessed by high-capacity agents and where that access is enabled by supportive institutions (see Figure 1). Figure 1 illustrates that vulnerability occurs in the darker areas of the diagram, where capacities are low, systems fragile and institutional support weak. By building resilience, through intervening with agents, systems and institutions, vulnerability is reduced and well-being strengthened.

A framework for urban climate resilience

All authors

Stephen Tyler & Marcus Moench

https://doi.org/10.1080/17565529.2012.745389

Published online:

14 December 2012

Figure 1. Urban systems enable agents to improve their well-being by adopting new strategies as conditions change. Resilience to climate change increases by building agent capacities, strengthening systems and reinforcing institutions that link agents and systems.

Display full size

2.5. Linkages between elements in the framework

The key elements of the framework for urban resilience as outlined above are infrastructure and ecosystems (both captured under the rubric of ‘systems’), social agents and institutions. For analytical purposes, we separate these elements, although we recognize that other analysts frequently refer to ‘urban systems’ that are integrated to include all these categories (da Silva et al., 2012). We find that analysing component dimensions separately is more practical. While we characterize infrastructure and ecosystems alike as ‘systems’ in our categories above, each of these elements requires quite different kinds of analytical methods in order to understand their resilience characteristics. In addition, by identifying and treating these elements separately it is easier for local government or civil society organizations, with limited sectoral or thematic interests, to engage in the framework and to identify relevant issues. With a variety of entry points and analytical approaches available, local experts and practitioners should be able to relatively easily identify starting points that relate to their own domain and expertise.

Within this conceptual framework, building urban climate resilience means:

While separating these elements of urban resilience simplifies both diagnosis and analysis for practitioners, they should not lose sight of linkages between them. The most important of these linkages are the institutional links between agents and systems. Local policy decisions, for example, may ignore the tenure claims of marginal groups, deny them access to supportive urban services, or subject them to involuntary climate risks while other groups are protected (Lebel et al., 2006; Moser & Satterthwaite, 2010). Measures to address the resulting ecosystem degradation or climate risk directly are unlikely to succeed without recognition of the underlying injustices.

Agent capacities can play an important role in system characteristics as well, as resourcefulness and responsiveness of key organizations can directly influence investment and operations of infrastructure systems, and lack of assets is a contributing factor to ecosystem degradation (Tyler, 2006).

These linkages may be difficult to assess because knowledge of these different factors and systems is held by many different actors. This highlights the importance in the planning process of multiple knowledge perspectives, and of meaningful engagement of multiple stakeholders (Berkes, 2007; Lebel et al., 2006; Tompkins & Adger, 2004). The integration of different factors and the comparison of key areas of urban climate vulnerability rely on participatory and deliberative processes to draw out different perspectives and ensure that stakeholder positions are considered. Social learning should be an essential element of the integration process because of its importance in adaptation and building a shared understanding of complex issues under uncertainty (Huntjens et al., 2012; Parson & Clark, 1995; Pelling, High, Dearing, & Smith, 2008; Tompkins & Adger, 2004).

The nature of vulnerability assessment, which needs to consider not only resilience elements but also climate exposure, suggests that such a process should bring together both global and local information about climate change, as well as information about urban systems (including their linkages beyond city boundaries), key organizations, low-capacity households and the institutions that enable or constrain their behaviour. The process should engage formal and informal knowledge holders and organizations responsible for different dimensions of adaptation action. It should be iterative in recognition of the time needed to build trust and develop responses across multiple scales of activity (Lebel et al., 2006).

One of the effects that can be missed by separating the resilience framework elements for analysis is self-organization. This is an emergent property of complex adaptive systems. An important example of this phenomenon is economic markets, where under a wide range of conditions the interaction of agents, institutions and infrastructure systems results in highly organized and efficient distribution of goods and services (Krugman, 1996). Self-organization and feedback enable a wide range of autonomous responses to climate change and other stresses, but these responses emerge from interactions across all three elements of the framework.

We emphasize the importance of building climate resilience in cities in order to foster adaptation, but we also recognize that resilience is not always a positive feature. Undesirable system conditions may be resilient (or resistant to change, which is not quite the same). Polluted water supplies, authoritarian governments or invasive species provide examples of conditions that are likely to prove both undesirable and possibly resilient (Walker et al., 2002).

2.6. The climate resilience framework – a tool for local planning

This conceptual framework, with three key elements and general components of each, provides a way to think differently about climate adaptation in cities. Instead of focusing on future climate projections and defining uncertainties and climate risks, local planners can address the enabling and service provision role of core infrastructure and ecosystems, together with the capacities of agents and the structure of institutions linking systems and agents, to identify the key factors that affect resilience in their city. By assembling and describing these conceptual features of climate resilience in cities, we provide a framework that can guide planning practitioners.

The premise of the conceptual framework is that, while it is abstract enough to be generalizable in a wide range of different contexts, it can be made operational by interpreting the details of local resilience factors represented by local actors engaged in the planning process. Systems design and performance will be familiar to engineers and managers (with the frequent exception of ecosystems, whose services in cities are often undervalued and unrecognized). Links between systems will usually be apparent to designers, operators or users. Agent capacities can be diagnosed and interpreted using socio-economic data and through engagement of the relevant organizations and social groups, or representative community-based organizations; and examples of institutions related to differential access, decision making and use of information can usually be identified by community organizations or other knowledgeable local actors.

As in any planning approach, issues of cultural reference point and of power relations come into play. But these are not issues specific to climate resilience, and we defer to other treatments of these questions in urban planning to point out that they are important considerations in resilience planning as well. Power relations inevitably constrain decision-making processes and options. However, there are always starting points and feasible actions that can increase resilience, and where institutional manifestations of unequal power are key contributors to vulnerability the framework will help intervenors make these explicit. The framework can also help to identify patterns of vulnerability that cut across social and power divides, enabling the identification of starting points that are likely to have broad support.

As the intention of this framework is to guide practice, a partial test of the approach is whether it can be used in practice for resilience planning.