Paper 3 Managing water resources versus managing water technology: prospects for institutional change

Linden Vincent

Irrigation Management Network, Overseas Development Institute, London

Summary: Water mobilisation has been the main focus of most state water institutions, usually for broader objectives in agricultural or energy development. This paper examines the prospects for institutional innovation to achieve improved management of water resources, and generate relevant technical and socio-economic data for planning. Shifts from a technology supply focus to a resource management focus are examined for water supply technology and agricultural technology.

Institutions and objectives

Institutions evolve to achieve certain objectives on behalf of the groups initiating them. Sadly, water resources management is one social action that many feel should take place, yet rarely has an administration evolved specifically for the purpose. In many countries, mobilising water has been a major government focus, but for various objectives. As a result, not only are there multiple organisations responsible for water but these are often linked into (or even subsidiary to) other institutional frameworks and objectives. Institutions for agricultural development and energy development often dominate the administration of water, although rural development institutions may play a major role in countries also committed to agrarian reform. It has often been difficult to develop an institutional framework to support and integrate domestic and industrial water with these other uses. Not only have needs developed later but interests may be based in the private sector located outside bureaucratic representation and control.

Even more than other natural resources, water has to be mobilised and combined with other inputs before benefits can be derived from it. Although supply of water technology should be part of an institution designed to deliver particular benefits, in many cases institutions have evolved only for this purpose, so that technology delivery often become separate to the operation of technology and support of users. Many new water technologies lie outside customary controls, or have been used specifically as a vehicle for land reform which has broken customary water rights and obligations. This has decreased the ability of many indigenous water resources management organisations to continue functioning. In agriculture, assumptions are often made about farming activities and water application techniques after interventions mobilising water. Sadly, these are often unrealistic, either in themselves, or because anticipated support services may not materialise. All these issues affect returns to investments and the 'values' derived from water use, and prospects to change current priorities in water allocation.

Institutional innovation will be necessary if governments are to make effective choices in mobilising and allocating water, and encourage the release of water from low-value to highvalue use. This may involve new actions from existing organisational frameworks, not necessarily new administrative structures. Institutional innovations create differences in access to the benefits delivered by institutions: they should ideally be organised, tested and evolved by controllers and users of the institutions but all too frequently have to be induced. This paper examines the room for manoeuvre of government and non-government organisations to encourage the evolution of more water resources management, and support the empowerment of groups to gain more, or different, benefits from water institutions. It looks first at the changing influence of central government over the organisations implementing institutional objectives in the water field, and then at key areas where an emphasis on water resources management is emerging in organisations once heavily focused on technology delivery. Finally, it looks at the role of data collection and use in promoting a water resources management focus, and in the evolution of institutions.

The changing prospects of governance

To understand options in institutional evolution, it is necessary to remember that government structures reflect very different political options and preferences in bureaucratic and constitutional development, and very different realities for central and local planning. One reason why attempts by donors to encourage water management reform have failed is that donors have wanted to encourage idealised and often inappropriate resource administrations (rarely found even in the west) and property rights inappropriate to local politics and local land and water tenure. Even though it is easy to see weaknesses in an existing institutional framework, it is not so easy to introduce improvements as producers of idealised bureaucratic structure believe.

There have teem several changes in both government policies and donor policies over the last decade which affect options in institutional evolution. Many countries have experienced financial reform restricting public expenditure on the development and operation of large water projects. On the one hand, decreasing financial resources increases bureaucratic competition for financial resources. This makes it less likely that new agencies can develop to deal with major planning and allocation problems. The study by Hirji and Ortolano (1991) in Kenya, highlights the major difficulties in creating any meaningful organisation for monitoring of the environmental impact of water developments, despite increasing donor concern for such studies. In Nigeria, the water authorities that once represented an ideal stereotype for regional water planning have been disbanded.

Where withdrawal of financial support has been rapid across both water supply and broader support services, deterioration in public utilities can be rapid, as has happened in several large irrigation schemes (Woodhouse and Ndiaye, 1991). Where institutional evolution at the centre is blocked, and planning capacity decreasing, the role of sector programmes may increase as sector loans become dominant again, raising important questions about the extent of water development planning scheduled outside the country by lending agencies, rather than within national capacity.

On the other hand, financial reform has induced better performance from many agencies. Under the twin effects of financial constraints and donor pressure, agencies have begun to give smaller projects more support. In irrigation, for example, government departments have been persuaded to offer help where there is existing demand, not simply to try to construct new schemes where 'irrigation potential' has been identified. Rehabilitation has been undertaken to address serious supply deficiencies with the understanding that farmers cannot, and will not, cover operation and maintenance costs unless supplies are adequate and reliable. Agencies have become more amenable to the participation of farmers in the design and implementation of schemes, something for which donors have long applied pressure.

The other change over the 1980s is the move, by donors, away from supporting public programmes through sectoral and rural development organisations to a growing emphasis on promoting the private-sector involvement and self-governing organisations or joint agency-user management in large water projects. For example, in Nepal, although various donors supported irrigation and water supply initiatives through integrated rural development programmes or special public works programmes, most work is now focused within the irrigation department. However, in a recent sectoral loan, the Asian Development Bank said it would no longer support the special office for screening programmes in the Agricultural Development Bank of Nepal anticipating that other organisations could provide this service to government departments (Pradhan and Thapa, 1992).

Indonesia has a special programme to turn over small irrigation schemes to user committees, where administrative procedures and training programmes have been developed through a successful partnership between NGOs and public agencies (Bruns and Soelaiman, 1992). However, one major issue is the replication of this experience in a shorter time frame and at lower costs, especially in countries which do not have the same extent of government and nongovernment organisations, nor the same financial circumstances. Formation of local associations for basic operational and maintenance work is a growing feature in many donor sectoral programmes. However, they are not always successful, especially where the management and cost load on users increases without much extra benefit.

In China, government units once responsible for duties are now encouraged to tender competitively for operation and maintenance contracts in irrigation sectors. There is growing interest in solving scarcity problems within irrigation commands by contracting the management of water at a block basis, with options appearing for private sector activities in operations and maintenance of water supply. However, what is proving problematic is the provision of support services through the private sector. Provision of mechanised services in land preparation and harvesting, and supply of seeds, fertiliser and credit have all been quite problematic (Woodhouse and Ndiaye, 1991). Technical advisory services, either for advising farmers on technology, or assisting state agencies in screening and implementing projects have also proved difficult, especially in areas where infrastructure development is low. NGOs appear to have taken a lead in providing the 'venture' capital that has helped support services develop among farmers. The role that bilateral and international donors can play in helping the private sector development by underwriting certain development costs and risks is still unclear as they have to work through government bodies.

While institutional evolution is taking place to deal with some allocation problems at the local level, institutional development to deal with reallocation at the regional level is proving less successful so far. As Sexton (1990) points out, the problem is not a lack of institutions or data for planning, as is often claimed. The problem is that economic development creates an actual shift in the responsibilities required of institutions. Bureaucracies which have been highly successful in mobilising water now find themselves required to make allocation decisions they are not equipped for, intellectually or politically. A strategy of increasing efficiencies can defer pressures, but eventually efficiency measures become too expensive relative to the value of water. In many countries where irrigation development has been a significant component of national development (and even identity), water resources were initially over-estimated with large quantities of water allocated to quite low-value agricultural output. The political difficulties in reallocating water are serious as many supply organisations have very close structural links with user groups who can block major changes in the structure and duties of institutions.

Delivery and use of new technologies

Many water allocation problems have developed through the rapid expansion of water use across the different sectors. However, very particular problems have emerged both through shifts in the availability of new water supply technologies and new agricultural technologies.

Very few governments have maintained a systematic technology policy whereby specific strategies are developed to use lift or storage technologies, or even monitored. One example of the effects of uncoordinated technology changes is Bangladesh, first to shallow wells and then to deep tube-wells, with the result that much shallow lift technology was unusable long before its economic life was over. The technology shifts also raised many (still unsubstantiated) fears about declining water levels affecting investments and equity effects on farmers (Aaron-Thomas, 1992; Morton, 1989). Most countries put no controls on the type of lift technology available in the markets (although import taxes are almost universal) and farmers can often install overcapacity. Even in countries like India, where some states have supported 'state agroindustries' specifically manufacturing pumps for rural development programmes, the limited range of sizes did mean that many small farmers had over-capacity pumps.

Conjunctive use is an increasingly important focus of attention to improve water availability. However, all too often, the strategy is left to evolve as 'joint' use to solve particular operational and drainage problems, rather than achieve any real optimisation (Vincent and Dempsey, 1991). Various strategies have been identified for commands where both canals and wells are present, but practical action depends on the local geology and available yield from wells, and on local political organisation. For example, experiments in Maharashtra, India, showed that farmers' water officials could not close canals periodically to induce groundwater use in certain periods, whereas this has been achieved in China (Sawant et al., 1991; Sun Fu Wen, 1992).

As water becomes more scarce, one major institutional option often cited is changing the property rights that permit uncontrolled development, specifically through the introduction of licensing. However, in many countries legal traditions and political circumstances make this unfeasible and most countries have looked at the control of water use by other means. It is rights to install the technology that have been emphasised or restricted, not rights to the actual resource in the ground, and in many countries there are few prospects in bringing fundamental reforms restricting water rights. Restrictions on siting often exist within the credit programmes which underlie technology of dissemination. Most countries have looked instead at the means to control water use by other means. Effective energy pricing measures are seen by many as essential. However, as wells become deeper and larger, energy costs become substantial. In one tube well scheme in the Philippines, electricity prices at market value cost farmers equivalent to 20-50% of the dry season yield (Flores and Mejia, 1992), and are still rising. In Tunisia, the government restricts the hours in which electricity is available (Boutiti Raqya, 1991). Although the use of tax incentives as well as fines has been discussed in more developed countries as a means of controlling water use (as it has for fertiliser to limit water pollution) few experiments are operating (Braden and Lovejoy, 1990). Making technology more efficient and less wasteful has actually led farmers to irrigate more land rather than reduce consumption (Boutiti Raqya, 1991).

In the next decade, there may be less emphasis on water supply technology and greater emphasis on realising the benefits from new agricultural technologies in seeds, fertilisers and pesticides. As the limits to water supply and yield ceilings are neared, attention to field-level water management in agriculture must increase. Many are aware of the gaps in research and extension for integrating advice on plant cultivation and water management. Many important issues in water resources allocation and management and efficient use will be debated on farmers' fields, not only in terms of regional water allocations.

Again different concerns emerge in different locations. For example, in Thailand crop diversification out of multiple irrigated crops is seen as an important concern. This reflects problems of inadequate water supply as catchments are increasingly developed or as river regimes change through upstream vegetation change. The Thai agricultural research institutions appear to have launched effective research and extension programmes that integrate irrigated and rainfed crops, or paddy with crops with lower water requirements, experimenting with varieties and crop rotations. In Nepal, an integrated approach is less common despite real gains where it has occurred (Whiteman, 1985). This may be because state interest to promote double cropping under irrigation is intense. In Thailand, institutional innovation in sectorally-based research appears high, improving access by, and relevance to, the small farmer. However in Nepal, despite attempts at inducing innovation through IRDPs as well as in sector programmes, actual innovation in agricultural research and extension is variable across the different farming systems and regions.

Elsewhere, for example in the Philippines, poor irrigation infrastructure and water delivery is seen as a major cause in the gap between theoretical yields and actual yields. Many organisations are being exhorted to ever better performance, making sure farmers co ordinate the supply of water and fertilisers at the optimum for crop growth. Initially the Philippines put major emphasis into contacting and organising irrigators through the use of social organisers but costs and logistics have ended this programme. Now, many irrigation organisations are taking increasing responsibilities as they co ordinate fertiliser as well as water, and monitor production. There is increasing debate on whether these organisations should have marketing and saving facilities also, making them supply and marketing co-operatives in all but name. The ironies of seeing one of the earliest development initiatives repeat itself (hopefully having learned from the experiments of the 1960s) is not lost on development workers. Nor is it lost on local administrators and politicians who form the main constraint to granting greater responsibilities to what are now seen as 'third generation' irrigation associations.

Data use and data collection

It may seem strange to hear information generation discussed in an institutional context. However, institutions have a fundamental influence on data generation: also changes in responsibilities (and funding) for data collection and data use can trigger major changes in institutional structure. Does the social and economic information exist to provide a basis for major water allocation questions in the future: if not, who will collect it?

One of the key changes of the last decade has been increased attention to the collection of socioeconomic information to aid technical design and programme delivery. Most socio-economic information is still collected either alongside specific assistance projects or as short-term contracts for government information, with little long-term commitment to data for monitoring or evaluation. However, there are also many weaknesses in the collection of social information and little research is being done which can assist major water allocation decisions. In many cases, pressures to spend funds available for technical infrastructure force a pace of implementation that rapidly outstrips socio-economic research. This is partly because few organisations employ social scientists on a permanent basis, but also because of the time necessary for many important social investigations. It is not easy for a small village to absorb a large group of social scientists, and real care is necessary in planning for information collection. In locations where considerable secondary data exists, the techniques of participatory rural appraisal have proved of enormous benefit in assessing local requirements for technical interventions and establishing some basic agrarian and institutional features. However, there are still many situations where rapid contact with villages will not yield accurate information, and will not provide adequate information for baseline surveys for monitoring and evaluation purposes. We still need systematic and longerterm social survey work, and many interventions do not support this. ODA has recognised the importance of having social scientists as project leaders as well as team members, but there are still many recognised problems in ensuring social scientists can play their essential role in country programmes.

However, it should be recognised that there are also problems within social research methodology. Michael Cernea (1991) has pointed out that anthropological paradigms and related research methodologies still do not serve development needs that well, and major contributions could be made if anthropologists were less heavily trained in ethnography, took less interest in villages, had a broader base in development studies and took more interest in studying organisations and processes. Social scientists in development get trained more in their work experience than in their courses. Anthropologists do not have a strong reputation for supplying their findings back into a milieu where they can be used, and often prefer to remain critics after the event rather than accept leadership roles. This may be one reason why, even when considerable information is collected, it may not be systematically organised to feed into programme design or monitoring and evaluation needs. While the hiring of suecessive different social consultants often fails to give continuity of information, equally the design and reporting of surveys is often weak.

Meteorological and hydrological information has always been an important focus of data collection, but shares many of the criticisms directed at social science training in development. Field workers have long been aware that there are many problems, not only in the absence or presence of relevant data, but also that there are major issues in developing models and equations to use available data in a meaningful way. However, the problem is that the studies that hydrologists want to perform are often not what is required for practical planning, or scientific results are not disseminated in accessible language.

One of the saddest results of the institutional focus on technology delivery rather than resource management is the poverty of the hydrological and environmental database, and the poor availability of relevant assessment techniques. After over 30 years of interventions in many countries, one must wonder why engineers still have to use very rough empirical techniques. One must also wonder why most hydrological training courses still teach so few of the empirical techniques that have to be used in many regions. Why has the hugh emphasis on construction not generated more systematic and relevant hydrological methods? For example, in Nepal, there are now several manuals on appropriate construction in hill areas related to local needs, but there are no equivalent training programmes for relevant hydrologiml assessment. Can it be because feasibility surveys are still so strongly associated with consultancy companies, or that, until recently, few donors have looked to introduce or develop relevant data collection in programmes which governments could take over? While some promising data-support programmes have evolved, we know little about which activities can best survive the funding crises and even wars that have often affected these programmes, and more accessible documentation would be useful.

The established systems that do exist remain strongly influenced by the donors that have assisted them so that data-sets wax and wane with project inputs. In India, for example, there are massive differences in staffing patterns and data collected across regions and sometimes catchments, reflecting different donor inputs and concerns. Often donors tried to establish an ideal programme of data collection for detailed catchment monitoring, unrelated to specific end needs or the practical funding capacities of countries. In reverse, sometimes the data collection system has become so geared to a certain end result, it is not useful for other purposes. In India, groundwater evaluation procedures are geared to identifying sites where new pump technology can be installed and continue in use even though compilation procedures are now widely recognised as too empirical.

In areas where data collection systems are well established, storage has become a major problem and a lot of material is starting to disintegrate. Both the volume of existing data and increasing financial restrictions mean that a radical review of data networks is feasible and necessary. However, a number of critical institutional and social issues influence the likelihood of changes in data collection systems and responsibilities. Often data collection organisations are under financial pressure - reform is threatening to their very identity. Computerisation is one answer to the data storage problem. However, it centralises available information which often fails to flow back to the field. This not only 'de-skills' the field worker who once analysed the data and removes his/her few creative jobs: it also removes information from the main contact point for the fanner and local administrator. Also, a considerable amount of data (for example, pumping tests and resistivity assessments in groundwater assessment) is not easy to computerise.

Complicated hydrological models may help to demonstrate general surpluses and deficits but these models need to be explained, with real opportunities for users to ask questions. They must also be inter-linked with improved studies of changing demand for water. Although it is now more common to find researchers holding a workshop to explain their results, there is rarely any systematic study of the subsequent use of information by administrators. Many local agencies have avoided the use of external agencies who can do sophisticated models and drilling tests because of their expense and lack of feedback. They prefer to use a combination of simpler, locally documented information, often in conjunction with remote sensing data showing geological and geomorphological features.

Are data analysis systems providing what is needed? Often the answer is 'no'. In groundwater, for example, methods exist to assess general yield parameters and regional aquifer resources, reflecting this major institutional bias in technology delivery. However, we still cannot tell a farmer accurately how the waterlevels in his well will fluctuate across the year, or how to adjust the pumping rate. While there are situations where more data may be necessary, better use of existing data is more important. Preservation of data must be linked with training for key operators in the field, not simply computerisation for easy modelling; and analysis must equally be linked to users. Such changes will raise difficult questions as organisational frameworks change relevant to the financial and political climates of the future.

Conclusions

Institutions for water resource management do not automatically require a single administration; and there is no such thing as an ideal administration, as needs evolve over time. The historical sequence of perceptions of benefits from water explains much of the structure of existing institutional controls over water, while the history of technology delivery, both in the public and private sphere, often explains many of the dilemmas for solving growing water shortages.

Prospects for improving water resources management and allocation are more likely to evolve through interacting organisations, as most countries have neither the political nor financial capacity to introduce an over-arching administration. These prospects depend on institutional innovation that not only encourages better interaction across and within the various bureaucracies involved in water, but also encourages evolution of more self-governed, jointly managed and privately-managed activities. While many governments are aware of financial and political dilemmas in changing bureaucratic capacity, most have unclear objectives in decentralisation and deregulation of existing powers, and in the creation of new responsibilities in water management within these new institutional frameworks. While financial reform may actually give donors greater potential to support effective sectoral programmes, donors can also play a major role in supporting these broader innovations outside the state sector. There appears to be more scope for relevant resource management at the local level in many countries.

While people often think that water scarcity problems are emerging due to increased demand, choice and performance of technology - both in mobilisation and delivery of water, and in related production technologies - have a significant influence on the rate and scale of changing water demand, and future development options. Effective water management must consider not only water technology but also catchment management and research, extension and support to improve water use. In agriculture, this means initiatives that integrate water management and crop growth in ways relevant to local farmers: this could mean crop diversification, or more integrated and inter-disciplinary advisory services, or more integrated support services that have good links with extension. In domestic water supply, hygiene awareness is essential to get the benefits of greater supplies. Monitoring technology choices and improving their delivery of benefits may become a much stronger requirement of these new institutional frameworks in the future.

Issues of water allocation and environmentally sound water development and use will have to be addressed within these interacting organisations. Thus effective and relevant training and data collection to support these new policy objectives and changing organisational structure become even more important. Increasing the awareness and information on socio-economic issues remains essential to create an enabling environment for water resource management. Training of technical staff to increase awareness of social issues and the importance of social survey is important. However, encouragement of relevant research and training for both social scientists and technical staff will be essential to achieve water interventions that are economically and environmentally sound. Relevant information and planning capability will be as, or more, important that conventional technical assistance in the next decade to achieve better management of water resources.

Finally, while new technical assistance and 'action-research' programmes may be necessary to answer these new policy challenges, much can be consolidated and learned from existing experience. Past donor programmes could be studied for institutional comparisons of the ways they have worked, not only their technical outcomes. Equally, studies of institutional evolution in countries where positive changes have taken place might demonstrate much about the forms of donor support that are most effective.

References

AKRON-THOMAS, M. (1992) The Equity Impact of Deep Tubewells: Evidence from the IDA Deep Tubewell II Project. Irrigation Management Network Paper 12, Overseas Development Institute, London.

BOUTITI RAQYA, A. (1991) Study on Rationalising Groundwater Use by Electrification of Shallow Wells. Irrigation Management Network Paper 5, Overseas Development Institute, London.

BRADEN, J. B. and LOVEJOY, S. B. (1990) Agriculture and Water Quality: International Perspectives. Lynne Rienner: Boulder, Colorado and London.

BRUNS, B. and SOELAIMAN, I. (1992) From Practice to Policy: Agency and NGO in Indonesia's Programme to Turn Over Small Irrigation Systems to Farmers. Irrigation Management Network Paper 10, Overseas Development Institute, London.

CERNEA, M. M. (1991) Using Knowledge from Social Science in Development Projects. World Bank Discussion Papers 114.

FLORES, V. S. and MEJIA, A. M. (1992) Farmer-managed groundwater pump irrigation systems - the case of Guimba-Cuyapo Network. Paper presented at the FMIS Regional Workshop Groundwater Farmer-Managed Irrigation Systems and Sustainable Groundwater Management, Dhaka 12-21 May, 1992. IIMI, Sri Lanka.

HIRJI, R. and ORTOLANO, L. (1991) EIA effectiveness and mechanics of control: case studies of water resources in Kenya. Water Resources Development 7(3), 154 167.

MORTON, 1. (1989) Tubewell Irrigation in Bangladesh Irrigation Management Network Paper 89/2d, Overseas Development Institute, London.

PRADHAN, U. and THAPA, G. (1992) Sustainable groundwater development and management in Nepal: major issues confronted by a development bank in Nepal. Paper presented at the FMIS Regional Workshop Groundwater Farmer-Managed Irrigation Systems and Sustainable Groundwater Management, Dhaka, 12-21 May, 1992 IIMI, Sri Lanka.

SAWANT, M. M. et al., (1991) Assessment of Conjunctive Use in Maharashtra Minor Irrigation Systems. Irrigation Management Network Paper 3, Overseas Development Institute, London.

SEXTON, R. (1990) Perspectives on the Middle East Water Crisis: Analysing Water Scarcity Problems in Jordan and Israel. Irrigation Management Network Paper 90/3f, Overseas Development Institute, London.

SUN FU WEN (1992) The development of groundwater resources and the farmer-managed irrigation wells in Shandong Province, China. Paper presented at the FMIS Regional Workshop Groundwater Fanner-Managed Irrigation Systems and Sustainable Groundwater Management, Dhaka, 12-21 May, 1992. IIMI, Sri Lanka.

VINCENT, L. and DEMPSEY, P. (1991) Conjunctive Water Use for Irrigation Good Theory, Poor Practice. Irrigation Management Network Paper 4, Overseas Development Institute, London.

WHITEMAN, P. (1985) The Mountain Environment: an agronomists perspective with a case study from Jumla, Nepal. Mountain Research and Development 5(2), 151-162.

WOODHOUSE, P. and NDIAYE. I. (1991) Structural Adjustment and Irrigated Agriulture in Senegal. Irrigation Management Network Paper 7, Overseas Development Institute, London.

Discussion

Much of the discussion centred around appropriate donor interventions. The speaker emphasised that both donors and governments needs to look at the lower levels of bureaucracy. Also that our analysis of institutions for prescriptive advice is very poor, especially at the middle level between national planning and water users. The role of cooperatives was questioned when experience with them has been so depressing. However, the state is handing over irrigation systems to such groups in many cases, but practical powers must be given to these organisations. The issue of whether organisational structures should match the nature of the water resource - lake, river system etc. was raised, but in reality management skills often do not develop for particular hydrological systems. As institutional conditions change, we have to change hierarchies too, but they should not be superimposed.

Paper 4 Water as an economic resource

J. T. Winpenny

Overseas Development Institute, London

Summary: The growing scarcity of water and the increasing environmental costs entailed by its provision, use and disposal are aggravated by a failure to treat it as an economic commodity. Traditional attitudes, based on supply-orientation and reliance on non-market controls, have contributed to the problem and cannot cope with the problems ahead. Although the obstacles are recognised, it is argued that what is needed is a combination of measures at three levels - the creation of enabling conditions, providing incentives and encouraging markets, and direct project interventions. Various suggestions are made for making the reforms more acceptable to the public.

Water as a commodity

Water may be a gift of nature but it is also a commodity. In the economic sense, this is a good or service exhibiting scarcity; unlimited amounts are not available at zero cost. Even where water is plentiful it has to be treated and distributed, and wastewater has to be safely disposed of. These activities use up economic resources. In many - perhaps most - countries at all stages of development, water is scarce and is rapidly getting scarcer. In many cases, the limits of existing supplies are being approached. More frequently, new supplies are only available at sharply increasing costs. In a recent review of projects in the water sector, the World Bank estimated that the cost of a unit of water from "the next project" was often two to three times the cost of a unit from the "current project" (Bhatia and Falkenmark, 1992). These costs are carried by water utilities and their sponsoring government agencies, the majority of which are in poor financial condition and unable to contemplate the large investments envisaged on current extrapolations: $ 10 billion is currently being invested annually in improving water supply and sanitation in developing countries. Even at current costs it would take five times this amount to provide universal reasonable services by the year 2000. Internal cash generation financed only 10% of the cost of World Bank-funded water projects in 1991 (World Bank, internal estimates).

The environmental costs of growing supply and consumption of water are mounting and in many quarters becoming unacceptable. The construction of dams and reservoirs, inter-basin water conveyances, the depletion of natural lakes, and the drawing-down and contamination of aquifers are all causing serious environmental concern. Many agencies have all but suspended aid for large hydro schemes and major water conveyance projects are coming under the closest scrutiny. Pollution from effluent, wastewater and sewage is another source of environmental cost from water use, in this case an 'externality' forced onto third parties. It amounts to using up the finite assimilative capacity of the aquatic environment, leaving none spare for other users and forcing high cost treatment on the public services. Alongside the mounting cost of supplying growing populations and servicing the rising proportion of urban dwellers is the large and growing backlog of services to the poor. Despite the achievements of the UN's Drinking Water and Sanitation Decade, over 1 billion people still lack access to safe water and 1.8 billion to proper sanitation (World Bank, 1992).

These symptoms all lead to water stress and user conflicts as interested parties contend for limited water supplies or for the preservation of threatened environments.

"..a large irrigation project in India does not operate because water has been diverted to the rapidly growing city of Pune. In China industries are having to reduce their production due to water shortages even though they are surrounded by paddy fields. In California selenium salts leached by irrigation are killing wildlife. [World] Bank irrigation projects in Algeria are now competing with Bank urban water supply projects for the same water, and many proposed irrigation projects and most hydro project proposals are on hold because of environmental concerns” (Rogers, unpublished).

Outmoded approaches to planning and provision

Water is universally scarce yet the majority of societies do not treat it as such. A commodity should be priced so as to cover its cost of production and to ration its use to those placing the highest value on it. Commodities are bought and sold in markets. Private agents are active in supply and distribution. In a well-functioning market, the benefit attached to the use of the marginal unit of the commodity (the last one to be sold) is the same for all consumers. These conditions are evidently not those in which water is supplied and used in most cases. The water sector is typified by supply-oriented provision, reluctance to make active use of pricing, allocation by non-economic means, the persisterce of low-value usage in important sectors, and the minor role played by private enterprise.

In most countries, the automatic response to water stress is to consider supply augmentation. The equivalent of the Hippocratic Oath for water engineers is to promise to meet all reasonable needs for water without question by enlarging and improving supplies. Prices are rarely used to allocate water supplies or to manage demand actively. Water pricing is usually seen purely as an aspect of cost recovery and in many cases (e.g. agriculture) does not even achieve that. The resulting paradox is that an increasingly scarce resource is subsidized, discouraging conservation or the reduction of waste. The average tariff in World Bank-financed water projects - probably a better-than average sample - is only about one-third the average incremental cost of supply. Most authorities respond to scarcity by non-price devices, such as rationing, prohibited uses, exhortation, or cutting-off supplies. Although these can be effective, they can also be costly and inconvenient to users, and do not take account of the relative value of water in different applications.

The benefits from using water typically vary widely from one sector to another as well as within sectors. Variations up to a factor of 10 or more are common in comparing the value of water for different uses within the industrial and agricultural sectors and similar differentials apply in comparing municipal and agricultural use values (Gibbons, 1986; Bhatia and Falkenmark, 1992). This is a sign that total benefit could be increased from reallocating supplies. Another sign of the under-development of markets is the minor role played by private enterprise in bulk supply and distribution. It is no accident that privatisation has made least headway in the water sector and, except in the UK, it has largely taken the form of concessions and management agreements rather than full-blooded ownership.

Why is a scarce resource underpriced and undervalued?

There are a number of reasons why water is not treated as other commodities and why water markets are so poorly developed. Why does water differ from electricity or bread?

· Customary attitudes are a primary obstacle in many cases. The idea that water is a gift of God is cited as an obstacle to pricing in some societies (it is less clear whether a dam, a treatment plant, a distribution pipe, or a tap are also due to divine providence). It is understandable that consumers used to getting free water from traditional, unimproved, sources (wells, streams) may resist paying for improved supplies, just as unmetered consumers resist paying proportionately for metered supplies. Whatever their justification, popular attitudes to water constitute a political hurdle to rational pricing. Because water is universally used, the inflationary effects of introducing pricing also need to be considered.

· The fact that water supply is in many circumstances a natural monopoly explains why public agencies are predominant. Difficulties of regulating private monopolies (e.g. in the UK) also account for the limited progress made by full privatisation. Public agencies can and do behave like private enterprises, but mostly don't. In general, water utilities fall far short of normal commercial behaviour.

· Institutions in the water sector have evolved to enable their employees to be rent seekers. Wholesaling water at below its market value to intermediaries and retailers is profitable to the parties involved (though not to the consumers reliant on vendors, who are amongst the poorest urban dwellers). The gains from monopoly power are shared in various anti-social ways. Transaction cost analysis has also produced insights into why the water sector resists market processes (Nickum and Easter, 1991). In certain towns in Ghana a political party extracts half the revenue from charges on the use of public toilets (Lovei and Whittington, 1991).

· Strong vested interests dependent on cheap water conspire to preserve the status quo. Irrigated agriculture and industries reliant on large volumes of water or cheap hydro-power can exercise great political influence.

· The physical barriers to developing a more integrated water market are often crucial. There may be no practical method of transferring water which is surplus to one sector - or used wastefully - to another which could make more economic use of it. In Beijing, surplus agricultural water would need to be collected from groundwater wells and pumped uphill to the city. This sets a limit on how much could be transferred (East-West Center, 1988).
· The physical barriers to the development of water markets are often underscored by legal obstacles, e.g. the existence of long-term entitlements to cheap water, ambiguity over the ownership of water, etc. The rights of third parties in water transfer cases is another consideration and, indeed, is a necessary part of recognising and 'internalising' environmental concerns into the transaction.

· Shifting water onto a more market-oriented basis entails transitional costs which can be heavy. Metering involves a heavy resource cost, which has to be weighed against expected water savings. Industries may need to spend sizeable amounts on recycling equipment or even introducing an entirely new water-efficient process. In households, campaigns to promote water-efficient devices are costly and time-consuming. Socially, ensuring the transfer of water from one sector to another may be disruptive (e.g. may lead to a decline in irrigated farm communities).

Notwithstanding the force of all the above factors, possibly the fundamental obstacle to treating water as a commodity is a lack of faith in the efficacy of economic instruments. It is widely believed that the price elasticity of demand is simply too low for water pricing to do an effective job in restraining demand and reallocating supplies.

Turning water into an economic resource

The interrelated problems described earlier call for action to recognise water as an economic (namely, scarce) resource, and to promote 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).

Enabling conditions

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.

· Water utility reforms would require them to behave more like commercial undertakings. This will require them to adopt more active pricing, metering and tariff restructuring, improved cost-recovery, and greater self-finanang. This will often entail managerial and organisational reforms. Drawing up corporate plans ('con/rats-plans') with the government has been shown to be useful in some cases. Privatisation is appropriate in some instances, though it can take many forms and full private ownership is an extreme - and rare - variant. The French model of concessions and lease contracts has influenced a number of developing countries, e.g. Morocco, Cote d'Ivoire, Guinea, Thailand and Malaysia. Regulated private companies also operate in Santiago de Chile and Guatemala City (Roth, 1987).

The best intentioned and designed reforms in the water sector will be frustrated if key economic signals prove to be countervailing. The more rational pricing of irrigation water will be negated by artificially high farm-support prices. Penalising wasteful industrial water use by pricing and effluent charges will be nullified by high protection on the output of heavy industry and 'soft' budget constraints enabling parastatals to pass on increased water charges and fines to their sponsoring ministries. Hence in those countries where water is becoming the scarce factor of production, action in the water sector should be consistent with other key economic signals.

Incentives, market creation and non-market inducements

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.

Tariffs: 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%. (Boland, 1991; OECD, 1987; Gibbons, 1986).

A pre-condition of economic tariffs is metering which is not always feasible or sensible. It is also difficult to apply to groundwater users - though there may be good environmental reasons for levying some charge on groundwater extraction.

Pollution charges: Setting economic charges for water may also be the best way of discouraging industrial water pollution by penalising excessive water intake. For example, in two private Indian fertilizer companies of a similar size, the one paying a high price for its municipal water achieved a unit water consumption per tonne of nutrient production only 40% of that in the other company, which depended partly on its own wells and partly on low-priced public supplies (Gupta and Bhatia, Unpublished).

The application of pollution charges proportional to the volume and quality of effluent is more rare, but has been shown to be effective in reducing water intake as well as discharge. In three industries in Sao Paulo, Brazil, the introduction of an effluent charge led within two years to a 40 60% reduction in water consumption (Miglino, 1984).

Water markets: There are various types of water markets. Their common feature is that water can be bought and sold, thus enabling it to find its highest value use. Groundwater markets are long established and widespread in certain parts of the Asian sub-Continent, e.g. Gujerat, Bangladesh. Farmers sell water surplus to their requirements to those in deficit. Surface-water markets exist in some Western states of the USA and Eastern, Australia, principally to transfer water from low-value irrigated farming to urban consumers. Sometimes the transfers are semipermanent arrangements, e.g. the efforts of the Los Angeles MWA to acquire long-term water rights from its agricultural neighbours.

Less well-known are the fledgling industrial water markets in India, whose existence depends on growing supply uncertainty and/or higher costs, together with variations between companies in the feasibility and cost of tertiary treatment of wastewater. In Jamshedpur, India, an integrated steel plant (TISCO) sells water at cost to other Tata firms in the area, as well as to a Birla subsidiary (Bhatia et al., In press).

Water auctions, although unusual, are well-established in parts of Spain and have been tried in Australia. Water banking has also been tried: as a response to the recent drought, the state of California bought up water rights to farmers to hold in reserve for urban and industrial use (and most of the stock was drawn down for these purposes).

Non market inducements: Despite the emphasis in this paper on the value of economic instruments in managing water resources there will always be a role for non-market devices, often 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 shortterm emergencies may be to ration supplies and ban wasteful uses.

In authoritarian societies the combination of prescriptive norms, approximating 'best practice' or reasonable usage in each case, and penal charges for users exceeding these norms, can be effective. In Tianjin, China, for example, 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 (Bhatia et al., In press).

Direct interventions and projects

The lowest 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.

Overcoming the objections; the art of reforming the water sector

It is safe to predict that turning water into a commodity will be controversial and call upon every ounce of a government's goodwill with its citizens. Governments should, however, seek every opportunity to remind their citizens of the alternative - which in many cases will be a grim scenario of growing and eventually disastrous water stress.

The following can facilitate public acceptance of the necessary reforms:

· Exploit complementarities and create synergy between the different elements in a reform programme; create virtuous circles

The profitability of UFW programmes is increased if tariffs are set at realistic levels. Consumers will more readily espouse water-efficient appliances if tariffs are at economic levels. Consumers will be more ready to pay higher tariffs if they see evidence of improved services. Industrialists are more ready to pay pollution fines and charges if the funds are earmarked for visible environmental clean-ups. A vigorous public campaign stressing the value of water and the dire consequences of allowing present trends to continue should form part of any reform programme.

· Create gainers as well as losers

The most effective way of doing this is to promote water markets. Farmers who decide to sell their water instead of using it on low-value crops are doing themselves, as well as society, a favour. There are many interests to be mobilised in favour of pollution penalties and reduced industrial water use: one firm's effluent is someone else's intake.

· Make the reforms socially equitable

In developing countries the poor are among the worst victims of existing systems. They regularly pay prices per unit for their water (from private vendors) many times higher than those paid by wealthier people with their own connections. Any reform that raises charges, improves cost recovery and generates funds for expanding and improving the system promises to be socially equitable, even if charges to piped consumers are raised. The structure of tariffs can further promote distributional goals by offering low 'lifeline' rates for minimum levels of consumption.

· Exploit environmental benefits

The better management of demand will postpone, or even obviate, the need for investment in new supply - with its attendant environmental costs. Reduced water stress will also alleviate user conflicts - between municipalities and farmers, power utilities and fishermen, industrial polluters and recreationists, as well as between nations. A reformed water sector confers environmental, as well as economic and financial, bonuses.

· Enlist private resources

Government has dominated the planning, supply, distribution and disposal of water. This has happened for a mixture of motives, both noble and ignoble. Outright privatisation is not feasible or even desirable in every case, though it is a way of deflecting heavy future costs - and public obloquy - onto the private sector (e.g. the UK). Privatisation stopping short of a transfer of ownership (e.g. concessions, management contracts, contracting out) can combine the advantages of public ownership with private management. The more active use of tariffs and effluent charges will encourage many more firms to invest in conservation and recycling, leaving less of a disposal problem for the public agency. Promoting water markets and raising tariffs decentralises the task of matching demand with supply, and mobilises every party behind solving the problem. Nothing less would seem to suffice.

References

BHATIA, R., CESTTI, R. and WINPENNY, J. (In press) Policies for Water Conservation and Reallocation: Good Practice Cases in Improving Efficiency and Equity. World Bank.

BHATIA, R. and FALKENMARK, M. (1992) Water resource policies and urban poor: innovative thinking and policy imperatives. Paper presented to the Dublin International Conference on Water and the Environment, January 1992.

BOLAND, J. (1991) Legislative and Economic Approaches to Water Demand Management. United Nations, New York.

EAST-WEST CENTER (1988) Water Resources Policy and Management for the Beijing-Tianjin Region. Joint Summary Report of the Chinese and East-West Center study teams. Honolulu, Hawaii.

GIBBONS, D. C. (1986) The Economic Value of Water. Resources for the Future. Washington, DC.

GUPTA, D. B. and BHATIA, R. (Unpublished) Water Conservation through Pricing and Pollution Control: a Case Study of Two Fertilizer Plants in India. Unpublished paper.

LOVEI, L. and WHITTINGTON, D. (1991) Rent Seeking in Water Supply. Discussion Paper, Report INU 85, Infrastructure and Urban Development Department, World Bank, Washington, DC.

MIGLINO, L. C. (1984) Industrial Wastewater management in Metropolitan Sao Paulo. PhD Thesis, Harvard University.

NICKUM, J. and EASTER, W. (1991) The Application of Transaction Cost Economics to Asian-Pacific Metropolitan Water Use Issues. Regional Development Dialogue, 12/4, Winter.

OECD (1987) Pricing of Water Services. Organisation for Economic Co-operation and Development, Paris.

ROGERS, P. (Unpublished) Concept Paper for World Bank Comprehensive Water Resources Management Policy Paper. Harvard University. July 19.

ROTH, G. (1987) The Private Provision of Public Services in Developing Countries. World Bank/Oxford University Press.

WINPENNY, J. T. (1992) Where No Water Is: the Economic Management of a Scarce Resource. A Development Policy Study, Overseas Development Institute, London (For publication late 1992).

WORLD BANK (1992) World Development Report, 1992, Development and the Environment.

Discussion

In response to a question on poverty alleviation, it was said that studies had shown a willingness on the part of the poor to pay for water supply. This seems surprising, but they would save a lot compared to buying from water vendors. So projects should look at wider coverage and fewer public standpipes. Customary attitudes to water supply are exaggerated as an obstacle, particularly by politicians. There are plenty of examples of supposed taboos being broken. It was pointed out that in California and Israel, water is very highly subsidised but very efficiently used. At the other extreme, in Bangladesh, farmers cannot even pay the cost of meters, let alone water. Models of water resources management were discussed; the privatised British water industry was compared with the French system, where ownership is retained but the market opened up; companies bid to manage. This does not throw up the problems of regulation which the UK has acquired. The ancient Roman system was also discussed; but it was suggested that these systems had deleterious environmental effects in the Middle East. It has been observed that subsidised water for farmers is a way of making farmers think that the government cares for them. In discussion it was said that, increasingly, resources are concentrating on the cities; this provides the dynamic for many of the problems which we see (despite the political weight of farmers). Raising the price of water to farmers would create food insecurity. But Egypt, for instance, is wasting precious water on low-productivity outlets. In addition, there was comment on the horrifyingly high percentage of water supply systems now out of use; this shows both lack of consultation and a failure to think through cost recovery.