Urban water security in Africa: The face of climate and development challenges

ABSTRACT

Resilience has been promoted as an important objective for the global development community, in part, as a response to concern about the potential impacts of climate change and related risks. A review of the challenges of achieving water security in urban areas of developing countries suggests that a specific focus on resilience may distract communities from more effective interventions. It would be more useful to support relevant institutions to address current service delivery priorities. This will better enable them to manage future climate change and the challenges that this may bring.

KEYWORDS:

• Water supply
• water resources
• climate change
• resilience
• management

1. Introduction

The building of cities and societies that are resilient, that can continue to meet the needs of their inhabitants in the face of changing circumstances, is important for children, women and men. The formal goal of resilience has become increasingly prominent in development policy (OECD, 2013). However, some commentators believe that it is neither clear what resilience is, in a socio-economic development context, nor how it can or should be promoted (ODI, 2012). This article, based on a presentation made to the ‘Towards Resilient Urban Communities’ conference in Harare in September 2014, suggests that the adoption of resilience as a goal may distract development actors from their more immediate priorities unless they are clear about the risks to be addressed and the strategic direction and sequencing of activities that will best achieve their goals. This danger is illustrated in discussions about strategies for resilience in rapidly growing urban communities in emerging economies.

The current development discourse about resilience has its origins in concerns about how the natural environment would respond to change (Holling, 1973) and still reflects an ecological focus. Since the late 1990s, there have been attempts to link human dimensions to the natural environment by considering ‘social–ecological systems’ (Berkes & Folke, 1998). This has been given additional impetus by more recent efforts to understand strategies for responding, and building resilience, to climate change often under the umbrella concept of adaptive management (Tompkins & Adger, 2004).

It is suggested that, in most sub-Saharan countries, a specific focus on resilience to climate change and similar global environmental challenges should be a low priority for local and national administrations. The proposition is that, rather than elevating ‘resilience’ to be an over-arching development goal and promoting global strategies to address it, the priority should continue to be addressing the more immediate challenges of, for instance, building cities which offer economic opportunities and have the infrastructure to provide the basic services that meet peoples’ needs. This will require the strengthening of both the physical and institutional frameworks of cities. Not only will this approach have immediately beneficial impacts on society but it is more likely to enable the longer term challenges to be addressed, thus building societal resilience.

This proposition is illustrated and supported using examples from different dimensions of the water sector which, in many different ways, reflects the situation in the wider society. These examples are drawn from Zimbabwe, Southern Africa more generally and, where relevant, from broader international experience.

2. The concept of resilience

2.1 Different meanings in different contexts

Before entering into a detailed discussion, it is necessary to be more precise about the terms used. Not only are current concepts of resilience in socio-ecological systems increasingly complex and widely debated but, together with the related concepts of vulnerability and adaptation, they are acknowledged to have different meanings in different disciplinary contexts. This has led to calls for greater efforts to work towards agreement between scientists in different fields on their use (Gallopin, 2006), but this has tended to focus only on specific elements of the social and environmental physical sciences, relevant to global change processes. There has been limited attention to the disciplines directly associated with the implementation of development programmes.

2.2 Resilience in the physical sciences and engineering

In physical terms, resilience describes the ability of a material to rebound from an impact; in materials science, a material's capacity to absorb and release energy when deformed elastically, returning to its original state when unloaded.

A resilient building is therefore one that can withstand an earth tremor, perhaps occasioned by rapid extraction of groundwater, and continue to perform its function without damage that reduces its functionality. A resilient water supply is one that is brought back to previous levels of functioning after an interruption, whether due to technical failure or climatic conditions.

2.3 Individual and social resilience

However, the concept of resilience is applied far more widely than in material sciences and engineering. It can also be considered at the individual level. In psychological terms, resilience has been defined as the ability to adapt positively to adversity (Fletcher & Sarkar, 2013).

In the context of the environmental discussion, social resilience has been defined as the ability of groups or communities to cope with external stresses and disturbances as a result of social, political and environmental change (Adger, 2000). But because the priority has been to address the challenge of global environmental change, approaches to social resilience in this context have been criticised for ignoring issues of power and authority and not answering the rather substantial question: ‘resilience for whom and at what cost to which others?’ (Cote & Nightingale, 2012:485).

This approach has been developed further (Hall & Lamont, 2013), using an explicitly political science perspective. This sees social resilience in terms of the well-being of groups of people, defining it as ‘an outcome in which a group sustain their well-being in the face of challenges to it’. They suggest that social resilience is often the result of political processes through which states moderate the impact of markets. More specifically, they also reject the notion that social resilience involves the return to a prior condition; their fundamental measure is the achievement of well-being, even where that involves significant modifications to social frameworks.

In these terms, a resilient people is a group that can withstand economic and social challenges and shocks and be better off afterwards. For many of its citizens, life in Zimbabwe over the past decade has required a great deal of individual resilience. But although the society has not returned to its previous state (societies cannot do this, not least because the world around them is changing), there are some signs that it is regaining the cohesion, optimism, energy and direction that characterised the country at independence. Such a movement back towards a different but more desirable state, if associated with a better quality of life, would be an indicator of social resilience.

2.4 Resilience in social–ecological systems

The idea that resilience can mean returning to a state that is different but still acceptable lies at the heart of yet another definition, the resilience of socio-ecological systems: ‘Resilience is the capacity of a system to absorb disturbance and reorganise while undergoing change so as to still retain essentially the same function, structure, identity, and feedbacks’ (Walker et al., 2004:2).

It is this definition that has dominated recent theoretical considerations of resilience (as opposed to development practice). It has obvious attractions for environmental scientists and policy-makers because it better describes many environmental systems and the challenges faced in the real world. Specifically, it directs us to consider ways in which a functional environment can be sustained without preserving an existing ecological and social environment in a constant state of equilibrium. This offers an interesting approach to understanding the challenges of global change but it is less obvious that it can be of assistance in devising practical policy responses at a local level.

2.5 Assessing resilience

By virtue of their explicit definitions, physical and engineering resilience can be measured. But it would be methodologically difficult to attempt a direct measurement of social resilience given the inevitable subjectivity on the part of both subjects and observers who would have to consider, for example, whether the current state is more or less desirable than the previous one. One approach might be to use practical and relatively uncontroversial indicators such as peoples’ access to safe water or other determinants of the quality of life. A measure of social resilience could thus be whether, after an economic or political shock or a natural disaster that has disrupted their habitual sources, people are again able to find sufficient, safe water on a reliable basis.

The experience of Harare, which suffered long periods of water shortages and related crises including serious disease outbreaks (ADB, 2010), suggests that this might indeed be a useful measure. Relatively reliable survey data are available, one set showing that between 2000 and 2012 the time taken for households to collect water had increased significantly (and that, by 2012, the time required in Harare was fourth highest of the 21 African cities surveyed); similarly, although starting with a relatively high level of access to safe water, this access had declined steadily (Hopewell & Graham, 2014). From these indicators, it might be concluded that Zimbabwean urban society has not yet recovered from the disruptions and shocks that it has suffered over the past two decades, raising questions about its resilience.

3. Resilience in practice – understanding addressing risk and vulnerability

3.1 Risk and vulnerability

Once a relevant goal or measure has been established, the assessment of resilience and the design of strategies to achieve it require consideration of two further concepts: risk and vulnerability.

Risk is defined in International Standard 31000 (ISO, 2009:section 2.1) as the ‘effect of uncertainty on objectives’. Vulnerability, in turn, is the product of the risk and the extent of exposure to that risk. In a practical context, strategies to increase resilience require the identification of the relevant risks and vulnerabilities and then of actions to mitigate them.

As an example, in the case of a piped water network, there is an obvious risk that a main pipe will break at some point in the future, interrupting water supply. While the exact time and location of the break is uncertain, its probability could be estimated from records of previous such events in similar systems. But the potential impact of that event cannot be determined from the probability alone. This will be a function of the number of people who depend on that supply. Even then, the impact of a single pipe failure will depend on whether there are alternative supplies.

The vulnerability of the community to water supply interruptions may be reduced by building a second pipeline able to supply a part of it. That would reduce the risk of the whole community's supply being cut off entirely. Reducing the number of households that depend on a single source would also reduce overall vulnerability. There remains a smaller risk that multiple pipelines could fail (which does happen, in practice, during extreme events such as floods and earthquakes). So the community would remain vulnerable, although less so. To further reduce vulnerability, alternatives such as the construction of local boreholes could be considered. A community with different types of water source will be more resilient to failures of part of a system than one dependent on just a single source.

3.2 Technical and physical responses to risk and vulnerability

3.2.1 Building-in redundancy to enhance reliability

The examples already presented illustrate approaches that may increase the technical resilience of a system. Engineers have traditionally considered resilience when designing large systems for water or electricity supply, although this was more often described simply as reliability or risk reduction (see Faber & Stewart [2003] for a discussion of structured technical approaches to risk reduction). One approach is to build ‘redundancy’ into systems so that they continue to work even if one component fails.

Instead of reducing the risk of a pipeline failure by specifying a stronger pipe, duplicate supply lines or alternative sources of water are therefore developed. When equipment is specified, a reserve capacity is provided. Rather than a single large pump, two or three smaller pumps are installed; if one fails, there is always one in reserve (see Wagner et al. [1988] for a description of traditional approaches to designing for reliability).

3.2.2 Mitigating design risks by robust assumptions and systematic monitoring

Technical resilience is not determined by the physical design alone but also by the assumptions which underlie that design. Water supply systems are therefore designed to meet the assumed needs of a community. These include, in the case of water supply, the quantity of water that will be used by each household as well as an allowance for system losses (see, for instance, Ashton et al., 2008).

However, a system may still fail to meet its objectives if the proportion of water lost in local distribution exceeds the allowance or if household consumption is higher than anticipated due to internal leaks or uncontrolled usage in the absence of effective metering and billing. If the assumptions made by the designers are wrong or managers fail to control water use and consumption grows faster than expected, the system may fail to meet the needs of the community it serves. This monitoring and response is a key technical function for water managers, and systems in which it is absent will be less resilient.

3.2.3 Recognising and dealing with uncertainty

Factors on the supply side also determine the technical resilience of a system. The threat of climate change has raised questions about the applicability of traditional technical assumptions about the probability of rainfall and the scale of floods and droughts. In the past, these have been estimated by applying sophisticated statistical analysis to historic data. However, the underlying assumption is that climatic phenomena vary around a constant mean value and that average temperatures and rainfall will stay the same over time. It is now suggested that this is no longer applicable and that, due to climate change as well as other factors, we have reached ‘the end of stationarity’ (Stakhiv, 2010). As a consequence, design parameters will require more careful consideration.

3.3 Beyond technical responses

3.3.1 A wider perspective

The challenge of achieving resilience in water supply specifically, and urban systems functioning more generally, cannot be addressed simply through the design and construction of technically resilient systems. A series of other factors must be addressed if resilience is to be achieved, assessed in terms of a reliable supply that meets the needs of its users. Many of these depend on the effective management of the system once implemented which is, in turn, impacted upon by broader societal forces.

3.3.2 Economic and financial considerations

Building a resilient water supply, like building a resilient society, is thus much more than a technical challenge. For a start, unless a community, settlement, city or country has an adequate economic base, it is unlikely to be able to manage adequately the risks associated with an uncertain natural resource.

Particularly in large urban areas, where self-sufficiency of households or local communities is rarely a feasible option, the availability and allocation of adequate financial resources determines whether a society can build the infrastructure – physical and institutional – that it needs for its resilience. If the society can only afford a system that provides enough water in normal years, it will not be able to meet the needs in a year of drought.

Most of sub-Saharan Africa has enough water, in aggregate, to meet societal needs. However, the concentration of populations in urban areas requires the construction of infrastructure to capture, store and transport the water to where it is needed. A review of the impact of rainfall and river flow variability and its implications for economic development concluded that: ‘Several of the nations with the greatest need for resilience to rainfall variability are among the poorest in the world, and therefore do not have the financial resources to take the necessary measures’ (Brown & Lall, 2006:315). Also, because they do not have the financial resources to manage water resource variability, poor countries lack the water security needed to underpin economic growth (Brown et al., 2013).

The primary risk in poorer communities and countries is therefore that there will simply not be enough money to address basic needs effectively; it is widely accepted that in most of sub-Saharan Africa, water scarcity is an economic construct rather than a physical reality (IWMI, 2007). This was demonstrated by a review of the financial requirements to meet the Millennium Development Goal target of reducing the proportion of people without safe water, which concluded that ‘ … there is clearly going to be a large gap between current financial flows and the investment estimates. The annual funds going into the sector as a whole would need to roughly double’ (World Water Council, 2003:13).

This diagnosis has been borne out by the limited progress subsequently made in Africa on the goals for water supply. Of the global total of 748 million people without access to improved drinking water in 2012, 325 million (43%) were in sub-Saharan Africa; the proportion of urban dwellers served in the region changed only marginally between 1990 and 2012 (from 15% to 17%). Since urban populations grew far faster, this means that the absolute number of urban dwellers without access to safe water increased significantly (WHO/UNICEF, 2014). The decreasing access to safe water in Harare, noted above, was closely associated with economic decline, a practical illustration of the impact of economic circumstances on social resilience.

Particularly in urban areas, a related problem is that although there is often the economic potential to support investments that would, inter alia, enhance resilience, finance is not available to translate this potential into investment. In part, this is because, even where users can afford to pay the costs, financiers are reluctant to accept the risk inherent in long-term loans of defaults caused by local political action or wider external impacts. The resilience of the financial system (NEF, 2015) then becomes a further determinant of the ability of communities to reduce their risks and increase their resilience.

3.3.3 Institutional considerations

Economic and financial considerations are important determinants of societal resilience. But water supply systems are tested by far more than simply technical and financial issues. Many of these relate to the way in which societies organise their affairs and manage change; in short, to institutional issues.

The quality of planning, to ensure that actions required to sustain supplies are identified and acted upon at the right time, will determine whether the system continues to meet its objectives. So too will the administration of operation, to ensure that consumption is monitored; that appropriate measures to control wasteful water use (tariffs or social pressure) are applied where necessary; and that maintenance is done when leaks are found. Any weakness in this technical administration will reduce the resilience of the supply – and of the community that depends on it.

Beyond these issues, which remain focused on ‘soft’ dimensions of technical systems, there are broader risks and vulnerabilities that, if addressed, may make an even greater contribution to societal resilience. These include security and political risks as well as risks inherent in social cooperation and institutional coordination.

3.3.4 The impact of political, security and administrative risks

Political risk is usually considered as something that impacts on the reliability of commercial or investment partners. But it can also contribute to water-related vulnerability and reduce resilience if political considerations are allowed to override technical advice. Common problems encountered specific to the water sector include external interference in administrative decisions, such as billing and metering or investment planning. These may contribute to weakening the resilience of the systems required to ensure that a community continues to have access to safe water.

Political reluctance to introduce unpopular measures such as water rationing can therefore have serious consequences. In 2010, the South African town of Beaufort West failed to curb water use before its dam ran dry or to repair the boreholes that offered an alternative source. This left its 50 000 inhabitants in crisis. A subsequent analysis found that the problem was not primarily a local drought but a failure of administrative planning and political action (Holloway et al., 2012).

Political risk of a similar kind led to water supply failures for 30 million people in the Brazilian metropolitan areas of Sao Paulo in 2015. Investments to increase the supply were delayed by complex institutional arrangements involving municipal, state and national governments, controlled by different political parties (Zero Hora, 2014). Water rationing was opposed by the national government, which faced elections (Brasil Post, 2014). The resulting supply failures showed how political imperatives can trump technical analysis and reduce the resilience of an entire region.

In Zimbabwe, a pertinent example was provided when government cancelled debts owed by water users to urban municipalities ahead of the 2013 national elections. This is typical of the political risks faced by water supply institutions. It may have been simply a populist ploy or a hostile act to weaken administrations controlled by the opposition party (see Human Rights Watch, 2013; New Zimbabwe News, 2013). Whatever the motivation, it reduced the resilience of urban water supplies specifically, and the cities more generally, by reducing the funds available for operation, maintenance and new investment.

Security-related risks are often considered at the macro scale and clearly, in a society in conflict, normal management processes may be interrupted and infrastructure destroyed. However, in water supply, local security often has tremendous impact. Vandalism and theft is a primary cause of supply interruptions in both urban and rural communities, in many African countries. Private water vendors often sabotage water systems with which they compete. In the Mirera-Karagita settlements of Naivasha, Kenya, steel pipes laid for peri-urban supply were stolen, cut up and used to build carts for private water vendors (personal observation by the author). In many developing countries, the diversion of limited supplies by strong groups using political influence or force is widespread (see WALINET [2013] for further examples) and undermines societal resilience.

The management of water resources and services is often impacted upon by security-related matters or political interventions. A particular problem is interventions to force inappropriate appointments to organisational positions which can rapidly undermine the ability of a water institution to discharge its functions.

Patronage is a general problem in many African countries. South Africa's National Development Plan makes specific reference to political patronage as part of the challenge of building a capable state which it considers essential to achieve its development goals, which include building a resilient society (NPC, 2012). Patronage politics is considered to be a significant cause of supply interruptions. While there are obvious solutions to this problem, such as regulating the competences of managers in key posts (Muller, 2009), the dynamics of local political economies often prevent this.

3.3.5 Cooperation as a source of institutional resilience

A particular challenge in the use of natural resources is the so-called ‘tragedy of the commons’ (Hardin, 1968), the failure of societies to find ways to manage sustainably public goods such as fisheries, forests, grazing and water resources (in many cities, piped water supply networks share many of the characteristics of a common pool resource). The economic argument is that, where it is not possible to prevent members of a community from using a limited natural resource, it is in each individual's interest to use as much as possible while it is still available, even if the resource is eventually destroyed.

For a society dependent on such common natural resources, their sustainable management is a matter of survival and determines the resilience of that society. Nobel prize-winner Professor Elinor Ostrom investigated how this could be achieved. Her field work showed that individuals in small communities often cooperated successfully to manage shared ‘commons’ rather than maximise their individual gain (Ostrom, 2008).

However, Ostrom also warned that organising such cooperation in larger communities was a lot more difficult than in the smaller cases which she studied, where many of the participants knew each other and were able to monitor each other's actions. As communities get larger and societies more complicated, it is no longer possible to rely on good neighbours, acting sensibly, to manage a scarce common resource in a way that maintains their water security and makes societies more resilient

To mitigate the risk of what other authors have termed ‘cooperation failure’, Ostrom (2009) recommended that central authorities, such as national governments, should establish frameworks which would support and guide efforts by users to create effective management mechanisms for their local resources.

3.3.6 Institutional cooperation succeeds in South Africa but fails in Thailand

The relevance of Ostrom's findings has been demonstrated in practice in South Africa (Muller, 2012). The intensively used Vaal River system, which connects a number of rivers and supplies the country's largest urban conglomeration, has successfully met society's needs over many decades. A feature of its management is that major water users of the Vaal river system (municipalities and large industries) work together with government agencies to plan and monitor its operation.

By participating in annual system modelling, they ensure that there is enough water stored in the dams to meet expected demand (and that restrictions are introduced and enforced if needed). The organisations involved are thus able to proactively manage their water risk, strengthening their resilience as well as that of the urban region more generally. This shows how a structured framework can enable a family of different institutions with common interests to work together to share a scarce resource. This is in contrast to the situation in electricity where one large national utility is dominant, with inadequate countervailing powers from users to challenge inappropriate decisions or investment delays. As a consequence, South Africa has suffered a series of power shortages while water supplies have been relatively reliable; the price of electricity spiralled dramatically, beyond the reach of many consumers, while that of water rose only slightly more than inflation, despite substantial investment in system expansion.

The consequences of failing to achieve effective collective management in the water sector were demonstrated by Thailand's 2011 flood disaster, the most expensive in history in terms of damage to insured property (Swiss, 2012) after many industrial estates were submerged. Experts cited ‘Poor governance and coordination of the national and local governments’ (Haraguchi & Upmanu, 2012) as well as ‘social and political involvement in unsystematic and unprepared ways caused confrontation and conflicts on flood operation’ (Koontanakulvong, 2012) as prime causes of the predictable and preventable disaster. Over 70 different public agencies were responsible for different aspects of flood prevention, from monitoring river flows to constructing protection works in different areas and planning land use.

The Thai example highlights the overarching challenge of water resources management, the need for coordinated public action guided by a goal of optimal use rather than the achievement of individual institutional goals. Reviewing the disaster, an insurance company executive framed the risks posed by floods as ‘ … the most frequent to occur and the most complex to model. This is due to its varied phenomena, ever increasing human interventions and climate variability’ (Lloyds, 2012).

The conclusion is that institutional resilience may come not from a single strong organisation but from interaction between a number of different organisations. However, there is a need for a framework within which they have sufficient ‘voice’, channels of communication and capacity to engage. Critically, they must share a common set of high-level goals.

4. Resilience may derive from simply meeting peoples’ needs

4.1 How short-term actions address long-term climate challenges

This review of the risks involved and the responses that are needed to ensure a safe, reliable and resilient water supply provides a context within which approaches to increasing the resilience of cities to climate change can be reviewed.

Climate change does present significant risks to water security in the longer term. However, it is suggested that these are complex and generally cannot be predicted with any great certainty. Further, in the communities most affected, there is usually a lack of capacity to deal with the challenges posed. In addition, in poor societies the overriding priority is the short-term one of achieving access to basic services. As a result, there is often limited commitment to actions that do not address this immediate goal.

However, actions to address short-term priorities are not necessarily incompatible with the broader objective of building urban resilience to longer term climate change. The provision of services requires a range of institutional capacities to be developed. In water provision, these include the capacity for long-term planning as well as the ability to build reliable (and therefore resilient) physical systems. The action of addressing immediate priorities will therefore in itself help to create the conditions in which the longer term challenges can effectively be tackled.

4.2 The values of uncertainty

Most of climate change's possible impacts are projected to occur over a longer term than the other categories of risk that have been identified. This introduces further uncertainty into the already uncertain business of estimating how much water is going to be available in rivers and underground, how extreme and frequent floods will be in the future.

Experts are unable to provide much specific information, beyond stating that the future is uncertain. As one practitioner noted, despite many efforts, current predictions can say little that is reliable about rainfall and river flows; the information available ‘is simply inadequate for most operational and design aspects of water sector decisions, and is not expected to be useful for at least another decade’ (Stakhiv, 2010:9).

The case of Zimbabwe is typical. A high degree of uncertainty is reported about the potential impacts of climate change. River flows are:

projected to decline by up to 40%, with the Zambezi Basin worst affected. At the same time, annual rainfall levels based on the 1961–90 average are projected to decline between 5–20% by 2080 in all of the country's major river basins. (Brown et al., 2012:9)

In Tanzania, a review of different studies (Noel, 2012) found that predictions of the potential impact of climate change on river flows varied from a 20% reduction to a 36% increase in flows. It concluded that ‘ … . coupled with widely-ranging predictions on population growth and the rate of urbanisation, (this) means that a key challenge for Tanzania in terms of planning adaptation strategies will be the high degree of uncertainty about its future climate’ (Noel, 2012:4).

Despite the uncertainty, some useful conclusions can be drawn. For instance, while the worst-case predictions may appear dire, they are not necessarily critical. Withdrawals from the Zambezi River system are currently tiny, 1 or 2% of total flow; there will always be water available for urban use given its high priority. Also, the potential impacts of climate change on water supply are often less serious than more immediate, man-made impacts. Planning scenarios for water supply to the city of Cape Town found that climate change might reduce water resource availability over the 30-year planning period. But a far greater reduction is required to provide (discretionary) flows to protect the environment (DWAF, 2007). If these priorities change, enough water will be available to meet urban needs.

This example shows once again that managing uncertainty is part of the job description of competent water managers as they seek to maintain supplies under conditions of (already quite extreme) short-term climate uncertainty and variability. Their ability to do this is a fundamental determinant of the ability of societies to withstand unexpected shocks and extreme events (Brown & Lall, 2006). While the majority of African cities do not yet have the ability to achieve this limited goal, addressing climate change cannot be considered a priority for the provision of reliable safe water supplies. But focusing on meeting short-term needs will help to develop the capacity to deal with these longer term uncertainties.

4.3 Ignoring short-term needs may weaken long-term efforts

The argument against giving priority to building resilience to climate change does not rest solely on the need to focus on the immediate challenge of water security. Rather, as some of the theorists of adaptive management for resilience have acknowledged, it is necessary to recognise that social preferences will place constraints on their approach (Adger et al., 2009).

Crudely put, it will be difficult to persuade people to devote their time, energies and limited financial resources to preparing for long-term threats while they are still grappling with today's challenges of finding water, food and safe shelter in chaotic, poor cities. Other commentators, more focused on the complexities of actually managing Third World cities, recognise that the goal of resilience cannot be achieved without addressing those prior conditions.

In Dar es Salaam, Kiunsi (2013:321) reports a huge ‘development deficit’. Eighty per cent of the city's four million people live in informal settlements with very limited piped water, sewers, drains and solid waste collection. The last City Master Plan was drawn up in 1979. Institutional challenges are illustrated by the fact that there has been extensive commercial and residential development in areas identified by that Plan as vulnerable to flooding. ‘Addressing this deficit (and building the institutional and financial capacity to do so) is also important for building resilience’.

Also, as Satterthwaite & Dodman (2013:291) point out, ‘ … a large part of the world's population lives in settlements that at present cannot develop resilience because they lack the institutions, technical competence and finance to do so'. It has thus been suggested that strengthening the management of today's climate variability should be seen as a specific contribution to addressing the potentially greater variability induced by future climate change (Muller, 2007).

5. Conclusions

Resilience as a development objective is receiving global attention, driven initially by concerns about environmental conservation in a rapidly changing world but, more recently, by the need to guide societal responses to climate change. Countries are being encouraged to develop adaptation plans to address these challenges and explicitly to build the resilience of their societies to the expected impacts.

It is suggested that caution and a more nuanced approach is needed. This article demonstrates that, for a start, the concept of resilience needs to be clarified since it has a range of relevant different meanings in the context of urban services such as water supply. The examples presented also show that the primary determinant of the kind of societal resilience considered in a climate change context is more often institutional (in the broadest sense) than technical.

It is suggested that interventions to build resilience to current ‘day-to-day’ risks may, paradoxically, often be the best strategy for the promotion of long-term societal resilience. Such interventions will meet immediate needs while also providing a foundation from which to address the longer term challenges that may arise under conditions of climate change. To the extent that efforts to build resilience to climate change distract from this prior focus and reduces its priority, they may actually reduce rather than contribute to urban resilience.

A previous review concluded that: ‘Today's investments in water security should be seen as an explicit part of a coherent longer-term strategy for adaptation that will build a more resilient world in the future’ (Sadoff & Muller, 2009:85). More important, a clear commitment to improving the immediate water security of poor urban communities is also more likely to mobilise the social and political support that will be needed to achieve longer term resilience goals.

Acknowledgements

This article is based on an invited presentation made in September 2014 in Harare at the University of Zimbabwe–Institute of Environmental Studies/UNICEF conference ‘Towards Resilient Urban Communities’.

Disclosure statement

No potential conflict of interest was reported by the authors.