Prepared for FAO by J.T. Winpenny - Edited to suit this e-mail conference

[W-1] In popular usage, "scarcity" is a situation where there is insufficient water to satisfy normal requirements. However, this commonsense definition is of little use to policy makers and planners. There are degrees of scarcity - absolute, life-threatening, seasonal, temporary, cyclical, etc. Populations with normally high levels of consumption may experience temporary "scarcity" more keenly than other societies, who are accustomed to using much less water. Scarcity often arises because of socio-economic trends having little to do with basic needs. Defining scarcity for policy-making purposes is very difficult.

[W-2] Terms such as water scarcity, shortage and stress are commonly used interchangeably, though have the following specific meanings:

1. water shortage: a dearth, or absolute shortage; low levels of water supply relative to minimum levels necessary for basic needs. Can be measured by annual renewable flows (in cubic metres) per head of population, or its reciprocal, viz. the number of people dependent on each unit of water (e.g. millions of people per cubic kilometre).

2. water scarcity: an imbalance of supply and demand under prevailing institutional arrangements and/or prices; an excess of demand over available supply; a high rate of utilisation compared to available supply, especially if the remaining supply potentials difficult or costly to tap. Because this is a relative concept, it is difficult to capture in single indices. However, current utilisation as a percentage of total available resources can illustrate the scale of the problem and the latitude for policymakers.

3. water stress: the symptoms of water scarcity or shortage, e.g. growing conflict between users and competition for water, declining standards of reliability and service, harvest failures and food insecurity. Difficult to capture in numbers, though a checklist approach is possible (FAO 1995, Ch. 2).

[W-3] One possible starting point is to stipulate a minimum amount of renewable water per head of the population, and to treat countries having less than this as "water short". At levels of internal renewable water availability of less than 1,000 cubic meters per head, FAO regards water as a severe constraint on socio-economic development and environmental protection. It has been estimated that 20 countries will be at or below this level by the year 2000 (FAO, 1995). Most of these are in North Africa and the Middle East. At levels of water availability of less than 2,000 cu.m. per head, water is regarded as a potentially serious constraint, and a major problem in drought years. 40 countries will fall into this category by 2000 (FAO, 1993).

[W-4] The picture changes if account is taken of the level of water use as well as availability. For instance, Tunisia, Algeria, Kenya and Rwanda, all countries with levels of availability of less than 1,000 cu. m., have levels of water use currently below this. Comparing levels of use and availability produces a different, and sometimes surprising, picture of scarcity (Kulshreshtha, 1993).

[W-5] Despite these problems of definition and measurement, the concepts of water shortage and scarcity have a hard kernal of reality for many countries. An annual level of renewable supply of 1,000 cu. m. per head is a useful benchmark for comparing countries. The reciprocal of this, viz. the number of people dependent on a given amount of water, may be more vivid and meaningful for certain purposes. For instance, 1,000 cu. m per head is equivalent to 1 million people reliant on a cubic kilometre annually.

[W-6] Whatever benchmark level is taken, the precise amount has no absolute significance; scarcity is a relative concept and can occur at any level of supply, depending on demand and other circumstances. A number of exercises have been done to define water-scarce countries. The more meaningful are those that allow for the inflow of water from other regions in their estimates of availability, and which take some account of the level of existing consumption or the rate of current utilisation of supplies (Engleman & Leroy, 1995, WRI, 1996, Kulshreshtha, 1993, Falkenmark & Lindh, 1993)

[W-7] The conclusion of the above discussion is that shortage is absolute whereas scarcity is relative. Shortages may or may not be capable of alleviation, e.g. by investment in storage or more efficient systems, but policy-makers are likely to be more limited in their options than in dealing with scarcity, in the sense used earlier in this paper. Scarcity may have its roots in water shortage, but it may also be a social construct, a product of affluence, expectations and customary behaviour. As will be argued below, scarcity has various causes, some of which are capable of being remedied or alleviated. A society confronting water scarcity usually has options. Scarcity is not necessarily inevitable or immutable.

[W-8] Strictly-speaking, most sub-Saharan African countries are classified as "surplus", mainly because water resources are ample, irrigation is not widespread, and usage is low. However, this broad picture conceals some problem areas. The underdevelopment of water infrastructure in many countries means that there are great regional differences between the availability and use of water, especially in the Sahel. The quality of water, especially for villagers and marginal urban populations, is a cause for concern and a public health hazard. There are signs that the balance of supply and demand is deteriorating and becoming critical in Kenya and some Southern African countries (e.g. Zimbabwe). South Africa is a special case, containing large areas of arid and semi-arid land, with most water under private ownership and control, and rapidly growing demand from a large urban population with higher expectations than in the past. inter-sectoral conflicts are becoming acute. Several island countries (Cape Verde, Madagascar, Mauritius) are already categorised as "marginal".

[W-9] Some causes of water scarcity are natural, others are of human agency. The impact of natural processes can be aggravated by human responses. Human behaviour can modify our physical environment in a way that makes useful water more scarce. The demand for water may be artificially stimulated, so that a given supply becomes "scarce". The brief discussion below illustrates these points.

[W-10] Growth in population and incomes: The world's population is growing rapidly - in 2020 it is projected to be 7.9 billion, 50% larger than in 1990 (Dyson, 1996). Most of this growth will be in countries the inhabitants of which have low levels of household water consumption, and in which the use of water-intensive appliances is likely to grow. Many of these countries are also rapidly urbanising, and the task of obtaining sufficient water and distributing it to the newly urbanised households will be a major financial and environmental challenge to many authorities. However, the growth in the demand for food is probably the single most important cause of pressure on water resources: over the last generation, most of the increment in food supply has been obtained by an expansion in irrigated farming and/or the growing use of crops dependent on agronomic packages based on irrigation. Looking ahead 20-30 years, in many regions, including the Far East and South Asia, most of the required increase in food production is expected to come from irrigated agriculture (Dyson, 1996). Although the growth of population and demand for food is the predominant factor in the growing scarcity of water, other factors, reviewed below, also contribute.

[W-11] Climatic change and variability: On one definition, a "desert" receives less than 100 mm of rain annually, while a "dryland" includes arid and semi-arid areas receiving annual rainfall of less than 1500 mm. Around half a billion people live in such areas, where water is an unavoidable constraint on everything they do (Winpenny, 1991, p. 21). The influence of climatic change on the availability of water is the subject of intense debate. There is a majority view - though not scientific unanimity - that global warming is occurring. If it persists, climatic zones are likely to migrate, leaving the climate of some regions dryer, others wetter, and all more variable and unpredictable (Parry, 1992). Certain regions dependent on water (e.g. major farming areas, or large population centres) will experience more water scarcity, while others will become more humid. It is an open question what the net effect on water supply will be, but in any case there will be transitional and frictional costs in regions that become drier.

[W-12] Modifications to landscapes and land use. The degradation and land use conversion of watersheds and catchments may reduce the amount of usable water available downstream, if there is greater run-off (e.g. temporary floods) which can not be captured. These same processes can reduce existing water storage capacity, e.g. by silting reservoirs. Major land use changes such as large scale deforestation, or the draining of large wetlands, may also induce microclimatic changes, leading to lower humidity. Localised desertification may result from unsuitable farming and animal husbandry. Rivers and lakes may be deprived of their usual flows of water by storage and diversion projects, leaving their hinterland starved of water. In some cases, these practices merely redistribute scarcity from one locality to another. Elsewhere, the damage to the hydrological cycle leaves everyone the poorer.

[W-13] Contamination of existing water supplies: Pollution of normal water supplies effectively destroys part of the water resource, and forces its users to turn elsewhere or reduce their consumption. This may happen to surface supplies (e.g. a river or lake used for drinking water or washing) or groundwater, and the pollution may be from industrial effluent, agro-chemical run-off from fields, the casual disposal of human excreta, or the release of insufficiently treated sewage from municipal works. However, stating the problem points to the solution - reducing water pollution can increase usable water supply.

[W-14] A failure to manage demand: Many instances of water "scarcity" are artificially created. Economics predicts that a useful commodity, such as water, which is offered free or at a low price will inevitably become "scarce" in the sense that demand will outstrip supply at the prevailing price. In many domains water is available free or at a price well below its true cost of production, and in that sense water providers have created the scarcity they grapple with today. They are literally subsidising the consumption of a good which is growing scarcer. Water-intensive sectors, such as irrigated farming and heavy industry, are being created in the often erroneous belief that ample water supplies are available.

[W-15] Financial and institutional obstacles: In many cases a country's water potential is not being realised because of financial shortages and institutional failures. Water is potentially available, but is not being fully captured because of the way in which water provision is organised and managed. Many water authorities are short of funds to invest in improving and expanding their systems, or even to maintain and operate their existing ones. They typically fail to recover their full costs, and to collect all of what is due to them. Many publicly-owned water institutions are inefficient and some are corrupt. Their capacity to implement cost-effective provision in tune with public demand is often lacking. The water they have at their disposal tends to be mal-distributed, favouring old-established customers, who tend to be the more affluent households and industries with good political connections. Waste and leakage is at a high level in many systems.

[W-16] The main point to highlight in this section is that "scarcity" is not just - or even primarily - an inevitable natural phenomenon, but is heavily influenced by human behaviour, social customs and institutions, and government policies.

[W-17] The impact of water scarcity (drought) is different, depending on whether it is chronic and long term, periodic and unpredictable, or regional and local. Societies and countries in arid and semi-arid regions have learned to accustom themselves to a permanent shortage of water, and have evolved ways of coping with this. In contrast, countries that have developed structures, modes of production, technologies and life-styles reliant on water will experience the shock of unexpected shortage more keenly: "...some of the relatively more developed or complex Sub-Saharan African economies, such as Senegal, Zambia and Zimbabwe, may be more vulnerable economically to drought shocks than least developed or more arid countries, such as Burkina Faso.." (ODI, 1996, p. 51).

[W-18] Affluent societies in dry regions may be able to secure their hydrological future by heavy spending on long distance water transfer. This is the basis of the growth of the cities in Southern California, New Mexico, Colorado, and - in a different region - Tripoli. However, such expensive solutions are not open to all. A recent serious drought in Southern Africa caused the city authorities of Bulawayo to close down the local power station, and to prepare emergency plans for closing and relocating the city's larger industries. In the drought-hit county of Yorkshire in the UK, the water company has also had to prepare drastic, expensive and unpopular plans for rationing, supplementing and redistributing its limited supplies. Since agriculture takes such a small share, industry and public services tend to bear the brunt of cuts.

[W-19] Historically, droughts have caused famines, but these are fortunately becoming rarer: "..the frequency and demographic impact of famines has declined greatly during recent decades. Modern food crises have had a diminishing and increasingly short-lived impact on population growth. The phenomenon of famine has become more and more restricted to sub-Saharan Africa. And even there, the chances of someone dying in a famine have been described as 'vanishingly small'" (Dyson, 1996, p. 75).

[W-20] Latterly, the worst food shortages have occurred in the context of civil war and insurrection, and have been caused by the destruction of crops and livelihoods, dispossession and migration, and difficulties in getting relief to victims. Food security complements water security: such measures as the diversification of food production and the storage of surplus food help to insure against a dry season. India, for instance, has greatly reduced the risk of future famine by these methods. Stored food is "virtual" stored water.

[W-21] Water scarcity, whatever its dimension or origins, makes societies feel vulnerable, hence the quest for greater water security. The essence of water security is that societies should have sufficient access to water, or that they should have the means to limit the damage caused by shortages.

[W-22] Water insecurity has various aspects. At its most basic level, drought can threaten the lives and the livelihoods of people. Although cases of people literally dying of thirst are, thankfully, rare, there are more frequent cases of famine caused by the loss of peoples' livelihoods, namely their cattle, their crops and common property resources they depend on. At one stage removed is the insecurity of important economic sectors dependent on water, such as irrigated agriculture, power, industry, or tourism. There is also the insecurity posed to the environment from water shortages, for instance, the loss of rare or beautiful habitats and the death of wildlife. Finally, the loss of water security can be costly and inconvenient to users who have built their lifestyles, households and working environment around ample water use.

[W-23] For these reasons, governments are concerned to improve national water security. Some of the motives and policy issues are similar to those surrounding national food security. Security, in the abovementioned sense of access to water, can normally be improved - at a cost. This cost may be in economic, social, political or diplomatic terms. Trade-offs need to be made.

[W-24] For instance, more of a country's renewable water resources could be tapped through investments in river diversions, dams, or groundwater pumping schemes. Or investments could be made in more water-efficient processes or irrigation equipment, or recycling and reuse. In these cases, security has an economic cost. An alternative is to persuade or compel people to change their attitude towards water and the way they use it, e.g. by metering its consumption, or banning its use for gardening. Security in these cases has social and political costs. Another option is to seek better international access to water, by importing it or by increasing a country's agreed share of common waters. In this case, improved international access to water is bought at economic, political and/or diplomatic cost.

[W-25] An increase in water security will normally improve the prospects for food security. However, the converse is not true: increasing food security by the expansion of the irrigated area will, unless it is accompanied by improved supply or management, reduce the security of other water-using sectors, and aggravate water stress.

[W-26] Food security is a state in which all households have the physical and economic access to adequate food for all their members, and are not at undue risk of losing such access. The ultimate determinant of food security is the purchasing power of the individual or family. At the national level, food security is usually pursued through a policy of self-reliance in food, in the broad sense of a combination of domestic supplies and the opportunities afforded by international trade.

[W-27] Between 30 and 40 percent of the world's food comes from the irrigated 17 percent of the total cultivated land. In addition, around one-fifth of the total value of fish production comes from freshwater aquaculture. Water development for food production thus constitutes an important element to increase food security and the security and stability in food supplies in next century will increasingly come from improved water management: rehabilitation of the inefficient water control systems and substitution of traditional irrigation systems stemming from a past era of plenty for systems based on accurate technology. Achieving this will require funds, qualified, capable farmers and managers, and politically strong administrations.

[W-28] In many countries situated in the arid regions, with limited potential in rainfed agriculture, water stress is already above the level that would allow enough food to be grown locally to feed the population. In some such countries, national policies call for enough food to be produced locally to protect against the contingency of inability to import food at any cost, as in time of water or trade embargo. Such local-production policies have been fostered through trade embargoes for political positions on water scarce countries, resulting in some cases in the exploitation of fossil water - a not renewable resource - for the production of low-value food crops. Generally, such policies result in high-cost domestic food production, thus decreasing the food security of the lower income sector of the population. National anxiety about enough water for food production is among the driving forces behind so-called "water wars".

[W-29] Food self-reliance requires an economy that generates enough exports to cover the cost of food imports in order to meet the food needs of the population. It also requires that i some places in the world the soil moisture, stemming from rainfall and irrigation, is available to grow enough food for all of the population in the world. In the context of food self-reliance, water security is obtained through a policy of social and economic development and rational, sustainable use of limited water resources, with the aim of meeting the needs for domestic and urban use, as well as of commerce, tourism and industry, so as to have the ability to supply employment for the population. Reliance on trade for food security has some perceived risks, such as deteriorating terms of exchange on world markets, uncertainty of supplies and price instability, and potential embargoes on food trade. Food and water security in water-stressed countries are thus closely tied to the solidity of the trading position anchored in a context of regional stability and collective security.

[W-30] A basic distinction can be drawn between approaches which are supply-oriented and those which rely on demand management. The differences between the two approaches can be exaggerated, and some solutions overlap the two categories (e.g. household leak detection, industrial recycling), Nevertheless, the distinction is a useful way of organising policy discussions. The main policy measures in supply and demand management are summarised below [Source: Winpenny, 1994].

[W-31] Summarising the main policy measures in a supply-oriented water management approach:

• surface water capture and storage
• long distance conveyance & inter-basin transfer
• groundwater exploitation
• watershed management
• conjunctive use of surface- and ground-water
• dual quality water standards
• desalination
• other non-conventional solutions
• pollution control
• new water-sharing agreements

[W-32] Summarising the main policy measures in a demand management approach:

I. Enabling conditions:

• institutional and legal changes
• utility reforms
• privatisation
• macroeconomic and sectoral policy

II. Incentives

Market based:

• active use of water tariffs
• pollution charges
• groundwater markets
• surface water markets
• auctions
• water banking

• restrictions
• quotas, norms, licences
• exhortations, public information

III. Direct interventions and programmes

• canal lining
• leak detection
• water-efficient user appliances
• industrial recycling, reuse,
• water efficiency

[W-33] Each country has unique options: its choice will depend on its level of development, hydrological situation, political and social institutions, management skills, financial resources, popular attitudes to water, and other factors. It is therefore unsafe to make general recommendations. On the other hand, it is possible to recognise certain general categories in which countries find themselves, which will shape their response to water problems.

[W-34] A comparison of water needs with its availability per head has led countries to be categorised, for the purposes of policy-making, into five groups, according to their level of demand and utilisation rates. So-called alpha countries have high levels of demand and high rates of utilisation ("mobilisation"). Their main future options are for demand management and for the avoidance of water pollution. Centralised management and allocation is likely to be called for.

[W-35] Beta countries have a high level of present demand, but a low rate of utilisation of available supplies. Their options will involve further development and modernisation of resources, and evening out regional imbalances. Gamma countries have low current demand levels, but high utilisation rates, and confront serious challenges as population grows. They need to implement policies on all fronts - supply augmentation, storage, demand management, reduction of pollution, recycling, etc.

[W-36] Delta countries have low demand levels and low rates of utilisation, and the greatest degree of freedom in managing future growth. There is a fifth category, the epsilon country, which is already overconsuming its renewable resources, usually by extracting fossil groundwater. The countries and regions concerned, many of which are affluent, face expensive alternative options to restore sustainability (Falkenmark & Lindh, 1993).

[W-37] The balance between supply-oriented and demand management measures, as well as the choice amongst options in each category, will be country-specific. For consistency, it is desirable that the different options are ranked and compared using common criteria, which are set out in the final section. The next two sections briefly review the main policy options summarised in the boxes.

[W-38] Although supplies can be obtained without major structures, e.g. by run-of-the-river schemes, this option normally involves erecting some structure to dam or divert a river, and store water in a natural lake or artificial reservoir. Such schemes can be multi-purpose, with supplementary benefits from power generation, flood control, navigation control, and recreation. However, most of the cheapest, simplest and most accessible projects of this kind have already been undertaken. Subsequent projects tend to have a sharply escalating incremental cost. There are also mounting environmental objections to this kind of project in many countries. The ideal, held by many engineers, of reducing river discharge into the sea to zero, is abhorrent to environmentalists. However, these arguments have to be balanced against the value of storing water from one year to another as insurance against drought.

[W-39] The river basin is often held up as a natural geographical area for planning water supplies, but the reality is that many cities and regions have outgrown this conceptual entity. Most major world cities now supplement their water supplies from other river basins and catchment areas. This often involves transfers over hundreds of miles, entailing major outlays on canals or pipelines, and the energy cost of pumping. Long distance conveyance may be inevitable, but is expensive, entails friction and conflicts with the region of origin, and faces environmental objections.

[W-40] Groundwater is not likely to be a major untapped source in most countries. It is typically close to its exploitable limit, and practically all major urban centres, and many rural areas, are running groundwater overdrafts (in the sense that extraction exceeds the natural recharge rate) (Frederiksen, 1996). The artificial recharge of aquifers with seasonally surplus water may be appropriate in some circumstances, but tends to be costly. Irrigation schemes, and leakages from their conveyance channels, are in practice the largest source of recharge, but this is often an unplanned result.

[W-41] The management of watershed and catchment slopes can regulate the flow of water in the interests of downstream users. There are many documented cases of municipal and irrigation water schemes running out of water because of deforestation and erosion in their catchment areas. Apart from water regulation, watershed management also has potential farming, forestry and environmental benefits (FAO, 1987). It should, however, be borne in mind that afforestation of watersheds is likely to reduce the net yield of water available to downstream users, though it would even out the flows.

[W-42] Studies of alluvial basins illustrate the potential gains available from the conjunctive use and management of ground- and surface water (O'Mara, 1992?). Groundwater can be used to supplement surface flows, when the latter are seasonally low. In the wet season, when surface flows recover, surplus water can be released into the aquifer to recharge it. The use of tubewells can also assist drainage, when otherwise the soil would become waterlogged or saline. In practice, conjunctive use requires the ability to transfer water between various parties, which may require the definition of rights, and calls for an active management role for public authorities. These are exacting preconditions (Frederiksen, 1996).

[W-43] It is wasteful and costly to treat all water to the standards required for drinking purposes. In developed countries gardens are watered and cars cleaned and toilets flushed with water that has been tested, treated and purified to a standard totally unnecessary for these purposes. This may commit resources which could have been used to develop new supplies, and hence contribute indirectly to scarcity. More generally, the application of standards which are unnecessarily high rules out water sources which could ease scarcity. The problem is the cost of installing and maintaining separate systems to deliver two kinds of water, especially for household use. In practice, industry and agriculture tend to tap water sources yielding a quality appropriate to their needs. The use of treated and recycled municipal wastewater in agriculture and industrial cooling is gaining ground.

[W-44] Affluent communities in arid and semi-arid regions lacking other alternatives are increasingly treating saltwater and brackish water for household use. Various technologies are available, and their commercial viability depends partly on their capital cost, and, crucially, on their source and cost of energy. Energy that comes as a by-product of power stations, waste heat from smelters, exhaust steam from nuclear power stations, burning refuse, etc. is of particular value for this purpose. Desalination is so far largely used in Middle Eastern oil producing countries, in isolated communities, and where demand is mainly from households and the tourism industry.

[W-45] Many ingenious proposals have been made for overcoming water scarcity, e.g. towing icebergs and melting them near water consumers. Many of these ideas are not yet feasible. The transport of water by sea tanker is, however, feasible for relatively small quantities, e.g. for islands without substantial water resources of their own. However, it is unsuitable for low value applications such as most agricultural crops.

[W-46] Reducing the pollution of water sources can serve several purposes: avoiding the loss of current sources, saving the cost of developing alternative supplies, and opening up previous sources once they have been cleaned and decontaminated. The link between contamination and supply is most direct in islands dependent on their freshwater "lens", which is easily contaminated by wastewater percolation and saline intrusion. However, there are many other circumstances where reduced pollution is a solution to supply, as well as conferring benefits to amenity, public health and the environment.

[W-47] Many countries depend on international rivers for their water supply. In most cases, the share of the water between the interested parties is subject to agreements. These treaties may have the force of international law, or they be less formal deals enforced by general diplomatic, political or economic means. It may be open to a country to seek to increase its agreed share of international water. The cost of an increased share might arise in financial, economic, diplomatic (or, at the extreme, military) terms, and the country may judge that it is still worthwhile to pursue this route rather than develop other sources of supply. Where the waters in question are already fully utilised, there is an element of "beggar your neighbour" in this approach, though it may be economically justified if one country uses the water more productively than the others. Where the waters are not fully used by the other parties, one country's larger share may not be at the others' immediate expense.

[W-48] Demand management starts from a recognition of water as an economic (viz. scarce) resource, and aims to optimise the use of existing supplies. In countries with a large irrigated farm sector, making more efficient use of irrigation water is of central importance. However, because other sectors compete for the use of the same resource, efficiency gains in industry and the household sectors can bring mutual benefits. Demand management includes market-oriented reforms in the way it is supplied, used and disposed of. Action is called for at three levels: the creation of enabling conditions; setting incentives and promoting markets; and direct interventions and spending programmes (Winpenny, 1992).

[W-49] In this context, the Enabling Environment comprises institutional and legal changes, the reform and privatisation of utilities, and sector-wide economic policies. Legal reforms may be necessary to remove ambiguities over the ownership of water, and the conditions under which it can be transferred. Planning for the development of water resources:

- usually dominated by the projection of fixed "requirements" and the inevitable rise of "gaps"

- needs to adapt to demand management, and allow for the operation of markets and - even- the entry of private operators.

[W-50] Policies towards agriculture will be critical, in view of the heavy water consumption of this sector. Decisions to open up new irrigated areas, and to allocate existing water for irrigation projects, need to take a realistic view of future water resources and the growing demands on them from other sectors, otherwise the resulting schemes will be uneconomic and unsustainable. The understandable pursuit of national food security has to be judged in this light.

[W-51] The permissive effects of enabling conditions may be sharpened by the creation of incentives for the more rational use of water. These may be positive or negative, market or non-market. They will be categorized below as: tariffs; pollution charges; water markets; and non-market inducements. Although water tariffs are in widespread use in countries at all stages of development, they are usually seen as a means of cost recovery rather than a way of actively managing demand. The principles of economic tariff setting are well established and accepted, and are similar to those in use in the power sector. They can be summarized as setting prices according to Long Run Marginal Costs. This usually entails adjusting the structure of tariffs to include a fixed and variable element, with the latter rising for successive increments ("progressivity"). There is evidence of enough elasticity of demand in the household sector to make tariffs an effective instrument for water demand management. A consensus is emerging from a variety of empirical studies that the price elasticity of demand for water by households falls in the range -0.3 to -0.7, implying that a 10% increase in prices leads to a fall in demand of between 3% and 7%.

[W-52] Demand management can, and often does, rely on non-market devices, either on their own, or working in tandem with economic measures. Education and publicity campaigns can help to convert the public to the need for water conservation, though the message will be powerfully underlined by the use of tariffs. In water pollution, some contaminants are so dangerous that they should be banned - pollution charges are not enough. The only feasible response to short term emergencies may be to ration supplies and ban wasteful uses.

[W-53] Some societies have successfully combined prescriptive norms, approximating "best practice" or reasonable usage in each case, and penal charges for users exceeding these norms, can be effective. For instance, in Tianjin, China, norms are set for industrial consumers based on regular detailed water audits, and users who exceed their quotas pay a penal water charge of up to 50% the normal level.

[W-54] The third level of the tier consists of direct intervention by the government or water utility to bring about the necessary conservation without further ado. This category of measure usually entails public spending. and absorbs administrative and technical resources. Many of these interventions can be thought of as projects. Examples include canal lining, programmes to reduce unaccounted-for-water (UFW), and the dissemination of improved household appliances. For public enterprises, the analogous action would be spending on recycling equipment. In practice, a large number of technological options for increased water efficiency are available for consideration (Mei Xie et. al., 1993).

[W-55] It is desirable for options, whether supply- or demand-oriented, to be assessed in a systematic way, if possible using common, agreed criteria. Supply-side options can be appraised using well-tried techniques of cost-benefit analysis (CBA) and/or cost-effectiveness analysis (CEA). Many schemes are justified by demonstrating least-cost methods of meeting project demand levels deemed to be fixed and unavoidable. Appraisals based on estimates of benefits can, in theory use a full range of estimators (Gibbons, 1986), though in practice tend to rely on estimates of willingness-to-pay, sometimes supplemented by quantified public health benefits.

[W-56] Demand-management policies (apart from system improvements that leave the level of service unaffected) tend to involve a reduction of consumption and thus in the level of consumer welfare. The proper assessment of these policies entails predicting consumers' response and placing economic weights on these changes. In theory the effects of a demand reduction are symmetrical to those of an increase, and can be appraised by analysing changes in consumers' surplus. Another common economic criterion is the cost of conserving water by the various methods (Winpenny, 1994). The economic criterion is not the only one relevant to choosing the best ways of planning and managing water systems.

[W-57] Other appropriate criteria are the following:

• efficacy
• distributional impact
• environmental impact
• fiscal effects
• political and public acceptability
• sustainability
• administrative feasibility

[W-58] In particular circumstances other criteria may also be relevant, e.g. impact on food self-sufficiency, regional development, the urban-rural balance, etc.

Allan, Tony, 1992: "Fortunately there are substitutes for water: otherwise our hydropolitical futures would be impossible", in Priorities for water resources allocation and management. Overseas Development Administration, London

Bhatia, Ramesh, Cestti, Rita & Winpenny, James, 1995: Water conservation and reallocation: best practice cases in improving economic efficiency and environmental quality. World Bank/ODI.

Dyson, Tim, 1996: Population and food: global trends and future prospects. Routledge, London.

Engelman, Robert & LeRoy, Pamela, 1995: Sustaining Water: an update. Population Action International, Washington DC.

Falkenmark, Malin and Lindh, Gunnar, 1993: "Water and economic development", in Peter H. Gleick (ed.) Water in crisis: a guide to the world's fresh water resources. Oxford University Press, 1993

FAO, 1993: The State of Food and Agriculture 1993. Rome

Fisher, Franklin, 1995: "The economics of water dispute resolution, project evaluation and management: an application to the Middle East" Water Resources Development, Vol. 11, No 4.

Frederiksen, Harald D., 1996: "Water crisis in developing world: misconceptions about solutions". Journal of Water Resources Planning and Management, March/April.

Kulshreshtha, S.N. 1993: World water resources and regional vulnerability: impact of future changes. International Institute for Applied Systems Analysis, Laxenburg, Austria

Mei Xie; Kuffner, Ulrich; & Le Moigne, Guy (1993): Using water efficiently: technological options. World Bank Technical Paper No 205. Washington, DC

Overseas Development Institute, 1996: The impact of drought on sub-Saharan Africa. London

Parry, Martin, 1990: Climate change and world agriculture. Earthscan, London

Serageldin, Ismail, 1994: Water supply, sanitation and environmental sustainability. World Bank, Washington, D.C.

Winpenny, J.T., 1991: Values for the environment: a guide to economic appraisal. HMSO, London.

Winpenny, James, 1994: Managing water as an economic resource. ODI/Routledge.

World Resources Institute (WRI), 1996: World resources: a guide to the global environment, 1996/7. WRI, Washington, DC.