Return to the Content of WRM course
Managing Water as an Economic Good: RULES for reformers
Running head: "Rules for Reformers"
The World Bank, Washington DC, USA
This paper extends a previously-developed framework for the management of water as an economic resource, by showing how positive and negative externalities can be taken into account. The main focus of the paper, however, is on assessing the lessons of experience which emerge from successful reforms. The following emerge as a tentative set of "rules for would-be reformers":
economics, managing transitions, politics, strategy, water markets, water resources management.
A recent paper to the World Congress of the International Commission on Irrigation and Drainage (ICID)  outlined the theoretical underpinnings of the idea of "water as an economic good", and presented information on the value of water in different end uses, and the supply and opportunity cost of water in different sectors and settings. This paper focuses on just two (but two major) water-using sectors -- urban water supply and irrigation.
1.1. Urban water supply
Urban water supply is a low-volume, high-value (typically between 10 and 100 US cents per cubic meter ) use. The supply costs (incurred in financing and operating the abstraction, transmission, treatment and distribution systems) are relatively high, while the opportunity costs (imposed on others as a result of use of the water) are quite low. Accordingly, the priority issue for the economic management of urban water supplies relates primarily to the supply cost axis (as illustrated in Figure 1).
Conventional economic wisdom suggests that users should be charged full marginal costs -- level IV in Figure 1. In most developing country situations, however, aiming for economic perfection is neither practical nor helpful. Instead, it is imperative that tariffs be set in a way that is understandable, transparent and legitimate and that forces suppliers to be accountable (and thus produce services efficiently). In the urban water supply sector, this "common-sense" pricing approach will therefore mean:
In developing countries most irrigation -- 90% in World Bank-financed irrigation projects -- is for foodgrains. This type of irrigation is a high-volume, low-value user of water (with values generally less than 1 US cent per cubic meter). But there is an important and growing sector of high-value irrigation (often fruits and vegetables), with typical values between 5 and 15 cents per cubic meter of water .
The supply cost of irrigation water for usually modest, but when there is competition with either urban uses or high-value irrigation, the opportunity cost is high. In the Limari Basin in Chile, for example, the supply cost (which is partially subsidized) is about 0.5 US cents per cubic meter, whereas water trades at about 5 US cents per cubic meter . In California, a typical financial charge to an irrigator is about $5 per acre foot (0.4 cents per cubic meter), whereas water trades at the equivalent of about $150 per acre foot (12 cents per cubic meter) .
From the perspective of treating water as an economic good, the great challenge in irrigated agriculture is how to ensure that farmers take into account the opportunity costs, which are often an order of magnitude higher than current charges. This is the essence of the appeal of the approach of water markets -- as described in the Chilean case  "the genius of the approach is that it ensures that the user will face the appropriate economic incentives, but de-links these incentives from the tariff, which is set on "common-sense" grounds". Or, in the case of the California Water Bank (which is a formal mechanism for pooling surplus water rights for rental to other users) the key is " the ability to increase the value of water without an increase in the cost to the farmer (that) is a politically acceptable way of sending the signal to users of the true value of water" 
There are many vital issues relating to charges for supply costs in irrigation. Experience  has shown that cost recovery in and of itself achieves nothing, unless the money collected is used efficiently to improve the quality of services provided. Recent experience in Mexico confirms this -- since management of irrigation systems has been transferred to usersí associations, recovery of operation and maintenance costs has increased from 30% to about 80% .
I: Operations and maintenance costs only
II: Average financial (capital + O&M) cost, with capital valued in terms of historical costs
III: Average financial (capital + O&M) cost, with capital costs computed in replacement terms
IV: Long run marginal cost of additional supplies
A: Water can be used only by an individual user
B: Water can be leased or sold to neighbors
C: Water can be leased or sold within a limited district
D: Water can be leased or sold to any urban or agricultural user
Fig. 2. Supply cost, opportunity cost and full economic cost for public
In terms of economic signals to irrigators, however, as illustrated in Figure 2, the "vertical axis" (supply costs) is relatively short, and the "horizontal axis" (opportunity costs) is often long. That is, from the perspective of the economic allocation of water, the key challenge is to ensure that farmers are aware of the opportunity cost of the resource, and that there are institutional arrangements for ensuring that water moves to higher-valued uses. This paper will, accordingly, focus primarily on the opportunity cost issues and not the supply cost issues in irrigation.
The ICID review  concluded that it was inappropriate, on a number of counts, to think of rolling opportunity costs into water tariffs (as has been suggested in several countries, notably Chile  and South Africa). This is so for three main reasons:
Discussions stimulated by the ICID paper have confirmed  that this is a useful and practical approach, but that the basic framework needs to be expanded to take account of return flows and the positive and negative externalities they generate.
2.1. Negative externalities
Use of water by one user commonly has negative impacts (externalities) on other users. For example, pollution from a town can mean that downstream users have to incur additional treatment costs. Similarly, drainage water from irrigation fields often carries high levels of salts, nutrients and pesticides, leading to losses of aquatic habitats . These externalities are easily incorporated into the conceptual framework of the ICID paper, by simply increasing the supply costs to include the cost of mitigating the negative externalities (Figure 3).
The implications of Figure 3 are evident in practice in the real world.
For example, in many countries cities are required to meet specified wastewater
(or receiving water) quality standards. The utility does this by treating
its wastewater, and passing on the costs of this treatment to its customers.
Similar practices are now starting to emerge for non-point sources of pollution.
In the Murray-Darling Basin in Australia, for instance, the major water
quality problem is the high levels of irrigation-induced salinity .
The Murray-Darling Basin Commission has now specified maximum salinity
fluxes from its different member states. Salinity control measures are
required to stay within these limits, with the costs of these control measures
being passed on to irrigators in their water bills. (These costs are considerable.
In the state of Victoria, for instance, irrigators pay roughly equivalent
amounts -- about 1 US cent per cubic meter for water and a similar amount
as a salinity levy . In the Colorado Basin it costs, on average, about
$70 to remove a ton of salt . For a program which aims to reduce salinity
levels by 200 ppm, this equates to a cost of about 1.5 cents per cubic
meter. This is a very substantial amount -- it compares to typical water
levies of about 0.4 cents per cubic meter ).
Fig. 3. Incorporating negative externalities
2.2. Positive externalities
Return flows constitute a vital element of many hydrologic systems. For example, farmers in the Gangetic Plain who apply more surface irrigation water than is required for evapotranspiration are, in effect, recharging the aquifer which underlies their fields, thus performing an aquifer recharge service for farmers who use groundwater. As Seckler  and Fredericksen  point out, analyses which concentrate on (apparently low) farm-level irrigation efficiency do not take into account the fact that basin-level efficiency may be high. This observation is important because it means that the benefits of apparent efficiency improvements at the farm level may be illusory or even negative at the system level.
Consider the case where surface water irrigation is both meeting the evapotranspiration needs of crops, and recharging a groundwater aquifer which is subsequently used for irrigation. It is apparent that the surface water users are providing a "recharge service" to the groundwater users. What should the groundwater users pay the surface water irrigators for this service? An upper limit on this charge would be the cost of a formal recharge system (with which the groundwater users could recharge their aquifer). In such a case, the relevant charges to surface and groundwater users would be as shown in Figure 4. Once again, this is a "common sense" approach, with surface water users effectively getting a credit from groundwater users for the recharge service they are providing.
In many circumstances, of course, surface and groundwater users are the same population, in which case the notion of a monetary transfer from a user (to himself!) would not be sensible. Furthermore, in many areas where there is conjunctive use, there are countervailing "distortions". In Indian irrigation systems, for instance, headwater and tailwater users are charged the same (very low!) price for water. On the one hand, since tailwater users make greater use of groundwater, they "get an implicit subsidy" from the headwater users. On the other hand, the quantity and reliability of the service to the headwater users is far superior, probably more than offsetting the implicit subsidy.
Real-world experience shows that where there is a will there is a way
-- there are practical solutions to the non-consumptive use issue .
In the Western United States there are a variety of methods for taking
account of return flows from irrigators. Under the laws of most Western
states, return flows "belong to the stream". The practice of the Northern
Colorado Water Conservancy District is consistent with this. While each
water user has the full right to purchase, sell, trade or rent rights to
the primary, consumptive flows, the District retains all rights to the
return flows of water.
Fig. 4. Supply costs when recharge is important
In this, as in all other aspects of water management, it is necessary to find a balance between the ideal and the practical. When water rights are sold in the Western US, the quantity of water that may be transferred to a new use will be limited to the amount of water which was deemed to be consumed historically. This is typically done by a State Water Engineer who makes an estimate based on the factors such as the type of crops cultivated, soil type and climate .
Finally, while the return flow issue is generally raised in the context of the pricing of irrigation services [11,12], the distinction between consumptive and non-consumptive use is not relevant only to irrigation. Indeed, taking the US as an example, consumptive use as a percentage of withdrawals was 56% for irrigation, compared with 17% for urban water supplies, 16% for industry and just 3% for thermoelectric power .
2.3. Negligible externalities
There are many situations in which return flows have neither a positive nor a negative value for other users in a basin. For example, where return flows accrue to low-value "salt sinks", they have no value. Or where return flows discharge into the ocean, they are also effectively "lost" from the system. In such cases there is no need for modification of the use-opportunity cost analysis presented on Figures 1 and 2. The 1989 agreement between the Imperial Irrigation District (IID) in California and the Metropolitan Water District (MWD)which serves Los Angeles, is a good illustration of such a case. Under this agreement, the MWD pays $120 million for conservation activities (mostly canal lining, but also operations modernization) in the IID. In turn, the MWD obtains the 130,000 ML of water which are conserved. This agreement was possible in part because there are few externalities -- the IID is at the lower end of the river system, and there are no opportunities to reuse return flows, which are therefore considered "wasted" .
There is an emerging global consensus on both the necessity for more effective management of water resources, and on the principles of effective management. The Dublin Statement of the pre-UNCED Conference on Water and the Environment  remains the clearest of such statements, articulating three principles. The three "Dublin principles" are:
3.1. There must be a demand for reform
The first requirement for reform is that there must be a demand for reform. Unless the shoe pinches, reform is unlikely to take place. This is an obvious and yet often-neglected fact, especially by water professionals who see basin-wide management, for example, as self-evidently necessary in all places. A couple of examples suggest where the impetus for managing water as an economic good might arise.
The most obvious type of water stress is scarcity. Accordingly, wherever there is scarcity, there has been an endogenous move by local citizens to develop some form of informal water market. That these markets have been invisible and usually illegal does not mean they have not existed, nor that they have not become very sophisticated, as illustrated by the water markets of Gujarat . In such settings, formalization of water markets greatly reduces transactions costs, and is thus welcomed by those who have been trading "in the black market". The Australian case is informative -- "the major impetus for the development of water markets came from users rather than the government,... all the government did was remove the legislative obstacles to transferability" . Similarly, in urban areas of most cities of the developing world there are sophisticated water vending systems which fill the void left by poorly-performing utilities . Because the unserved (usually the poor) typically pay 10 times as much for a liter of water than do the served, there is a tremendous implicit demand for reform which can be tapped. (As in all black markets, however, there are those who thrive on the distortions. Jakarta, where vending is very widespread and very lucrative, proves a general point to which we will return later, namely that there will be losers in reform processes , and that their interests have to be identified and dealt with.)
Scarcity is a great impetus for change, and so is pollution. In 1857 the slogan of the day in London was "India is revolting and the Thames stinks". The result was Chadwick and the sanitary revolution in the United Kingdom. Similarly, in the early part of this century, the pollution of the Ruhr River in Germany, and its threat to the operation of the industrial heartland, mobilized industry and society to do something. The upshot was the Ruhrverband, a revolutionary approach to water quality management which was built on the twin principles of management by stakeholders and using economic instruments to provide incentives for efficient use and for waste reduction . And when France started recognizing the importance of its river pollution problems, it adapted, in 1964, the Ruhr approach on a national level, with the result being the famed "River Basin Financing Agencies". And now, when the water quality situation in the Paraiba do Sul river (which is the sewer of Sao Paulo and the drinking water supply for Rio de Janeiro) becomes intolerable, the states involved (Rio, Sao Paulo and Minas Gerais) are forming a Ruhr-French type basin agency, again grounded on participation and the use of economic instruments, to deal with the problem .
All stresses and challenges do not have to be the "natural challenges" of scarcity and pollution. Stresses in the economic and institutional machinery can also act as an impetus to change. Of particular importance here is the overall view which a society takes of its economic development process. Thus, in settings as diverse as Chile, Australia and Peru, it has been the articulation of an open, export-oriented, growth-driven economic development strategy which has been critical in providing the impetus to improving the economic performance of water (and other factor) markets.
The possibilities of reform are, therefore, greatest when there is a confluence of "natural" challenge (scarcity or pollution) and institutional reform-mindedness, as exemplified by Chile in the 1980s and Australia today. The basic point here is simple -- creative responses in dealing with water as an economic resource will only happen when there is a problem to be addressed, when that problem is perceived to be important, and when there is a political climate conducive to reform.
Water is a good with special properties -- it is the basis of life itself, it is not produced, it is unitary, it is fugitive . These particular attributes have long made water "special", in symbolic, religious, and legal terms. It is no wonder, therefore, that there is much skepticism and concern about the effects of reforms which purport to treat water as an economic good.
3.2.1. Concerns about "the end of irrigated agriculture"
A common concern is that treating water as an economic resource will mean "the end of irrigated agriculture". While this will certainly be true in some local areas, the overall effect on the quantity of water used in irrigated agriculture will be small. There are two reasons for this -- first, because irrigation is a dominant withdrawer and even more dominant consumer of water in all countries where irrigation is important. The changes in the United States illustrate the general point: between 1960 and 1990, while withdrawals for municipal purposes almost doubled, irrigationís share of total consumption fell by only 4%. As concluded by the US Department of Agriculture: "Growth in non-agricultural water needs, particularly in areas with limited supply-enhancement opportunities, may be met by relatively small shifts in national irrigation water use. However, small national shifts may mean large adjustments in local irrigated activity." The implications of this are obvious -- it is essential that the reform debate be informed by solid analysis which can distinguish the legitimate (and in this case local and limited) concerns from extravagant "doomsday" claims.
3.2.2. Concerns about the thinness of the market
Another set of skeptics have the opposite concern. They claim that water markets are, in practice, very "thin", so thin that the notion of a functioning market allocating resources is an illusion.
First some facts. In Chile the number of transactions varies very widely, in a way that is systematically related to water stress and structural changes . In most basins stresses are not yet great and so transactions are rare . But in highly-stressed basins, there are many transactions. For example, in Santiago County, over a one-year period, 3% of total water rights were transferred (with 94% of transferred water moving from one farmer to another) . In Australia (where there are currently both area-of-origin restrictions and restrictions on trades out of agriculture) about 2.5% of all water is leased each year and about 0.3% sold. The Northern Colorado Water Conservation District illustrates how important the cumulative effects of transfers can be -- in 1957, 98% of deliveries from the Colorado-Big Thompson Project were used in irrigation; in 1989 this figure was down to 73% .
3.2.3. Concerns about the impact on the poor
There is a common and reasonable concern that treating water as an economic good will inevitably be damaging to the poor, especially in terms of supplying "basic human needs". This concern acquires particular validity because most interventions to "keep tariffs low" are defended in the name of the poor.
In the urban water sector the evidence is compelling and consistent throughout the developing world. Giving politics a central role in determining tariffs has meant three things -- costs are too high because utilities are not accountable to users; coverage is low, with the poor always the last to get services; the underserved have to resort to buying from water vendors, typically at prices 10 times those which they would have to pay an efficient utility . This has aptly been labeled "the hydraulic law of subsidies" -- the subsidies follow the water, and the water flows to power and influence and away from the poor.
Recent experience in the urban sector has shown that with commitment and imagination, the poor can be much better off when water is managed as an economic resource. Three examples illustrate this.
In Santiago, Chile, the government realized that it was inherently contradictory to require that an urban water utility (EMOS) function as a commercial entity and provide subsidized services to the poor, since each subsidized person served would represent a loss of revenue to the utility. Accordingly, the government decided to institute a targeted, means-tested, government-administered "water stamps" program, whereby poor people would get "stamps" which would cover part of their water bill. The utility then not only strengthened its focus (getting out of the welfare business and focusing on becoming the most efficient utility it could), but it now had a clear incentive to serve the poor, who became revenue-generating customers like all others. The system works very well .
In Conakry, Guinea, the performance of the water utility in the late 1980s was catastrophic -- water for only a few hours a day, with the poor, as always "at the end of the line". The familiar "low-level equilibrium" prevailed -- service was poor, people were not willing to pay, revenues were inadequate, service got worse and so on. The government made creative use of a World Bank credit to get to a "high-level equilibrium". The assets were leased to a private operator who was paid a fee which reflected the full cost of the service. Users initially paid only about a quarter of this fee, with three quarters of the operatorís fee covered by the World Bank credit. Users were informed that service would improve, and that as it did tariffs would be increased to cover costs over a five-year period. Although problems remain , this innovative approach worked well for the poor -- coverage increased from 15% in the 1980s to 52% in 1994.
In Buenos Aires, Argentina, the public water company performed poorly for years -- coverage was low, water was rationed every summer, and prices were high. The Government gave a concession contract to a private operator in 1993. At the end of 1995 the water tariff was 27% lower than it was when the utility was publicly run, 650,000 new water connections and 340,000 new sewerage connections had been made.
The experience of the urban water sector is clear and well-documented -- the poor are much better off when water is managed as an economic good. What of irrigation?
The inequities of existing command-and-control mechanisms for water allocation in irrigated agriculture have been widely documented (for instance by Wade  in South Asia). Because water has rarely been formally managed as an economic good in developing countries, however, there is little information on the equity effects of a market-oriented management system.
In the case of Chile, there are differing positions on the equity implications of water markets. As always, the counterfactual is of central importance and debatable! Proponents of the Chilean water markets  argue that the counterfactual is of subsidized infrastructure which inevitably (just as in the case of urban water) differentially favors the well-connected. They further argue that it is the poor that suffer disproportionately from the fiscal deficits and inflation to which such subsidies contribute. Critics of the Chilean markets , focus on the fact that no effort was made to address the specific problems which faced the poor when the market system was introduced: "In the 1980s the government undertook no campaign of public information or education about the Codeís new features, not offered legal or technical advice about how to apply for new rights or regularize old ones". What is striking in Chile today is that post-Pinochet social democratic governments (who have a strong commitment to equity) have remained firmly committed to the use of water markets, while being equally committed to addressing the informational deficiencies which disproportionately affect the poor .
A rare empirical assessment of the equity impacts of water markets was done for Spanish and Western US irrigation in the 1970s. The authors  concluded: "although it is a doctrine of many welfare economists that procedures that rank high in efficiency will do poorly in distributing income equally among beneficiaries, while procedures that do well in distributive terms will be inefficient... this conventional wisdom does not apply to a wide variety of conditions in irrigated agriculture".
3.2.4. Concerns that the environment will be neglected or damaged
In the past there was a widespread perception that there is some inherent contradiction between "the capitalist economy" and the environment, and, therefore, concern with the environmental effects of treating water as an economic good. After the devastating effects of command-and-control policies on the environment became clear in the Soviet Union and elsewhere, perceptions have changed dramatically. It is now widely understood that it market mechanisms induce efficient resource use, and that inefficiency is the enemy of the environment. And it is equally widely understood [18,32] that this is particularly true of water.
Many sophisticated environmental groups have, accordingly, become vigorous advocates of the concept of water as an economic good. In the case of some of these groups, such as the World Resources Institute  and the Environmental Defense Fund , the principal issue is that treatment of water as an economic good per se will mean much more efficiency, and far fewer environmentally-destructive investments. For others, the purchase of water rights becomes a cost-effective and practical method through which environmental requirements can be handled without expropriation. The Nature Conservancy, for instance, has spent $1.5 million to purchase water from farmers to leave instream in the Carson River in Nevada . The US EPA is also starting to follow the same path -- Federal Clean Water Act funds now being used to buy water rights from irrigators in the Truckee River area (near Reno, Nevada), to increase river flow in dry summer months .
3.2.5. The "exceptionalism syndrome" -- "markets may work for selling cars, but they donít work for water; water markets may work in California, but they canít work in India,....."
Institutional change has always been resisted, precisely because it involves change. A universal argument against change is that "water is different" or "India is different". It is, however, becoming increasingly clear that there is a remarkable degree of commonality between the ingredients of successful reforms in quite different contexts, ranging from Taco Bell, to General Electric, to the New York Police Department, to the NGO Aravind Eye Hospital in Tamil Nadu , to the state government of Ceara in Brazil . The "uniqueness" idea is not standing the test of time!
In the water sector there is a remarkable degree of similarity in the nature of "the problem" throughout the world. The political economy of contemporary public irrigation systems in, say, India , is remarkably similar in many respects to the political economy of public irrigation systems in the Western United States . And there is much in common between the groundwater markets of Gujarat  and those of New Mexico . And there is little distinction between the reform recipes for urban water supply in Adelaide and Abidjan. It is this very convergence which underlies the consensus around the Dublin principles!
This does not imply that the instruments used in different cases will not be strikingly different. The institutional arrangements are not the same in short, unregulated river basins in Chile as they are in continent-wide rivers like the Murray-Darling, but the underlying principles are similar.
What is striking in the contemporary water management arena is how a few hegemonic ideas are becoming universalized very rapidly. The Ruhr-French model (of participatory management, with the use of user and pollution charges) is proving to be well adapted to quality problems in the Paraiba do Sul river in Southeast Brazil  and for the management of coastal wastewaters in the United States . And the water market/water bank model is finding wide application where there is scarcity -- in Chile, Argentina, Mexico and South Africa, just as in Australia and the Western United States.
3.2.6. Discussion, debate and development of a consensus for reform
There are, essentially, two reform paths. Either a dictatorial government can simply declare that the resource will, henceforth, be managed differently. Or there is a process of open debate, with different stakeholders expressing their concerns and views. There are important examples of water reform which have taken place via the former mode (notably Chile), and there are certainly some countries which could follow that route today. But, happily, in the vast majority of countries of the world today, changes require the consent of the people. In these circumstances a vigorous, open debate is necessary for reform to take place. Three contemporary examples are illustrative of "good practice" in this regard.
The first example is Australia. As in every other situation of scarcity, informal water markets existed for many years. Starting in 1983, the government formalized and legalized this practice, simply by removing legal obstacles to transferability. These markets have worked well in many respects, and have become a fixture of the institutional landscape, supported by both sides of the political spectrum . The markets have, however, been limited because there have been restrictions on inter-sectoral and inter-state trade. In recent years the Coalition of Australian Governments (COAG, comprising the Prime Ministers of the Federal and State Governments) has looked carefully at Australiaís overall economic and trade policies, and concluded that economic growth requires that water (and other factor) markets operate efficiently. Accordingly, COAG has decreed that "the major goal of water resources management is to achieve the highest and best value of the limited resource... (and that)... the move towards property right regimes is intended to link the responsibilities and accountabilities for decisions on water use with the incentives and sanctions for achieving highest value use" . A central element in achieving this goal is to ensure that the geographic and sectoral reach of water markets is much broader, with inter-state, inter-sectoral trade a prime objective. What was clearly understood was that high-level commitment was a necessary but not sufficient condition -- it was vital that there be a broad-based discussion of the why and how. And in this discussion it was vital to identify different stakeholders and their interests and concerns and address these specifically and systematically in a consultation process. The work of the "Murray Darling Basin Commissionís Water Market Reform Working Group"  is a model of how this stakeholder identification and information/consultation process should be done.
The other two examples are developing countries which are now embarking on similar reforms. When Peru initiated discussions of "following the Chilean model", there was concern that the Pinochet dictatorship was a necessary condition for such radical reforms. The (as yet incomplete) experience in Peru showed that, on the contrary, broad-based discussions with interested parties strengthened rather than weakened the support for market-based reforms in water management . South Africa is also undergoing a major reform in its water management practices, with the management of water as an economic resource a major objective. The debate on this policy issue is very broad and very open and is making use of a variety of traditional (public meetings) and modern (internet-based) instruments .
Finally, it is imperative that this debate be conducted in the appropriate language. As noted elsewhere: "Above all else, policy-makers need to demystify the academic literature, to strip away the jargon and explain to politicians, public servants and private stakeholders the advantages and disadvantages of market solutions to natural resource allocation" .
While there are clear and universal principles on what constitutes effective water management, the details of what can and should be done are enormously variable. It is obvious that context -- historical, cultural, legal, institutional, political, economic and hydrologic -- matter a great deal, and that the particulars of appropriate solutions require careful and ongoing adaptation to particular circumstances. A couple of examples illustrate this general point.
Within the Ruhr basin, pollution charges are the major economic instrument for managing point sources. But as non-point source pollution has become a more important and recognized problem, different and creative approaches (such as subsidies for changed land-use practices) have been invented, and used successfully .
Within most water markets, a mix of different instruments are used. Short-term leases are effective for addressing the needs for higher reliability, but long-term sales are more appropriate when there are structural shifts (in the location of agriculture, or from agriculture to industry) . And as water markets mature, other niches are emerging -- with options and futures contracts now coming onto the scene in the Western United States, Australia and Chile.
In urban areas, there is a similar need for inventiveness and adaptation. Thus there is innovation in how to direct subsidies to the poor (see the Santiago example discussed earlier), and innovation  in how to charge for water from public standpipes.
Treating water as an economic resource is desirable for a wide variety of economic, equity and environmental reasons. And the benefits of this approach are substantial. But precisely because context matters so much, there are no ready solutions which can simply be plucked off the shelf and no "final solutions".
Reform requires a complex mixture of impatience and patience. Impatience is required to make paradigm shifts, but then it must be realized that implementation is a very long-term process, which requires persistence, patience and adjustment. This is well illustrated by the contemporary processes in Australia and Chile. In both places formal water markets have existed since the early 1980s and in both places these have been relatively successful. But both countries are now in the process of major adjustments -- in Australia to extend and deepen what were initially relatively timid reforms ; in Chile to adjust the framework to account for some distortions (hoarding of rights by hydroelectric companies) and to embed the markets in a more effective river basin management framework [2,23].
There is a further, very clear lesson from all efforts to deal with water as an economic resource, whether it be through pricing or marketing mechanisms. In all cases where economic instruments for water management work well -- such as the Ruhr, the French River Basin Financing Agencies, the Northern Colorado Water Conservancy District, the New Mexico water markets, the Murray-Darling Basin, the Elqui Basin in Chile -- this happens in the framework of an effective overall river basin management system. The issues of governance and technical management of the resources are at least as important as, and essential complements to, the use of economic instruments .
Experience shows that blind advocacy of, say, water markets as "the silver bullet which will solve all problems" is not only misguided but actually counterproductive (as has been apparent in the debate over water markets in Chile). Acknowledging broader issues and keeping expectations realistic is not a recipe for inaction, but essential if there is to be effective reform.
Reforming water management systems is never easy. Early successes are vital in demonstrating that change is possible and in building a broader constituency for reform. The strategy being followed in introducing inter-state water trading in Australia is a good example.
The background to this stage of the Australian reform process is recognition of the gains to be had from broadening and deepening water markets. As has been pointed out in the Western United States "since localized markets such as those within water conservancy districts and small river basins have been active for many years, some of the greatest opportunities for increased efficiency lie in interdistrict and interstate markets". In the lower Sevier River in Utah, for example, the actual gains in efficiency were measured following a relaxation of exchange restrictions -- exchanges were allowed between four irrigation districts (rather than just within a particular district). The average real rental price in the period after the free inter-district exchanges was more than three times that in the exchange-restricted period .
The practice of water trading in Australia is now well established and supported by all major stakeholders, and all political parties. However, the water markets in Australia are relatively restricted (to sales within agriculture, and to sales within specific states). High-level commitment to inter-state trading has been made; the question is where to start. The Murray Darling Water Market Reform Working Group  is clear and strategic:
"Initial community consultation on permanent interstate trade focused on the Mallee region for three reasons:
Similar strategies are evident in other parts of the world. In Brazil, for example, the establishment of tradable water rights in being undertaken on a pilot basis on a single-state basin, in a state (Ceara) where water is scarce and where there is a recent track record of effective, modernizing government . In India the establishment of the first formal water market has been proposed as a method for effecting the voluntary transfer of water from low-value agriculture to high-value urban uses in the Madras Metropolitan area. The chances of success are increased by the fact that it is a classic "win-win" situation (in which the city could buy water at a fraction of the cost of alternatives, and farmers could get paid much more than the value of the water in irrigation) .
There is no such thing as the perfect water management system. Insisting on perfection is a recipe for inaction -- the best can become the enemy of the good.
All water management systems have to face many difficult issues -- systems which manage water as an economic resource are no exception. Consider, for instance, the thorny issue of allocation of initial rights in publicly-financed irrigation systems. In virtually every country in the world, these systems have been heavily subsidized, with (as always) the privileged getting disproportionate benefit from these subsidies. One legitimate perspective on the allocation of permanent rights is thus: "these people have had a privileged position for long enough; now is the time to allocate the rights on a more equitable basis." However, those who enjoy those rights never see it like this. Their generic position is most persuasively argued by a recent purchaser of land in the irrigation district: "When I bought my land, I implicitly paid for the right to water at the historic (subsidized) price. To take this right away from me now -- either by pricing or by re-allocation -- is expropriation, which I will resist fiercely and honorably."
There is reason (and often commitment) on both sides of this debate, and thus no elegant "perfect" solution which will be just, efficient and politically acceptable to all. A practical outcome to the solution will be different in different places. In New South Wales, for example, there has been considerable controversy regarding one-off capital gains some water-entitlement holders have been able to make on the sale at market prices of water rights acquired under government-financed and subsidized projects. After much debate, the government approach been to accept a one-off imperfection in return for continuing future improvements in efficiency .
A similar debate is under way in South Africa, complicated greatly by the fact that it was whites who were the beneficiaries of apartheid-era pork barrel politics. Sanctification of these past inequities is out of the question, yet a policy of confiscation, as in other parts of the world, would meet with serious political opposition. While this debate is far from over, the task is clear -- try to "start playing the right ball game" (perhaps by buying out initial rights at the opportunity cost of water in irrigation, and then auctioning off all rights). This would probably mean, as in Northern Colorado , that high-value urban users would buy the rights, and then lease them back to farmers until such time as they were needed by the cities.
Finally, it should be recognized that these difficulties are not created by a water market system -- it is simply that a market system brings these issues to the surface. The bottom line is that there is no perfect solution -- the challenge is to find a reasonable, practical second-best solution which will start the vital process of treating water as an economic resource.
3.7. Ensure that legitimate third party interests are addressed, but that transactions costs are minimized
3.7.1. Concerns with economic and employment effects in "areas of origin"
The rationale for tradable water rights has been clearly articulated by the Water Science and Technology Board of the US National Academy of Sciences: "The classic rationale for all economic activity -- gains from trade - motivates most water transfers. Buyers perceive that the cost of purchasing existing water rights and transferring water to new locations, seasons, or purposes of use is less than the cost of alternative means of securing needed supplies. Conversely, sellers -- generally farmers -- sell when the price offered is greater than the economic value of the crops or livestock they produce. The net result is that the new use generates higher economic returns than the old use" .
Where they have been established, water markets have performed this re-allocation function well. In the State of Victoria in Australia, for example "water is tending to move away from badly-salinized mixed-farming land to dairying and horticultural areas where the returns are higher" . In California the Water Bank has meant the transfer of water from low-value fodder and foodgrain crops to high-value fruit, vegetable and nut agriculture, and to municipal uses . A study of trades in the Arkansas River Valley illustrates the typical imbalance between costs and benefits well -- net income losses in the area of origin were about $53 per acre foot (4.4 cents per cubic meter), and the market value of water in the urban areas $1,000 per acre foot (80 cents per cubic meter) .
Precisely because of these imbalances and shifts, the economic benefits from trades is substantial -- an estimated $104 million in 1991 alone from the California Water Bank ; about $5 million per year in the (presently-restricted) market in New South Wales  and an estimated $100 million per year in additional agricultural production in the southern part of the Murray-Darling Basin in Australia once inter-state trading becomes a reality . In Chile, the gains-from-trade in the (small) Limari Valley are about $2 million per year . The impacts on employment follow a similar pattern -- in the case of the 1991 California Water Bank about 1600 jobs lost in the area of origin, offset by 5,400 jobs gained in the importing regions .
These large net benefits notwithstanding, there are losers in water trades (as there are whenever there is an adjustment in any market). Naturally these third parties would like protection, usually by way of legislation on "area-of-origin restrictions". As has become evident in the case of Australia "advocates of restrictions on transfers ... argue that unrestricted transfers impoverish less-productive regions. Yet this should, in fact, be viewed as a vindication on the transferability of water -- it shows the resource moving to higher-valued use". Whatever the conceptual shortcomings of the area-of-origin claims, they remain a fact of life  and a potentially serious impediment to managing water as an economic resource because of the way in which they can increase transactions costs.
3.7.2. Minimizing transactions costs
While externalities are (as the above discussion suggests) too often the "first refuge of scoundrels", the very nature of water means that real externalities are, in fact, pervasive. Accordingly, "the main administrative problem in water markets is the existence of Ďthird-partyí effects that take the forms of changed return flows, changed groundwater levels and water quality changes, and the main issue in making markets work more efficiently is to identify and quantify these effects accurately and quickly and to get agreement on their magnitudes so that compensation and/or adjustments to the original property rights can be carried out without excessive transactions costs" .
Because transactions costs are so important, the choice of institutional arrangements for dealing with these is critical. Detailed empirical investigations of transactions costs in the Western US are revealing. Policy-induced transactions costs (include attorneysí fees, engineering and hydrological studies, court costs and costs paid to state agencies) range from about $50 per acre foot (4 cents per cubic meter) in New Mexico (where the State engineer adjudicates these issues) to nearly $200 per acre ft in Colorado (where they are adjudicated by the courts). The delays follow a similar pattern -- an average of 4 months in New Mexico, and 30 months in Colorado 
The issue of transactions costs is particularly vital where short-term markets are dominant (as with the California Water Bank), since the problem is a rapid response to drought conditions . Accordingly, much attention is now given to streamlining the systems that impose superfluous restrictions, costs and delays on the transfer process, and, at the same time, to devise new ways to account for the important interests that are now left out. For example, in California, "one mechanism being considered would decouple transactions and compensation by establishing a compensation fund from which third party claims would be paid. The fund would be kept solvent by a standard charge on all inter-basin water sales. While the system is open to aggregation inefficiencies and moral hazard costs, it is hypothesized that the reduction in transaction risk will more than compensate for these costs" .
A related, important issue for third parties is whether trades deal with consumptive use or withdrawals. One way of reducing transactions costs is to tie transfers to consumptive use . While there is no perfect solution to the problem of determining consumptive use, the problem can be addressed effectively where there is a will. In New Mexico, for instance, all rights are consumptive, with a typical description of the water right as follows: "This dedication is for 6.82 acres of irrigated land having a diversion right of 20.46 acre feet of water per annum and having a consumptive use of 1.5 acre feet per irrigated acre for a total of 10.23 acre feet per annum of consumptive use" . The consumptive use is estimated by the State Engineer, who considers the type of crops cultivated, soil type, climate and other factors that affect water consumption . Protests are rare. In California, where farmers are paid not to use water, "(aerial) photography used by the US soil conservation service to verify and monitor crop areas for subsidy payments were used to establish the past cropping pattern of an area, and to verify that irrigation was discontinued once the fallowing agreement was in force. Using this data base, the Water Bank was monitored at low cost ($4 per acre ft, or 03. cents per cubic meter)" .
This paper extends a previously-developed framework  for thinking about management of water as an economic resource, by showing how positive and negative externalities can be taken into account. The main focus of the paper, however, is on assessing the lessons of experience which emerge from successful reforms.
The conclusions of this review are that strategy is vital in implementing a market-oriented water reform. More specifically, the following emerge as a tentative set of "rules for reformers":
1. Briscoe, J. (1996) Water as an economic good: The idea and what it means in practice, Special Session R 11, International Commission on Irrigation and Drainage, Sixteenth Session, Cairo, Egypt.
2. Briscoe, J. (1996) Water Resources Management in Chile: Lessons from a World Bank Study Tour, Washington DC.
3. Water Strategist. (1989) Evaporating water markets? New contingencies for urban water use, Water Strategist, Claremont CA, July 1989, Vol 3, No.2, p 11,12.
4. Howitt, R.E. (1996) Initiating Option and Spot Private Water Markets: Some examples from California, paper presented at the IDB Seminar on Economic Instruments for Integrated Water Resources Management, Inter American Development Bank, Washington DC., 15 pages.
5. World Bank (1995) The World Bank and Irrigation, A World Bank Operations Evaluation Study, Washington DC.
6. Conley, A. (1996) The future of irrigation under increased demand from competitive uses of water and greater needs for food supply. International Commission on Irrigation and Drainage, Sixteenth Session, Cairo, Egypt.
7. US National Research Council. (1989) Irrigation-Induced Water Quality Problems. Washington DC., National Academy Press.
8. Murray Darling Basin Commission (1996) The Murray Darling Initiative, Canberra, Australia
9. Sturgess, G.L. and Wright, M. (1993) Water Rights in Rural New South Wales: The Evolution of a Property Rights System, The Centre for Independent Studies, Policy Monograph No.26, St. Leonards, Australia
10. US Dept of the Interior (1995). Quality of Water: Colorado River Basin, Progress Report No. 17, Washington DC
11. Seckler, D. (1993) Designing Water Resources Strategies for the Twenty-First Century, Center for Economic Policy Studies, Winrock International, Arlington, Virginia.
12. Fredericksen, H. (1996) The water crisis in the developing world: Misconceptions about solutions. Journal of Water Resources Planning and Management, Vol 122. No 1.
13. US National Research Council. (1996) The Future of Irrigation. Washington DC., National Academy Press.
14. Gollehon, N., Aillery, M. and Quinby, W. (1995) Water use and pricing in agriculture, pp 45-56 in Agricultural Resources and Environmental Indicators, Agricultural Handbook No. 705, United States Department of Agriculture, Washington DC.
15. US National Research Council. (1992) Water Transfers in the West. Washington DC., National Academy Press.
16. International Conference on Water and the Environment (1992) The Dublin Statement, World Meteorological Organization, Geneva
17. Shah, T. (1993) Groundwater Management and Irrigation Development: Political Economy and Practical Policy, Oxford University Press, New Delhi, India.
18. World Bank (1992) World Development Report: Development and the Environment, Oxford University Press, New York.
19. Porter, R.C. (1996) The Economics of Water and Waste: A case study of Jakarta, Indonesia, Avebury, Aldershot, UK
20. Ruhrverband (1992) Tasks and Structure, Essen, Germany
21. Kelman, J. (1996) Building a water resources management system in Brazil -- A status report, Federal University, Rio de Janeiro, Brazil.
22. Rogers, P. (1992) Comprehensive water resources management: A concept paper, World Bank Infrastructure and Urban Development Department, WPS 879, Washington DC.
23. Pena, H.T. (1996) Water markets in Chile: What they are, how they have worked and what needs to be done to strengthen them, Paper presented at the Fourth Annual World Bank Conference on Environmentally Sustainable Development, Washington DC.
24. Gazmuri Schleyer, R., and Rosegrant, M.W. (1996) Chilean water policy: The role of water rights, institutions and markets. Water Resources Development, Vol 12, No.1, pp. 33-48.
25. Water Strategist (1990) Trading Federal Project Water: The Colorado-Big Thompson Project, Water Strategist, Oct 1990, Vol 4, No 3. p 1-13.
26. Serageldin, I. (1994) Water supply, sanitation and environmental sustainability: The financing challenge. Directions in Development, World Bank, Washington DC.
27. Empresa Metropolitana de Obras Sanitarias S.A. (1995) Memoria 95, Santiago.
28. Brook Cowen, P. The Guinea water lease -- five years on, Viewpoint Note No 78, World Bank, Washington DC.
29. Wade, R. (1986) Corruption, delivery systems and property rights. World Development, XIV(1), pp 127-132.
30. Bauer, C. (1996) Bringing water markets down to earth: The political economy of water rights in Chile, 1976-1995, University of California at Berkeley, California.
31. Maass, A., and Anderson, R. (1978) And the Desert Shall Rejoice, MIT Press, Cambridge, USA.
32. Repetto, R. (1986) Skimming the Water: Rent-seeking and the Performance of Public Irrigation Systems, World Resources Institute, Washington DC.
33. Moore, D. and Willey, Z. (1991) Water in the American West: Institutional Evolution and Environmental Restoration in the 20th Century. University of Colorado Law Review, 62,4, pp 775-825.
34. Anderson, T.L. and Snyder, P.S. (1995) Water Crisis: Ending the Policy Drought (Revised Edition, draft), Political Economy Research Center, Mozeman, Montana, US.
35. New York Times (1996) Novel use of clean-water loans brightens outlook for a river, New York Times, Oct 31, 1996.
36. Harvard University School of Business Administration (1997) Teaching materials for World Bank Executive Development Program, Boston Massachusetts.
37. Tendler, J. (1997) Good Government in the Tropics. The Johns Hopkins University Press, Baltimore, Maryland.
38. Worster, D. (1992) Rivers of Empire: Water, Aridity and Growth of the American West. Oxford University Press, New York.
39. Dumars, C. (1992) Water marketing in the Western United States: A description and commentary. University of New Mexico School of Law, Albuquerque, New Mexico.
40. US National Research Council. (1993) Managing Wastewater in Coastal Urban Areas. Washington DC., National Academy Press.
41. Coalition of Australian Governments (1995) Water Allocation and Entitlements: A National Framework for the Implementation of Property Rights in Water, Canberra.
42. Murray Darling Basin Commission (1996) Report of the Water Market Reform Working Group, Canberra.
43. Department of Water Affairs (1997) Home page on the World Wide Web (http://www.gov.za/dwaf/). Pretoria, South Africa.
44. Briscoe, J. (1995) The German water and sewerage sector: How well it works and what this means for developing countries, World Bank ,Report TWU 21, Washington DC.
45. Simpson, L. (1992) Water resources marketing: The Northern Colorado experience and its applicability to other locations, Northern Colorado Water Conservancy District, Fort Collins, Colorado.
46. Howe, C.W., Shurmeier, D.R. and Shaw, W.D. (1986) Innovative approaches to water allocation: The potential for water markets, Water Resources Research, Vol 22, No 4, pages 439-445, 1986.
47. Gardner, B.D. and Fullerton, H.H. (1968) Transfer restrictions and misallocation of irrigation water, American Journal of Agricultural Economics, 50(3):556-571.
48. World Bank (1996) Water resources supply and management issues in Chennai (formerly Madras), Internal Working Document, Washington DC.
49. Stringer, D. (1995) Water markets and trading developments in Victoria. Water, March 1995, Melbourne, Australia.
50. Howe, C.W., Lazo, J.K., and Weber, K.R. (1990) The economic impacts of agriculture to urban water transfers on the area of origin: A case study of the Arkansas River Valley in Colorado. American Journal of Agricultural Economics 72(5):1200-1204.
51. Rios Brehm, M. and J. Quiroz (1995) The market for water rights in Chile. World Bank Technical Paper No.285, Washington DC.
52. Colby, B.G. (1990). Transactions costs and efficiency in Western water allocation, American Journal of Agricultural Economics, 1184-92.