Peru - A user-based approach to water management and irrigation development

Report No. 13642-PE Peru A User-Based Approach to Water Management and Irrigation Development Country Department ilI Natural Resources Management and RLural Povertv Latin America and the Caribbean Region Document of the World Bank  GLOSSARY OF ACRONYMS AACH Autonomous Hydrographical Basin Authority Autoridad Aut6noma de la Cuenca Hidrografica DGAS General Directorate of Water and Lands Direcci6n General de Aguas y Suelos EMOS Municipal Water and Sewerage Company Empresa Municipal de Obras Sanitarias FAO/CP Food and Agriculture Organization/Cooperative Program FONCODES National Social Compensation and Development Fund Fondo Nacional de Compensaci6n y Desarrollo GDP Gross Domestic Product IDB Inter-American Development Bank INADE National Institute for Developmznt Instituto Nacional de Desarrollo INRENA Directorate of the National Institute of Natural Resources Instituto Nacional de Recursos Naturales MAG Ministry of Agriculture Ministerio de Agricultura MRTA Tupac Amaru Revolutionary Movement Movimiento Revolucionario Tupac Amaru NGOs Non-Government Organizations WUAs Water User Associations Junta de Usuarios de Agua WUO Water User Organizations Organizaciones de Usuarios  PERU A USER-BASED APPROACH TO WATER MANAGEMENT AND IRRIGATION DEVELOPMENT Table of Contents Page No. Preface Executive Summary ....... .................. i - viii Introduction ................................. 1 1. Overview of Irrigation Development in Peru .... ........ 2 A. Agricultural Land and Water Potential and Use .... .... 2 B. Synopsis of Agricultural Policies ................. 3 C. Irrigation Policies and Institutions ................ 6 D. The Current State of the Irrigation Sector .......... 11 2. Tradable Water Rights For Improved Water Management . . 13 A. The Draft Water Law ...................... 13 B. Resolving Issues in Establishing Tradable Water Rights . . 24 3. A Strategy for Irrigation Development .... ........... 36 A. Small and Medium Irrigation Investments .... ...... 36 B. The Development of Large Hydraulic Projects ... .... 39 C. Conclusion ......... .................... 45 TABLES Table 1. Investments and Project Components of Special Projects ....................... 40 Table 2. Peru: Investments and Outcomes of Large Hydraulic Coastal Projects .... .......... 42 ANNEXES A. The Draft Water Law: An Unofficial English Translation B. Water Reform and Poverty in the Peruvian Highlands C. Description of the Large Coastal Hydraulic Projects   PREFACE This report was prepared by Mateen Thobani based on the findings of missions to Peru in May and October 1993 to prepare an Irrigation Sector Project. Loretta Sonn (FAO/CP) and Regina Martinez (Consultant, FAO/CP) participated in the missions and contributed some of the material for Chapters 1 and 3. Paul Trawick (Consultant), drawing from his four years of field work for his PhD thesis, prepared the annex on Water Reform and Poverty in the Peruvian Highlands. Carlos Emanuel, Gabriel Montes (IDB) and Hernan Roman (Consultant, IDB) provided constructive inputs in the analysis of the draft water law and in the formulation of a strategy for the development of large hydraulic projects. Janice Molina translated the draft water law. The Government sent comments on an earlier draft in February and March 1995. The managing division chief is Nicholas Krafft, the Lead Economist is Norman Hicks and the Department Director is Yoshiaki Abe.   Executive Summary Overview 1 Immediately upon assuming office on July 28, 1990, the Fujimori Administration implemented a strong stabilization program that succeeded in halting hyperinflation and improving the country's foreign exchange situation. Over the following years, the Peruvian Government implemented a number of structural reforms aimed at reintegrating Peru into the international community, fostering openness of the economy, broadening the scope for private sector development, and enhancing efficiency in resource allocation. The reforms reversed the decline in GDP that had occurred since 1988 and improved prospects for longer-term economic growth and poverty reduction. 2 The structural reforms were implemented in virtually all sectors of the economy. The agricultural sector underwent a major transformation as a result of far-reaching reforms in land tenure, agricultural pricing and marketing, agricultural credit, and agricultural research and extension. While some reforms were also implemented in the irrigation area, the Administration's proposed water reform that will most profoundly affect irrigation activities -- establishing a system of tradable property rights to water -- has been delayed. 3 This report argues that the proposed water reform has the potential to resolve many of Peru's widespread water problems. However, the draft law would benefit from some modifications. In addition, the report provides some suggestions to ensure that the subsequent regulations to the law are well-designed and implemented. Finally, the report analyzes a proposed strategy for irrigation development, which complements the water reform and is based on sound economic principles. Background 4 Although Peru's water resources are abundant, they are poorly distributed. The bulk (98 percent) of the country's water is located in the Amazon basin, which has limited agricultural potential, while the fertile, low- rainfall coastal region is entirely dependent on variable supplies from rivers originating in the Andes. Since pre-Columbian times, Peruvians have invested heavily in irrigation systems to store and divert water from Andean basins to the coastal areas for use by coastal cities and industry, as well as by agriculture. Agriculture accounts for roughly 80 percent of water use in Peru. Presently, almost half of Peiu's cultivated area is under irrigation, and these areas account for the bulk of agricultural output.  ii 5 By 1992, following several years of virtually no public spending for maintenance or rehabilitation of public irrigation structure, many public irrigation systems faced a high risk of failure. Water delivery became more irregular, water quality deteriorated and water conflicts grew. There was also widespread reporting of water theft. Even in areas where water was scarce, it continued to be used wastefully. Although Peru's existing water law: (i) specified priorities for water use; (ii) mandated that water tariffs cover both operations and maintenance (O&M) costs and construction costs; and (iii) established a comprehensive institutional framework, the system of priorities for water use has proved unworkable, water tariffs typically do not cover even O&M costs, and the public institutional structure is weak. Despite low water tariffs, many users were unwilling to pay them and the Government realized that it would be extremely difficult to increase water charges substantially or to create strong institutions to plan, assign, monitor and enforce efficient water use. Moreover, it was unwilling to devote substantial public resources to subsidizing irrigation activities. 6 Not only has the current system of water allocation resulted in inappropriate water use and poor system O&M, it has discriminated against the poor. Under a system whereby new water rights were assigned at no charge (except for modest water tariff charges) at the discretion of public officials, the politically influential farmers were more easily able to obtain these rights even when there was insufficient water available to justify the granting of new rights. This led to an effective reduction in the rights of existing, less politically influential farmers. In addition, the threat of having one's water rights being expropriated by the State for higher priority uses has discouraged investments in water-intensive activities. At the same time, many of the large public investments in hydraulic projects, which were supposed to make Peru self-sufficient in agriculture, have proved to be ineffective, uneconomical and a large fiscal drain. Proposed Water Law 7 To address these problems, the Ministry of Agriculture has proposed, as the centerpiece of its water management and irrigation development strategy, a new water law modelled along the 1981 Chilean water code, which has successfully improved water delivery and use, stimulated private investment, and reduced water conflicts. Under the proposed Peruvian law, existing water users are to be given property rights to water without charge. Rights to new or unallocated surface water are to be distributed via a public auction. The rights may be traded at freely negotiated prices provided that the trade would not reduce water availability to others and that there is enough water to maintain a minimum ecological flow and to maintain the accustomed quality of life in cities and towns. Rights may also be mortgaged  iii or leased. The law prohibits altering water quality to the detriment of flora or fauna; however, rather than proposing specific sanctions and fines, it defers to the Environmental Code and Environmental Authority to set and enforce water quality standards. 8 Under the draft law, water rights are to be acquired by being recorded in a public Water Rights Registry, specifying, inter alia, the flow or volume (which may be specified in terms of percentage of stream flow or in shifts); the point at which the water will be diverted; whether. it is for consumptive or non-consumptive use and whether it is for permanent or temporary use; the point and form in which the water will be returned to the river system; and the amount paid for the rights. The law also establishes a property tax on water rights. In contrast to the current law, the new rights do not have to be used for any specific purpose, there any no priorities among water rights for different purposes, and the water right is separate from the land right for both surface and ground water. 9 Although opposed by some because of its innovative nature, Peru's proposed system of tradable water rights is likely to have many advantages over the current system of administered water allocation. Because a market system increases the value of water, there are greater incentives to define water rights and to improve their measurement and enforcement. The conveyance infrastructure required for water transfers is likely to be built more cost-effectively by the private sector. Similarly, water user organizations, which have an important role under either system, will become stronger and better-organized when water rights are well-defined and made transferable. In principle, and reflecting the Chilean experience, tradable and secure property rights to water should: (a) Provide an effective way of changing water use towards higher priority uses without creating conflicts and without having to raise water tariffs. Thus it would provide farmers financial incentives to switch to improved irrigation methods or to less water-intensive crops and to sell their excess water to a neighboring city or other farmers. (b) Help reduce environmental damage. By inducing water conservation, the construction of new hydraulic storage or diversion infrastructure can be avoided, thereby averting possible environmental damage. In addition, water conservation by farmers will help control waterlogging and soil salinization caused by over-watering. (c) Assure poor farmers that their effective water availability will not be reduced. Under the current system, water administrators often give new water rights to politically influential water users even when there is  iv not enough water in the system to meet existing water rights. This reduces the water available for others. By defining water rights appropriately and restricting auctions to those instances where surplus water rights are clearly available, the proposed law will help reduce current abuses of the water rights system. (d) Be used as a source of collateral. This is especially beneficial for poor farmers who have few sources of collateral for cheaper loans. It also provides a large farmer the flexibility of being able to mortgage only part of his infinitely divisible water rights as opposed to his entire land and water holdings. (e) Stimulate private investment in activities requiring an assured supply of large quantities of water (e.g., farming, mining, some types of industry, water supply companies) since the uncertainty of access to water will be reduced. The increased investment and improved productivity of water use are expected to benefit the poor indirectly as a result of increased employment opportunities. (f) Stimulate private investment in hydraulic projects as potential investors will be assured that their water rights will not be expropriated in future and that the water generated as a result of their investment can be sold at freely determined prices. This would be particularly useful if governments wished to encourage private investment in hydropower, domestic water supply, and the privatization of large public hydraulic projects (see below). (g) Facilitate the transfer of irrigation infrastructure to water user groups and allow them to be solely responsible for system O&M and for the setting of water tariffs. (h) Provide the basis for better protection against overexploitation of underground aquifers as compared to the current system, whereby owners of the land above an aquifer have unlimited rights to its water. 10 To allow for passage of the water law, the 1993 Peruvian constitution contains a clause (Article 66) which treats land and water resources equivalently. Thus private property rights to water can exist in the same way that private property rights to land exist. A draft water law has been prepared. However, its submission to Congress has been delayed. The Government could usefully use this time to make some modifications to the draft law and ensure that its subsequent regulations are well-designed and implemented. It is important that the Government proceed expeditiously to draft the regulations to the law and that it devote sufficient resources for the  V effective and fair implementation of the law and its regulations. Suggestions for modifying the law and drafting the regulations include: (a) a public information campaign and debate to make policy makers and water users aware of the problems with traditional approaches and understand the operation and potential benefits of water markets. This can help ensure that the final design and implementation of the legal framework is done in a transparent and participatory manner; (b) the establishment of effective public institutions to draft the regulations and to implement the law efficiently and fairly after approval. This may require modifying the law so as to avoid appointing the same people that served as water administrators under the present water rights regime; (c) understanding and addressing the implications of various ways of dealing with the issue of third-party effects arising from "return flows". (d) the suggestion that in the regulations relating to the proposed water law, the tax: (i) exempt farmers holding very small quantities of water rights; and (ii) apply to both consumptive and non-consumptive rights, although at different rates. The existing tax on irrigated land should simultaneously be removed; (e) implementing safeguards to prevent monopolies. While there is no danger of widespread monopolies, they could occur: (i) following privatization of the large hydraulic works and water companies; and (ii) in the area of non-consumptive water rights. To avoid this: (i) an appropriate regulatory framework should be developed prior to any privatization; and (ii) a tax on non-consumptive water rights should be introduced, accompanied with regulations determining power tariffs; (f) streamlining the law by: (i) transferring some of the institutional detail in the law to the regulations to allow them to be changed easily if they turn out to be inappropriate; (ii) specifying early in the law (Article 9) that water rights will be given without charge to existing users; (iii) clarifying that even when water storage and transfer works are performed by individuals, there needs to be an auction of the associated temporary water rights; and (iv) addressing the issue of what will happen to unregistered water rights if users are unable to register their rights 180 days after publication of the law -- one option is to implement the new law by river basin only if the relevant water user organizations in the watershed choose to do so, with the remaining water basins continuing to function under the prevailing law; and  vi (g) measures relating to the initial allocation of property rights. Where there is sufficient water to honor all existing rights, as in much of the Selva region, the initial allocation could be done simply by re- registering them in the new Public Registry. Where the existing registry contains many overlapping property rights, as in the Costa, the initial allocation should be based on past usage. In situations where there have existed gross abuses of water rights, as in much of the Sierra, water rights should be assigned to communities based on historic use, who would then allow individual registration according to the irrigated land area. In addition, to reduce water losses, the initial rights in the Sierra should specify water distribution patterns that are contiguous, or concentrated in one area at a time, according to the order of fields along each canal. In all cases, the Government needs to help small farmers to register their water rights, making maximum use of "block titling." Finally, the auction that would allocate new water rights should be carried out in an open and transparent manner with information on prices and volumes made publicly available, and with zero or minimal costs charged to enter the auction. Strategy for Irrigation Development 11 A strategy for irrigation development that is under discussion proposes treating water and irrigation services no differently from other commodities or services. According to this strategy, the Government will subsidize water and irrigation services only to the extent that they exhibit public goods characteristics and/or positive externalities (e.g. flood control, drainage, soil erosion and siltation reduction, riparian flows, the formation of water user associations and dispute resolution mechanisms, and water supply), or to the extent that a subsidy is an effective poverty alleviation measure. Moreover, the Government will not itself carry out irrigation activities or investments if the users or private investors are capable of undertaking them. In addition to providing possible subsidies, the strategy requires establishing a legal and regulatory framework to reduce uncertainties to users and potential investors and to avoid monopolies from developing. This approach is consistent with sound economic policy and with the Government's approach in other sectors. Tradable property rights to water will greatly facilitate the implementation of this strategy. 12 Under the proposed strategy, the Government would gradually transfer without charge all the irrigation infrastructure in its completed projects to water user associations (WUAs) and would provide them with technical assistance to operate and maintain it. Giving the infrastructure away without charge recognizes the existing situation whereby the Government has been unable to impose water charges to recover investment costs -- water charges  vii rarely cover full O&M charges. In addition, until the rural banking system is sufficiently developed, the Government is considering making long-term funds available for lending toward financially viable off-farm irrigation investments. This would help avoid a possible failure of the irrigation system as a result of the past government neglect of O&M activities. While the mechanics of the proposed operation are still being worked out, some of the principles to be followed are: (i) the loans would have to be guaranteed by collateral; (ii) WUAs or concessionaires, not individuals, would be the final borrowers of the funds; (iii) the funds would be lent at commercial interest rates; (iv) WUAs would need to demonstrate their ability and willingness to repay the costs of rehabilitation and to cover the incremental O&M costs; and (v) the WUAs would be responsible for designing the projects and either the WUAs or the concessionaires for executing the civil works -- the Government would provide technical assistance for this. 13 For projects in poor communities that may have high social but low private returns, (e.g., irrigation rehabilitation, contour terracing, reforestation, pasture improvement), the Government would continue to provide poor communities grant-assistance through its Social Investment Fund, FONCODES. Under this program, communities organize themselves to identify and implement small-scale infrastructure investments, contributing to the costs of the investment, usually in the form of unskilled labor, and organizing themselves to operate and maintain these investments. 14 The strategy for the on-going large public hydraulic projects is also consistent with the principles above. By the end of 1993, these projects had already consumed US$3.4 billion (in constant 1993 dollars) in public funds. In order to complete them, an additional US$8 billion would be required, of which US$7 billion would be needed for the coastal projects alone. Since this is substantially more than Peru's total annual public budget expenditure, both capital and current, it is becoming increasing apparent that without substantial private involvement, most of the projects will never be completed. 15 These projects (or stages of projects) may be divided into four categories. The first category include those that are largely complete and have established WUAs (initial stages of Chira-Piura, Tinajones, Jequetepeque, Majes, and Chavimochic). Under the proposed strategy, most of the hydraulic infrastructure would be transferred to the Associations without charge just as for the completed schemes above (and for the same reasons). The WUAs would then be responsible for completing any outstanding works and for operating and maintaining the system. However, any unallocated land and water rights and any infrastructure needed for further development would be auctioned (through competitive bidding). The Government would provide  viii association members with technical assistance and would make available collateralized loan financing and/or matching grants to the extent that there exist activities with positive externalities or of a public goods nature. 16 The second category includes projects where, because of past large public expenditures, the incremental financial rates of return of investments are acceptable (Chinecas, Pasto Grande, Tacna, and the new stages of the first category). Here, the proposed strategy calls for the sale of the hydraulic infrastructure and unallocated land and water rights to private investors through a transparent process that involves competitive bidding. In addition, a regulatory framework for the privatized enterprise would be prepared that will allow for the "unbundling" of water services so that the investing consortium will be able to sell, for example, the non-consumptive rights to a hydropower generation company and the consumptive rights to a urban water supply company and a land developer. The regulations would also define any monopolistic pricing rules, define whether and for how long any monopoly rights will be awarded, and define how fees for system O&M will be determined. If there are to be any public subsidies to address market imperfections, these would be clarified and the financing mechanism spelled out. The privatization would free-up scarce public resources and increase the likelihood that the project will be completed faster and cheaper. 17 The third category includes projects that are at an early stage of implementation (Olmos, Alto Piura, Sur Medio). Their incremental costs per hectare are estimated to be much higher than their expected benefits as reflected in the projected increase in land prices. There is little chance that private investors would be willing to take over the schemes without enormous, and totally unjustifiable, government subsidies and guarantees. Under this approach, the projects would be shut down for once and for all, thereby saving Peru tremendous fiscal resources as well as stopping the wasteful lobbying efforts from contractors, financing agencies and potential beneficiaries. 18 The fourth category includes those projects with a strategic or social rationale (Tumbes, Rio Cachi, and all the Selva projects). Here, the strategy suggests that the Goverrnent first evaluate whether the non-economic objectives for each project are still valid. If so, the Government would commit sufficient budgetary funds to implement them as rapidly as possible, using local participation in the design and implementation and cost sharing with beneficiaries to the extent possible. If the Government no longer regards these projects as priority, it would see whether it can transfer such projects to beneficiaries to complete the works and operate and maintain the project. In such cases, the Government may provide credit guarantees or matching grants. If there is no interest from beneficiaries despite such efforts, the Government would shut down the projects.  Introduction 1 Peru's water resources are not being used well and its public irrigation infrastructure, which has been costly to build, is not being operated or maintained well. A lack of incentives to conserve water, combined with poor institutional capacity to plan, assign, monitor and enforce water distribution and use, has led to a wasteful use of water. Furthermore, the lack of secure water rights has resulted in conflicts over water and discouraged private investment in irrigation and in water-intensive activities. At the same time, many of the large public investments in hydraulic projects, which were supposed to make Peru self-sufficient in agriculture, have proved to be very uneconomical and a large fiscal drain. Despite various policies aimed at ensuring adequate operations and maintenance (O&M) of irrigation infrastructure, the operation of most irrigation systems is poor and the maintenance so inadequate that many systems face a high risk of failure. 2 The objective of this paper is to identify an irrigation sector strategy to address the above-mentioned problems. Resolution of these problems is expected to increase output and incomes, thereby helping to reduce poverty. Although the focus of the paper is on irrigation, some of the reforms discussed here will have a profound effect on other uses of water (water for domestic use, industry and mining, and hydropower) as well as on public finances. The primary audience for this report is Peruvian policy-makers, both in the executive and legislative parts of the Government. In addition, this report is expected to be valuable to the local and international community involved in Peru's development efforts. Finally, since there are some novel aspects to the strategy, the report is expected to be useful to policy advisors, researchers and policy-makers working on other water-scarce countries as well. 3 The structure of this report is as follows: Chapter 1 provides an overview of irrigation development in Peru. It describes Peru's land and water resources, discusses past agricultural policies, provides a brief history of irrigation development, and analyzes how Peru's irrigation policies and institutions have affected water use and irrigation development. Chapter 2 describes the proposed water law which would establish a system of well- defined property rights to water that can be traded relatively freely. It shows how effective implementation of such a law could increase the productivity of water use, help improve the O&M of existing irrigation systems, resolve water conflicts, stimulate private investment, and benefit the poor. It then suggests some reforms to the law and discusses how best to implement the legislation. Chapter 3 discusses a strategy for the rational development of irrigation in Peru. It suggests measures to improve the effectiveness of existing irrigation schemes and to ensure the rational development of on-going and future irrigation schemes. Finally, Annex A provides an English translation of the May 1994 version of the draft water law, Annex B focusses on water reform and poverty alleviation in the highlands of Peru, and Annex C summarizes key characteristics of the large coastal hydraulic projects.  -2- 1. Overview of Irrigation Development in Peru A. Agricultural Land and Water Potential and Use 1.1 Peru's twenty-three million people inhabit an area of 129 million hectares, greater than the combined areas of France, Italy and Germany. About 10 percent of Peru's area is located in a narrow coastal strip between the Pacific ocean and the foothills of the Andes (Costa); 30 percent is mountainous (Sierra), made up of the valleys and western plateaus of the Andes; and 60 percent is in the Amazon watershed (Selva), comprised of the valleys and eastern plateaus of the Andes (Ceja de Selva) and the low Amazon rainforest (Selva Baja). 1.2 The Costa is largely arid with an annual average rainfall of only 38 mm. However, it enjoys moderate temperatures and fertile soils and receives water from 53 rivers that flow from the Andean mountains. It also has limited underground water in some areas. The Sierra receives an average of 600 mm. of rainfall but has a harsh topography and climate. The Selva suffers from torrential rainfall (annual average 3500 mm), fragile ecosystems, acid soils, pest problems, and is isolated from the Pacific and Atlantic Oceans by the Andean mountains and Amazon rainforest. As a result, the potentially cultivable area in Peru is estimated at only 6 percent of its total area. 1.3 The availability of surface water in Peru is estimated at over 2 trillion cubic meters, equivalent to a runoff of 64,800 cubic meters per second. This represents 5 percent of the world's runoff. Although water resources are abundant in national terms, they are geographically poorly distributed in terms of agricultural needs, which constitute about 80 percent of total water demand in Peru. The bulk (98 percent) of the country's water supply is located in the Amazon basin where the potential for agricultural expansion is limited while the area of greatest potential expansion, the Costa, is entirely dependent on uncertain and highly variable seasonal supplies from rivers originating in the Sierra -- over three-quarters of the water flow is between January and April. 1.4 Peru's development of water resources and agriculture reflects its regional characteristics. The scarcity of surface water in the fertile coastal region and Peru's concern with food self-sufficiency has led successive Peruvian Administrations to invest heavily in irrigation systems to regulate and divert surplus water from Andean basins and convey and distribute it to coastal users. Some of these systems are hundreds of years old. Currently some 717,000 hectares are irrigated in 53 valleys in the coast. Of these only about 320,000 hectares have permanent surface water resources for irrigation purposes. Nevertheless, given its better access to domestic and international markets, as well as its milder weather and fertile soils, the Costa accounts for 60 percent of Peru's agricultural GDP. The region's most important  -3- agricultural outputs are cotton, rice, sugarcane, yellow corn, beans, and poultry. Although the region has significant potential to develop non- traditional exports such as fruits and vegetables, it has only recently begun to diversify into some high-value export crops such as asparagus. 1.5 The majority of usable land in the Sierra is suitable only for grazing. However, the region does have about 310,000 hectares of irrigated land, mostly in very small irrigation schemes of 100 to 3,000 hectares that are managed directly by community organizations (See Annex B). The Government is also executing three large projects in the Sierra, oriented mainly at irrigation. In spite of such irrigation efforts, crops on almost 80 percent of the 1.9 million cultivable hectares are rainfed. This is due to predominantly mountainous terrain which lends itself mostly to cultivation of small plots on the mountainsides of the Sierra. The region accounts for about 25 percent of agricultural GDP, with important crops being potatoes, white corn, wheat, barley, and several native crops. 1.6 Agriculture in the Selva is primarily in the valleys of the Ceja de Selva, at altitudes of 500 to 1500 meters. Of the one million hectares of cultivated land in the Selva, only ten percent is irrigated and three hundred thousand hectares are utilized as pasture land. The bulk of the area remains under rainforest (72 million hectares). The region produces 15 percent of legal agricultural output, with the main crops being coffee, cocoa, rice, yellow corn, and cassava. During the last 15 years, this region has witnessed tremendous growth in the cultivation of coca, mostly for illegal export in the form of paste, base or cocaine. Finally, the Selva has significant un-exploited potential to diversify into non-traditional exports such as palm oil and Brazil nuts. However, erratic government policies, a deterioration in public infrastructure, and violence spawned by the illegal drug trade has severely hampered the development of this region. B. Synopsis of Agriculture Policies 1.7 While Peru cannot be characterized as an agricultural economy, its agricultural potential is far greater than has been exploited. Peru's agricultural share of GDP fell from 22 percent in 1950 to 13 percent by 1970. By 1990, it stood at only 6 percent of GDP at current prices, although almost a third of Peru's labor force was involved in agriculture. While part of the decline can be explained by unfavorable terms of trade, erratic and inappropriate government policies have played a major role. There were five basic policy regimes during the last half of this century. The first period, during the 1950s and through most of the 1960s, was characterized by little government intervention in the sector. Agricultural GDP grew by 3 percent annually over this period. However, there was widespread social unrest, mainly because of growing income disparities between rich and poor. In 1968, following a military coup, General Velasco came into power and began  -4- implementing a far-reaching land reform program, known as the Agrarian Reform, aimed at redistributing wealth and reducing income inequality. The second period from 1969 to 1978 saw polices aimed at redistribution and state control. 1.8 Under the Agrarian Reform, about 9.5 million hectares of agricultural land, previously held by 10 thousand owners, were expropriated and transferred to over 400 thousand families. The holdings represented about 85 percent of all privately held land. There were few large individually-held farms left following the land reform, especially in the richer coastal area. The 1984 rural household survey showed that only 1 percent of individually-held coastal farms were greater than 20 hectares and that only 3 percent of the land area was in farms of greater than 50 hectares. In order to try to maintain technical efficiency and economies of scale, the large land holdings were not broken up into small units but were organized into cooperatives. The Agrarian Reform prohibited the use of land as collateral (except to public development banks) and disallowed corporations from owning rural land, thus limiting vertical integration of agro-industry. With the exception of some privately held land, agricultural land transfer and rental were effectively prohibited in Peru. These restrictions severely affected the efficiency with which land could be used and inhibited innovation and private investment. 1.9 In addition to the land reform policy, Velasco's agricultural policies included: (a) state control of production, marketing, and pricing of major food crops and agricultural inputs; (b) industrial import substitution, which led to a general erosion in agricultural terms of trade with the rest of the economy; (c) reorganization and disruption of extension and research activities; and (d) the initiation of massive public investments in hydraulic infrastructure, some of which have yet to be completed. Although the policies were aimed at food self-sufficiency, agricultural output grew by only 0.5 percent annually during the 1970s. Similarly, although the land reform was meant to reduce poverty, rural poverty actually worsened. While land reform succeeded in reducing incomes of the richest and in temporarily increasing incomes of those that received land, the landless poor became worse off. 1.10 When General Morales Bermudez ousted Velasco in a coup in 1975, he initially continued with the thrust of Velasco's policies. However, in 1978, following large fiscal and external imbalances, the Morales Government implemented a stabilization and liberalization program. These market-oriented reforms were deepened by the civilian president Belaunde, who was elected to a five-year term in 1980. The key reforms affecting agriculture included elimination of nearly all quantitative import restrictions and significant import tariff reductions; improvements in agricultural research and extension; and the introduction of legislation allowing cooperatives to be broken up into independently owned and managed smallholder units. However, State control of agricultural marketing and credit continued. Agricultural growth during this  -5- period was sporadic. While there were some reversals in trade liberalization, the third period (1978-1985) can be characterized as one of market liberalization efforts. During this period, terrorist activities of the Shining Path and MRTA began to take hold, especially in the Sierra, and the illegal cultivation of coca for export increased rapidly. 1.11 Upon being elected President in 1985, Alan Garcia began a heterodox stabilization program. This fourth period (1985-1990) consisted of renewed state controls and widespread subsidies. An important component of the Garcia Administration's program was price controls that discriminated against agriculture, accompanied with other input, output and credit subsidies to help farmers. Some large hydraulic projects were also initiated during this period. Although the heterodox program was initially successful in reducing inflation and stimulating growth, it was unsustainable and the subsidies had to be abandoned as the fiscal situation deteriorated and inflation reached 1700 in 1988, 2800 percent in 1989, and accelerating further during the first seven months of 1990. The early gains in growth were also rapidly eroded as GDP fell by 7.4 in 1988 and an additional 12.4 percent in 1989. Moreover, it was the richer farmers that had benefitted from the subsidies, while the price controls and multiple exchange rate system had discriminated against the poor. This period also witnessed a sharp increase in coca cultivation and terrorism. 1.12 Since his election in 1990, President Fujimori has implemented a series of broad-based reforms to restore macroeconomic stability and long-term economic growth. The budget was rapidly brought under balance; inflation fell steadily so that by 1994, it had been brought to an annual rate of 15 percent; gross official international reserves rose dramatically to US$ 7.1 billion; and GDP grew by 6.5 percent in 1993 and by 12.9 percent in 1994. The main reforms affecting agriculture included: macroeconomic and trade policy reform to improve the fiscal balance and eliminate import duty exemptions and quantitative restrictions; * pricing and marketing reforms which dismantled consumer price controls and the costly guaranteed producer price mechanism and replaced them with a variable import surcharge scheme; * new land tenure legislation which removed restrictions on the sale or lease of agricultural land, allowed landowners to mortgage lands in excess of 5 hectares, raised maximum landholding limits, and greatly simplified land registration procedures; * rural financial reforms that eliminated preferential agricultural interest rates, allowed banks to freely set interest rates, and liquidated the Agrarian Development Bank, which had contributed to inflation and benefitted the wealthier farmers disproportionately;  -6- * research and extension reforms which drastically reduced staff, converted research and extension stations to private foundations (composed of associations of agricultural producers, exporters, extension agents and NGOs) so as to make these activities more responsive to farmer needs, and provided public funding for foundation research and extension staff; coca eradication efforts which involved farmers in the design and execution of coca substitution programs and included land titling programs, and improved interdiction efforts; and irrigation policies described below. C. Irigation Policies and Institutions Water Rights Legislation 1.13 The use of water in Peru today is determined by the 1969 General Water Law (Legislative Decree 17752) along with the 1991 Agricultural Investment and Promotion Law (Legislative Decree 653) and their respective regulations. The 1969 law abrogated a 1902 law under which water rights were privately owned''. Under the law currently in force, which stems from the 1969 law, all water resources in the country belong to the State. They are not susceptible to either private ownership or transferable rights. The use of water is subject to specific social and development needs. The law prescribes preferential uses of water, in the following order: (i) for human consumption and basic necessities; (ii) for livestock development; (iii) for agriculture; (iv) for the energy, industry and mining sectors; and (v) for other uses. 1.14 Water resource allocation amongst the different uses is made by granting: (i) licenses, for permanent uses; (ii) permits, for temporary use of excess water and, in agriculture, for annual crops (cultivos transitorios); and (iii) special authorization, for the implementation of specific studies and/or works. Private investment by an individual in groundwater development (e.g., construction of wells on private property) provides the investor with an assured use of this water, without interference from the State. Licenses and permits are revoked or declared null and void when: (a) the use of water is transferred to other parties or differs from that originally prescribed 1/ However, under the 1902 law, the water rights were given in conjunction with land rights and hence could not be traded separately from land.  -7- (revocaci6n); and/or (b) the water tariff has not been paid for two consecutive years or when water is not used according to the cropping and irrigation plans (caducidad). 1.15 While the 1969 law focusses exclusively on water and considers water in all its uses, the 1991 law has only a small water-related component aimed at making the irrigation sector more modern and efficient. A key aspect of this component is that it fully transferred responsibility for the management and financing of all completed irrigation systems to water user associations2'. However, the law does not clearly define user's rights with regard to the irrigation infrastructure. This law also specified how water tariffs were to be implemented and established a new administrative structure at the water basin level (see below). 1.16 The current system of priorities for water use has some conceptual and practical shortcomings. It is not sensible that a high-value use for water such as for hydropower generation or industry should be denied water rights in favor of a lower priority activity such as livestock development or agriculture. Similarly, since new water rights can, in principle, be obtained without charge (except for the annual water tariff) by confiscating them from other lower priority uses, there is less incentive for water companies to reduce water losses. In practice, there is no mechanism to transfer water rights from one user or sector to another to meet priorities and changing demands. For example, although legally permissible under the law, it is in practice very difficult to expropriate irrigation water from farmers to give to a neighboring city. Thus cities such as Lima suffer from water shortages whereas farmers in the nearby rural areas continue to grow water-intensive agricultural crops. Furthermore, since water tariffs are very low and unrelated to the volume of water used (irrigation charges are based on area under irrigation), the system provides little incentive to conserve water. 1.17 There are other practical reasons for why the current system has failed to rationalize water use. Licenses for water use are often obtained from public authorities (see below) even when it is clear that the available water in a stream or canal is already fully allocated. It is not unusual to find that the amount of water rights along a water course is twice the amount of water available in a normal year. This intensifies water conflicts and leads to uncertainty over water availability. Finally, the current law, by giving land owners to unlimited rights to underground water, ignores the linkage between ground water and surface water even when the well is close to a river. It also provides inadequate protection against aquifer depletion. 2/ O&M for the large hydraulic projects ("proyectos especiales") that have not yet been fully completed continue to be the responsibility of project authorities under the direction of INADE (Instituto Nacional de Desarrollo)  -8- Institutional Legal Framework 1.18 The 1969 and 1991 laws mentioned above entrust the management of water resources to public institutions at different levels: national, regional and local. 1.19 National level: Since 1902, water resources management has been the responsibility of the General Directorate of Water and Lands (DGAS), initially under the Ministry of Public Works and, since 1968, under the Ministry of Agriculture (MAG). The role of DGAS, however, has been limited to the agricultural sector, leaving other Ministries (such as Fisheries, Health, Energy and Mines) to regulate the use of water in their own sectors. Although a multi-sectoral Water Council (Consejo Superior de Aguas) was established by the 1969 General Water Law, this council has apparently never met. Since the promulgation of the a law reorganizing MAG in December 1992, DGAS has been a Directorate of the National Institute of Natural Resources (INRENA), which is in charge of the rational and integral management of renewable natural resources within MAG. 1.20 Regional/River basin level. The 1991 Agricultural Investment and Promotion Law established regional inter-sectoral water entities. Officially under MAG, these Autonomous Hydrological Basin Authorities (AACHs) are to be responsible for formulating master plans for the rational management of natural resources and for implementing irrigation and conservation activities in their specific areas. An AACH consists of a Board chaired by the irrigation district's Technical Administrator. Its members are the representatives of regional governments, the Ministries of Energy and Mines, Housing and Construction, the National Development Institute, and five representatives of producer groups or associations. As of October 1993, only three AACHs had been established (in Jequetepeque, Chira-Piura and Chancay-Lambayeque). 1.21 Local level: At the local level, responsibility for irrigation rests with an Technical Administrator (Administrador Tecnico), who is appointed by and reports to the Ministry of Agriculture. The Technical Administrator is to: * ensure the rational and efficient use of water resources; * approve cropping and irrigation plans and supervise their implementation; * authorize and approve the studies and infrastructure construction associated with requests for licenses and permits for water;  -9- * issue water licenses and permits; * approve and keep water use registers up to date; * establish, modify or cancel water rights; * impose restrictions on water use for conservation purposes; * resolve conflicts amongst water users; * support and approve the creation of water user associations; * propose and set water tariff levels; and * approve plans for the O&M of the irrigation systems. 1.22 The Technical Administrator oversees an "irrigation district", and its Water User Association (Junta de Usuarios). There are 97 irrigation districts in the country, each including one or more river basins and/or sub- basins. Water User Associations (WUAs) are responsible for the development, conservation, preservation and rational use of water and land resources, in coordination with the Technical Administrator. The WUAs are private organizations and are legally entitled to collect tariffs, operate and maintain the irrigation works, and to take out loans to undertake works and studies, either directly or through contractual agreements. They also have the authority to cut off water to users under certain conditions (see below). Membership, statutes and by-laws of these associations are regulated by law. WUAs are to be comprised of representatives of all users, agricultural and non-agricultural, within a district, and theoretically, only one WUA can exist within each irrigation district. In practice, there is sometimes more than one WUA in a district and non-agricultural users do not participate in the WUAs, which generally comprise only Irrigator Commissions (Comisiones de Regantes). Currently at least one WUA exists in each of the 97 irrigation districts. 1.23 Conceptually, the formal establishment of the AACHs could help to ensure that water resource management decisions within the same hydrological basin are made in a concerted manner among water users and between the public and private sectors. Currently, however, the structure of the AACH may limit its effectiveness. There is too much public sector control in the AACH, which is likely to inhibit dynamic and voluntary private participation. The Technical Administrator designated by MAG is the President of the AACH. This is inappropriate because the Technical Administrator deals only with irrigation while the AACHs have a wider, inter-  -10- sectoral function. In addition, the Technical Administrator is low within the public hierarchy. Moreover, the Technical Administrator, as President of the AACH, is placed in the powerful position of considering appeals on cases on which he as ruled in his capacity as Technical Administrator. It would be more appropriate for the President of the AACH to be elected by its members, with the Government either abstaining from voting or having only a minority vote. Finally, the funding for the AACH's operations is uncertain. AACHs' financial resources are to come from: (i) the water levy or "canon," which is only a small fraction of the low irrigation water tariff; (ii) loans, grants and any revenues that the AACH can generate for themselves; and (iii) contributions from the various groups that form the AACHs. 1.24 At the irrigation district level, the Technical Administrator receives little technical support and few human and physical resources to carry out his functions. This technician generally operates alone, with an extremely limited resources. As a result, water resources are poorly managed at the irrigation district level. Situations of conflict often arise, as out-of-date water user registers do not correspond to the actual use of water in the district while new licenses and permits are issued without due regard for the available water resources, often resulting in over-exploitation and severe shortages during periods of drought. Water Tariffs 1.25 The existing legislation prescribes two classes of water tariffs, one for agricultural use and the other for non-agricultural use. Neither reflects the true cost of water. In agriculture, the water tariff includes three components: * A so-called water user association component ("ingresos junta de usuarios"), to meet O&M expenses and to finance the operating budget of the Administrador Tecnico. * A water levy or "canon de agua" component, representing 10 percent of the first component. * An "amortization" component, to recover the cost of public investment in storage structures ("obras de regulaci6n") of irrigation systems. This component is to represent also only 10 percent of the water user association component, except when otherwise determined by "special" project authorities.  -11- 1.26 Irrigation water tariffs (tarifas) are proposed by the Technical Manager to the water user association and in case of disagreement, are decided by an ad-hoc, informal committee. In addition, special water levies (cuotas) are collected by the WUA or Irrigator Commission as mandatory contributions from the water users for specific works or activities in the irrigation system or sub-system. According to the existing law, these cuotas are to meet only extraordinary expenses. In reality, as the tarifas are too low to cover O&M expenses, the cuotas often constitute the only means for a WUA or Irrigator Commission to operate and maintain its irrigation system or canal. Since the transfer of irrigation infrastructure O&M by the Government to water users, the cuotas have helped to improve O&M of canals and irrigation systems. 1.27 As the portion of the water tariff allocated to amortization is restricted to only 10 percent of the amount required for O&M, the amount is unlikely to recover construction costs. In fact, water tariffs actually paid by farmers account for only 1 to 3 percent of the total construction costs. Although the water tariffs are low, an estimated 30 percent of farmers do not pay their dues. This is probably due to: (i) poor organization of water user associations; (ii) perception of the State as provider; (iii) little or no involvement of beneficiaries at the time the schemes were built; (iv) frequent conflicts and uncertainty over the use of a resource that is often unavailable; and (iv) minimal penalties in case of default. D. The Current State of the Irrigation Sector 1.28 While the institutional and policy framework described above appears rational, in practice it has proved unworkable. This has been partly due to the inability to set and collect adequate water tariffs, which has resulted in: (i) a lack of financial resources to operate and maintain the irrigation systems adequately, so that most of these systems are now in an advanced stage of deterioration; (ii) excessive water consumption at farm level, which may be due both to the lack of an incentive to use water efficiently and to high conveyance losses; (iii) preference for crops (such as sugarcane and rice) with large water requirements; and (iv) few incentives to move towards more efficient irrigation techniques such as sprinkler or drip irrigation. 1.29 It is estimated that at least one half of the irrigated area in the Costa is badly in need of rehabilitation and modernization. Furthermore, excessive use of water in the upper reaches of irrigation systems has also contributed to soil salinization problems in the lower parts of the valleys. An estimated two to three hundred thousand hectares in the Costa are  -12- detrimentally affected by waterlogging and salinization. This is almost equal to the planned new area to be brought under irrigation from the completion of the all of the expensive large irrigation schemes in the Coastal Region (Chapter 3, Table 1). 1.30 Scarcity coupled with misallocation of the water resource within the same river basin or sub-basin has created conflicts amongst the different water users (for water supply or electricity versus irrigation, e.g. in the Arequipa region and in particular the Zamacola-Chili Regulado irrigation area). In certain cases this has led to the illegal multiplication of WUAs, and in some irrigation systems such as La Joya, more than one user group has been established. Similarly, while some of the WUAs are strong and dynamic, the majority of them are weak, organizationally and technically as well as financially, and are unable to maintain the irrigation infrastructure adequately. As a result, the infrastructure continues to deteriorate, with some infrastructure facing a high risk of failure. 1.31 In the Sierra, some 310,000 ha are currently irrigated, mostly under traditional communal administration. Traditional irrigation practices and difficult climatic, topographical and geotechnical conditions have constrained the optimal use of the soil and water resources available in the region (see Annex B). This situation is aggravated by the way the current water law is implemented, by the fragmentation of the rural properties and by the lack of appropriate agricultural extension services in the region. 1.32 The policy of having the Government design, evaluate and execute the large hydraulic systems has proved to be very costly and has failed to bring about even a small fraction of the estimated benefits. A primary example is the Majes project, which was started in the 1970s. Although about US$ 1 billion (in historical dollars) has been spent on this scheme, it has achieved only one-sixth of its targeted new irrigation area and none of its planned hydropower potential. Perhaps the biggest reason for its failure was that this project, like most others, was selected on political rather than economic considerations. Furthermore, project implementation suffered from delays in the budget release process due to macroeconomic imbalances and a changing of priorities for the project with each new Administration. Moreover, there were few incentives for project managers to control costs or expedite construction as well as reports of corruption in the awarding of contracts. These problems are typical of Peru's large hydraulic projects.  -13- 2. Tradable Water Rights for Improved Water Mnagement A. The Draft Water Law Introduction 2.1 By 1992, following several years of virtually no public spending for maintenance or rehabilitation of public irrigation structures, many public irrigation systems faced a high risk of failure. Water delivery became more irregular, water quality deteriorated and water conflicts grew, both within agriculture and between the various water users. There were also reports of widespread water theft. Although water was scarce, it was often used wastefully or in low productivity activities. Despite low water tariffs, many users were unwilling to pay them and the Government realized that it would be extremely difficult to increase water charges substantially in order to generate funds for system operations and maintenance (O&M). Even if water tariffs could be increased substantially, they would fail to provide incentives for water conservation since the tariffs were based on irrigated area rather than volumetric measurement. At the same time, the Government realized that it would be very difficult to create strong institutions to plan, assign, monitor and enforce efficient water use. Moreover, the Government was unwilling to devote substantial public resources to subsidizing irrigation construction, rehabilitation or maintenance as it had done in the past. 2.2 Not only had the system of publicly allocated water rights failed to provide incentives to conserve water, to use water for high-valued purposes, or to carry out adequate O&M activities, it discriminated against the poor and caused insecurity to users, thereby discouraging investment in water-intensive activities and water-related infrastructure. Since valuable water rights have been given without charge, with only modest tariffs to be paid subsequently, there has been tremendous demand for additional water rights. With water rights being allocated at the discretion of public officials, it is not surprising to find that the richer, more influential farmers were more likely to obtain the new water rights, even when the water along a watercourse was already fully allocated. This resulted in a proportional reduction in the availability of water to other farmers, typically the poorer ones, and led to water conflicts. In the Sierra, where the bulk of the small poor farmers reside, richer farmers were able to use their influence over the Technical Managers to receive water out of turn (Annex B). 2.3 In addition, the threat posed by the ability of the State to divert water for higher priority uses under the existing law has probably discouraged new investments in water-intensive activities. An export-oriented farm or an industrial company that requires large amounts of water is not likely to invest  -14- if there is a threat that their water rights could be taken away without compensation in future for some "higher priority" purpose. Indeed, some beer companies in Lima have been obliged to pay the Lima Water Company substantial sums in order to prevent the water company from exercising its priority right to water. 2.4 In order to address the problems above, several Government officials, under the leadership of the Minister of Agriculture, realized that a substantial reforn of the water sector, including the 1969 water law, was required. The Peruvian Government sought assistance from the World Bank and Inter-American Development Bank (IDB) to understand world-wide experience with water markets. Responding to this request, the World Bank invited selected congressmen, senior officials and private sector executives to a workshop in Washington to learn from international experts about the experience of other countries in the sale of water rights and the development of irrigation systems. The objective of the workshop was to improve the way in which water was used in Peru, to improve the O&M of existing irrigation systems and to ensure the rational and cost-effective development of large hydraulic projects. International experts discussed the advantages and disadvantages of water markets and the experiences of Chile, Mexico, United States, France and Spain in irrigation and water markets. The Peruvian Government was particularly interested in the 1981 Chilean water law, which appeared to be working well and which had a private sector orientation. Under the Chilean law, private entities are allowed to buy and sell property rights to water with few restrictions. 2.5 Although there were differing points of view expressed at the workshop, the Government of Peru came away with the impression that some modified version of the Chile water law would be suitable for Peru. It therefore used technical assistance funds under an IDB loan to hire consultants to prepare a draft water law based on the Chilean model. To allow effective implementation of such a law, the Government also introduced into the 1993 Peruvian constitution a clause (Article 66) that treats water and land rights equivalently, thereby opening the way for property rights to water. The submission of the draft water law to Congress has been delayed because of opposition to the law. This opposition stemmed partly because there was insufficient participation of some important stakeholders at the design stage. A lack of full dissemination of the law during the early phase also led much misinformation about the law, emanating particularly from those who are opposed to Fujimori's market oriented reforms. By clarifying the law and its rationale in this chapter, and by pointing out the difficulties with alternate ways of addressing water problems, it is hoped that the legislature will be able to make a more informed decision on the design of a new water law.  -15- Key Characteristics 2.6 This section highlights some of the important characteristics of the law and provides a rationale for the law. For greater clarification and discussion of issues not mentioned here, the reader should refer to the unofficial English translation of the law provided in Annex A (May 1994 version) or the original draft law in Spanish. Assignment of Terrestrial Waters 2.7 The law deals with the use, preservation and conservation of terrestrial waters found in liquid state, either surface or ground water, either flowing or standing. While water is a national asset, the right to use it is a property right which consists of the authority to hold or use it in an exclusive manner and designate it for any purpose, as stipulated by law. Thus the rights do not have to be used for any specific purpose; nor are there any priorities among water rights for different purposes. A user may only use water in an exclusive manner to the extent that it is the holder of the corresponding water rights. However, provided that surface waters are found at their natural source, anyone can drink from, bathe, navigate or water animals in them, in observance of municipal or other legal regulations. 2.8 The water rights may be for consumptive or non-consumptive use, temporary or permanent use. Holders of consumptive rights do not need to have the water returned to the river system. Holders of non-consumptive rights must return the same quantity of water to the river system (allowing for normal losses). In order for the right to allow water of inferior quality to be returned or to alter the normal runoff conditions or the timeliness with which water flows down the river, any injured parties must receive compensation. Temporary water rights may be exercised only when the natural source is able to supply permanent rights fully. If there is not sufficient water for all holders of temporary rights, preference will be given to upstream users in the case of existing temporary rights. In this, the draft law simply acknowledges the system in the prevailing law. However, for new temporary rights, preference will be based on the seniority of rights. 2.9 Water rights are established by the Watershed Director either upon its own initiative or at the request of interested parties, provided that the water is available and no rights to it currently exist, that third parties are not affected, that there is enough water to maintain a minimum ecological flow and maintain the accustomed quality of life in cities and towns. Water rights may also be obtained by adverse possession under the civil code (prescripci6n). For lakes, the water rights are established with respect to waters above minimum levels. For groundwater, the Deputy Watershed  -16- Director may declare certain areas restricted from new groundwater rights if data suggests that increased exploitation of the aquifer would jeopardize third party water rights. Once these conditions are met, the Deputy Watershed Director publicly auctions water rights to new or currently unallocated water. However, water that has already been allocated to users (via licenses or via custom) are to be given without charge to those already receiving water (see Final Provisions below). 2.10 Water rights are acquired by being recorded in a special public registry (Water Rights Registry), specifying, inter alia, the name of the title holder; the name of the riverbed or natural source; the flow or volume (which may be specified in terms of percentage of stream flow or in shifts); the point at which the water will be diverted; whether it is for consumptive or non- consumptive use and whether it is for permanent or temporary use; the point and form in which the water will be returned to the river system; and the amount paid for the rights. The supply source or intake point may be changed at the request of the holder of water rights, provided that third parties are not injured. 2.11 The water right is separate from the land right for both surface and ground water. However, to obtain ground water rights, one must either own the land on which the well is to be built or have the owner's explicit authorization. Water rights are transferred by registering the respective title with the Water Rights Registry. The establishment and transfer of mortgages or other rights on water also require registration. The sale or mortgage of land does not include water rights, unless explicitly stated otherwise, and the mortgager is owner of such rights. Institutional Matters 2. 12 Under the draft water law, water distribution according to established water rights is to be done by the various types of Water Users Organizations (WUOs). WUOs include the Watershed Users Committees, (currently known as the WUAs), with jurisdiction over each watershed, and the Canal Commissions (now known as the Irrigator Commissions), with jurisdiction over canals or other hydraulic works. In addition, there may be WUOs for underground aquifers and for drainage works. There may be no more than one WUO for each watershed, canal, underground aquifer and drainage work. The Watershed Users Committee are to also perform hydrological measurements and studies and protect riverbanks and riverbeds. 2.13 These organizations must be legally recognized entities, registered in a Registry of Water Users Organizations, and they are governed by bylaws approved by a Board of Directors. At meetings, votes are issued in  -17- terms of volume of flow, with non-consumptive rights having one-third the weight of consumptive rights per unit of volume. Members of WUOs have rights and obligations, including organization dues. After two reminders, the Board of Directors has the authority to deny water allocations to those members who are late in paying. Compliance with resolutions taken by the Boards of Directors is mandatory for these organizations. 2.14 The National Water Council, the Watershed Directors, Deputy Watershed Directors and the Public Registry of Water Rights has administrative jurisdiction over water. The law specifies the minimum qualifications of the Director of the Water Council, the Watershed Director, and the Deputy Watershed Director. The National Water Council responsibilities include, inter alia, establishing criteria for the initial allocation of water rights, issuing environmental and other regulations needed for the proper understanding and enforcement of this Law, and appointing Watershed Directors. The Council possesses technical, economic, budgetary and administrative autonomy and is comprised of the representatives of the Ministries of Agriculture; the Presidency; Energy and Mines; Industry, Tourism, Integration and Trade; Health; and Economy and Finance. 2.15 There is to be one Watershed Director for each one of the five Watershed Bureaus to be established under the law. The Watershed Director is responsible for: (i) establishing the original water rights in accordance with the regulations to the law; (ii) authorizing the transfer or water from one watershed to another; (iii) authorizing the construction or modification of intakes, retaining walls, dams or any other works in riverbeds; and (iv) verifying river depletion. The resolution authorizing or denying the transfer of water from one watershed to another is to consider and mention not only the availability or water and damage to third parties, but also how the transfer will affect the recharging of the ground water table and the natural enrichment of the lower parts of the original watershed. However, the statements in this law do not affect enforcement of the Environmental Code and other related laws. Other important functions of the Watershed Director include: (i) coordination with public or other authorities with water-related expertise in the watersheds under their jurisdiction; (ii) dispute resolution; and (iii) proposing to the National Water Council a list of candidates for the position of Deputy Watershed Director. 2.16 The Deputy Watershed Director reports to the Watershed Director and is responsible for authorizing changes in the supply source and intake point for specific water use rights, for decreeing restricted areas with respect to ground water, determining mandatory rights of way, levying fines and sanctions, and delimiting the banks of riverbeds. He is also responsible for resolving first administrative appeals on disputes, for overseeing the  -18- application of water quality regulations and for hiring persons to perform hydrological studies in the area under his jurisdiction. Fines and Sanctions 2.17 Persons violating this law will receive administrative sanctions, separate from any civil or criminal proceedings. The Deputy Watershed Director shall, by means of administrative resolution, levy fines on those who violate the provisions of this law. The amount of fine shall vary according to the severity of the infraction, the damage caused, and whether the violator is a repeat offender or in default, and shall be repeated as long as the law continues to be violated. The money collected from fines shall be used for watershed protection. Final Provisions 2.18 The law is to become effective 180 days after publication and will revoke all previous laws and regulations dealing with similar issues. The regulations of this law will be prepared by the National Water Council and approved by Supreme Decree, with the Council of Ministers' vote of approval. Supplementary Provisions under the law include the establishment of a property tax on water rights and the clarification that when water regulation and transfer works are built that allow surplus water to be stored, this water becomes the property of the executor of these works. The draft law's Temporary Provisions explain that in consideration of the availability of water resources based on hydrological studies, the Water Authority will allocate the corresponding water volumes to Water Users Organizations. They, in turn, will allocate them to their members according to criteria adopted by the National Water Council. Once the water allocations of each user are determined, the Watershed Director shall validate them and acknowledge corresponding water rights. Holders of acknowledged water rights do not pay for such rights. For Peasant Communities (Comunidades Campesinas) in the Sierra and Selva regions, water rights will be acknowledged based on the average water volumes used historically. Water rights corresponding to real estate that is mortgaged at the time that the law is published will remain legally mortgaged together with the real estate. Operation and Rationale 2.19 Once property rights to water are assigned to existing water users (such as farmers, water supply companies, large industrial users and hydropower companies), those individuals or other legal entities will be free to buy and sell them in the open market at a price determined freely by supply  -19- and demand. They can also lease them in the secure knowledge that such leasing will not jeopardize their rights in future. 2.20 In addition to the price of the rights, purchasers will have to pay the cost of any changes to infrastructure to effect the transfer as well as compensation to any injured parties. The sale price has little relation to the water tariff that must be paid to ensure adequate O&M of the system. To use an analogy from the condominium market, one can think of the price of water rights to be the purchase price for the apartment and the water tariff to be the annual condominium fees. Just as a condominium association decides on its fees for its operations and maintenance activities in line with prudent financial management, the WUO will decide on the level of water tariffs. The tariff revenues will be retained by the WUOs rather than be handed over to a Government agency. This will increase the user's willingness to pay the tariff. Another important difference with the prevailing system is that the tariff will be proportional to the amount of water rights held. Under the prevailing system, water tariffs are typically based solely on irrigated area, which provides less incentives for water conservation. 2.21 With the establishment of tradable water property rights, water will have an implicit value or "opportunity cost", creating built-in incentives to conserve water and put it for the most productive uses. If farmers are able to sell their water rights at freely negotiated prices, some farmers will choose to generate extra income by selling any surplus rights to a neighboring city or to more efficient farms, where the water has a higher value. They can sometimes generate a surplus by using more efficient irrigation techniques or by switching to less water-intensive crops. This is indeed what happened in parts of Chile and in the State of Colorado in the United States. For example, the city of La Serena in Chile, which needed large amounts of water rights to meet its rapidly growing demand for water, purchased 30 percent of its water rights from farmers, rather than contribute towards the construction of a new dam (the Puclara) to obtain additional water rights. At the same time, farmers were able to switch to use more efficient irrigation techniques and to increase their incomes. In this way, Chile was also able to avoid the conflicts that often come about when governments attempt to take water away from farmers to divert for human consumption. Tradable property rights thus helped shift water to higher-value uses in a cheaper and fairer way than such alternatives as building expensive new hydraulic infrastructure, confiscating water from farmers, or substantially raising water charges to force farmers to conserve water. 2.22 A market for water rights also helps reduce environmental damage. By obtaining additional water rights from farmers instead of via the construction of the Puclara dam, the city of La Serena succeeded in postponing  -20- its construction indefinitely, thereby averting possible environmental consequences. Although the conveyance infrastructure to carry out the transfers must exist or be built, its cost and environmental implications are often less than that of generating new water rights by constructing dams, new reservoirs, or tunnels to bring water from another watershed. 2.23 Another environmental benefit is that the incentive to conserve water, such as by farmers switching to drip and sprinkler irrigation, helps control waterlogging and soil salinization, which is caused mainly by over- watering and excess use of fertilizer. As seen earlier, soil salinization is a particularly serious problem in the coastal region of Peru. In addition to helping reduce fertilizer costs and avoid salt build-up in downstream lands, such improved irrigation systems also benefit farmers by allowing a greater surface area in which to plant, reduced topsoil erosion, greater control over the water, and the ability to switch to higher value-added crops. 2.24 Purchasers of water will also have a strong incentive to conserve water. If a water supply company is able to obtain water rights without payment based on its higher priority use, it has little incentive to conserve water. When it has to pay to obtain new water rights, it will try to find ways to conserve water. Thus, EMOS, the largest water supply company in Santiago, Chile found it cheaper to invest in mechanisms to reduce water losses than to pay the higher prices to induce farmers to sell them part of their rights or to build new hydraulic infrastructure. In the past, it had been able to obtain new water rights without charge. Similarly, farmers in Chile that purchased water rights for their export-oriented fruit businesses, typically use efficient irrigation technology. 2.25 Whereas raising water charges substantially so as to cover both O&M and investment costs (its long-run marginal cost) would provide similar incentives to conserve water, it would have vastly different income distribution and social consequences and be politically difficult to introduce. Raising water prices substantially would transfer incomes from heavy water users such as farmners to the Government. Moreover, such a policy requires infrastructure that allows for water measurement and institutions that are able to (i) decide on appropriate water tariffs by area, (ii) adjust them over time as required and (iii) enforce the tariffs -- something that is particularly difficult if water delivery is not reliable. 2.26 Establishing a regime of tradable property rights to water is likely to help the poor, especially when compared to the alternative of raising water charges substantially. Once poor farmers are given property titles to water, they are much more assured that their access to water will not be reduced in future. Under the current system, poor farmers have been  -21- particularly vulnerable to a reduction in their water allocations over time caused by water administrators granting additional rights without charge to politically influential farmers, even when water rights along a certain river, stream or canal were already fully allocated. 2.27 Another advantage of tradable property rights to water is that it can be used as a source of collateral. This is particularly beneficial for poorer farmers who have few assets that can be used as collateral for obtaining access to lower interest loans. In addition, the system provides a larger farmer the flexibility of being able to mortgage only part of his infinitely divisible water rights as opposed to his entire land and water holdings. 2.28 The ability to purchase water is expected to stimulate private investment in activities requiring an assured supply of large quantities of water (e.g. farming, mining, some types of industry, water supply companies) since the uncertainty of access to water will be reduced. This has shown to be the case in Chile where farmers' increased security in their supply of water has led to investment in fruit crops and in efficient irrigation technology. By stimulating private investment and by providing incentives to improve the productivity of water use, both urban and rural poor will also gain indirectly from having increased employment opportunities as a result of increased economic activities. In addition, the ability to buy options on water leases, as occurs in Chile, also has a positive impact on private investment. An options contract will allow a fruit farmer to protect against a drought by paying an annual fee to a neighboring corn farmer for the option of purchasing the corn farmer's water at a pre-determined price. The advantage to the corn farmer is that he will receive an income supplement during the non-drought years and be compensated for the loss of corn production by the sale of water during the drought year. 2.29 Private investment in hydraulic projects will also be stimulated as potential investors become assured that once the water rights are obtained, they will not be expropriated and that the water generated as a result of the investment (e.g. storage reservoirs) can be sold at a freely determined price. Moreover, establishing tradable property rights to water could attract private investment in some of the partially-completed large public hydraulic projects. Once property rights are established, the Government could choose to privatize the large hydraulic projects by auctioning, through a sealed-bid process, the existing hydraulic infrastructure and unallocated water and land rights associated with these projects. The privatization would free-up scarce public resources and increase the likelihood that the project will be completed faster and cheaper. Because of changing Government priorities, budgetary shortages and limited incentive to control costs, these publicly executed and financed projects have run into enormous delays and cost overruns.  -22- 2.30 Presently, there are few incentives and resources for these projects to be operated cost-effectively and maintained adequately after completion. For this reason, even in completed public projects, it is desirable to transfer the infrastructure and water rights to water user organizations, who could then become solely responsible for system O&M and for the setting of water tariffs. Having property rights in water would facilitate this process since it is a logical outcome of a system where the users own tradable water rights and irrigation infrastructure. Virtually all the irrigation infrastructure in Chile is owned by the users themselves and they have found that they are able to carry out O&M activities at a much cheaper cost than the Government. Given Peru's experience with public investment in hydraulic infrastructure, the ability to transfer existing infrastructure to users, privatize on-going projects schemes, and provide incentives for the private sector to execute future projects is valuable (see Chapter 3). 2.31 Through the establishment of the Groundwater Users' Commissions, the law also provides the basis for better protection against overexploitation of underground aquifers. Under the current law, owners of the land above an aquifer have full rights to its water, even if their use were to result in depletion of the aquifer and even if its extraction infringes upon surface water rights. Regulations issued pursuant to the draft water law are expected to limit the ground water rights to the recharge of the aquifer and prohibit the drilling of new wells once the safe yield of the aquifer has been reached. While there are limited opportunities for large-scale exploitation of ground water in Peru, in the few areas where there is substantial use of ground water (e.g., Ica), aquifers are being exploited at unsustainable rates. 2.32 The law has been criticized for discriminating against the poor and indigenous populations, who have limited funds with which to buy new water rights. However, the current system, whereby water rights are given without charge (except for some nominal water tariffs) at the discretion of the Technical Administrator, has led to rights being given to the politically influential, often at the expense of the poor. Experience has shown that when valuable public resources are given for free, it is the relatively wealthy that tend to get them. If the Government would like to provide additional water rights to certain groups, it is far better to do this via direct income transfers, which are transparent and less distortionary, and with which the farmers can purchase water rights. In fact, proceeds from auctions of water rights could be used for such purposes. 2.33 The law has also been criticized for not including provisions that allow for the confiscation of rights if the water is used wastefully or in an environmentally damaging way. Although it is conceptually easier to confiscate rights for misuse under an administered system, the current system  -23- Questions and Answers on the Draft Water Law Q. Since Article 66 of the Constitution water is rationed even in Lima, while in the specifies that natural resources are a national asset outskirts of Lima, farmers grow water intensive (patrimonio de la Naci6n), does not this imply crops such as rice. This is because there is that water, a natural resource, cannot be sold or currently no mechanism to effect the transfer assigned to individuals as a property right? required under the current law. Instead there is talk of building very expensive infrastructure to A. While lawyers disagree on the legal transfer water to Lima from another water basin. interpretation of this Article, for practical The proposed system, by providing incentives to purposes, the authority to use natural resources conserve water, is likely to result in increased subject to certain norms is in fact a property right water availability for city dwellers. Given that (derecho real). If land -- another natural resource water is valued much higher in urban rather than -- can be privately owned under the Constitution, agricultural use, water companies will be able to why not water? buy water rights more cheaply from farmers as compared to the alternative of building new Q. If water rights are to be auctioned, will hydraulic infrastructure. not the rich buy them all, thereby depriving poor farmers of their livelihood? Q. Since the draft law fails to specify environmental regulations related to water quality A. In describing procedures for initial and to clearly define the institutions that would allocation, the draft law clearly specifies that establish and enforce environmental laws, would holders of acknowledged water rights will not this not lead to a deterioration in water quality? have to pay for them. Thus existing water users will receive their rights without charge. Once A. Peru presently faces many water-related water rights are obtained, farmers are free to sell environmental problems. The proposed law is them; however, there is no reason to expect that likely to have a slight positive effect on the they would sell their means of livelihood any environment since it provides incentives to more than they would sell their joint land and conserve water, thereby reducing the need to water rights presently. Auctions will be permitted build new infrastructure and helping reduce soil only where there is surplus water, and buyers will salinization. Furthermore, unlike the current law, be paying a market price for the rights. Under the draft law clearly states that for water rights to the prevailing system, whereby rights are given be established there must be enough water to for free, the system yields no public revenue and maintain a minimum ecological flow and the virtually ensures that the poor will be excluded accustomed quality of life (Article 8). It also and that water will not be used for the most explicitly recognizes the ability of the productive purposes. Environmental Authority to set water quality standards and for authorities to enforce the Q. When water is scarce, how will the State Environmental Code (Articles 61 and 62). be able to fulfill its obligation to guarantee water However, to reduce environmental problems for human life, health and welfare under a system substantially, there must be clear standards and whereby financially-strapped water companies the political will and institutional capacity to have to buy water rights rather than receive them enforce them. It is best that the mechanisms for without charge? the setting of such standards and for their enforcement be developed separately rather than A. Although the current system is supposed as part of the draft law, whose main goal is to ensure that there is water for domestic use, improving water use.  -24- has been unable to halt wasteful use. In fact, a market system, by raising the value of water, provides far greater incentives for rational water use than the present "use it or lose it" system. Similarly, the threat of confiscation of water rights for polluting one's water has been unable to halt the decline in water quality. The enforcement of water quality is a political decision; a system of fines (and possible jail sentences for repeat offenders) under a system of property rights for water may be more effective a deterrent than a threat of confiscation that is almost impossible to enforce. B. Resolving Issues in Establishing Tradable Water Rights 2.34 Because of water's peculiar characteristics, there are challenging issues to be addressed if a well-functioning market in water rights is to be established. These issues relate to: (i) defining rights for a commodity such as water which varies in volume and which is often difficult to measure; (ii) enforcing contracts for the sale of water; (iii) building the necessary infrastructure to change water intakes and transport water; (iv) minimizing damage to third parties; (v) protecting against environmental degradation; (vi) avoiding monopolistic pricing practices; and (vii) resolving disputes between consumptive and non-consumptive uses (i.e., hydropower). There are also important transitional issues relating to how one moves from the current system to the new regime of tradable water rights. Finally, a market will not lead to adequate investment in some activities which, because of their public goods nature or because they yield positive externalities, may have a high social return but may not be privately profitable (e.g., flood control, drainage, soil erosion and siltation reduction, domestic water supply, riparian protection). These will have to continue to be financed by the Government. 2.35 As will be seen below, the draft Peruvian water law should handle most of the above concerns adequately. Given the potential advantages of the new system over the current system, it is now important that the law, with the proposed modifications, be passed as soon as feasible. It is also important that the Government proceed expeditiously to draft the regulations to the law and that it devote sufficient resources for the effective and fair implementation of the law and its regulations. This section describes how the draft Peruvian water law proposes to handle the concerns mentioned above, makes suggestions to improve the draft law, and offers guidelines on how to implement the law. 2.36 The bulk of the above issues occur even under alternative ways of allocating water. Under Peru's prevailing administered system of water rights, the rights still have to be defined in a way that can be measured and the resulting allocation of water rights still needs to be enforced. The draft  -25- law (Article 6) allows rights to be defined in volumetric terms (in regulated systems), as percentages of stream flow, or in terms of hours that the gate to a canal may be left open (shifts or turnos).2' The conveyance infrastructure that will be required to effect transfers in line with priorities has to be built regardless of whether the priorities are determined by the market or by legal and administrative means. 2.37 Similarly, the same environmental laws and institutions that are needed to enforce environmental quality under an administered regime can operate under a tradable water rights regime. The conflicts between consumptive and non-consumptive rights exist even under an administered allocation system. The damage to third parties when water use is changed in line with priorities can affect the return flow and water availability to downstream farmers under either system. For the bulk of the issues then, the question becomes which of the two approaches -- tradable property rights or an administered regime -- is likely to yield better results. 2.38 There are several reasons to believe that a water market is likely to function better than an administered allocation. Because a market system increases the value of water, there are greater incentives for defining water rights clearly, for improving their measurement and enforcement, and for establishing mechanisms to resolve disputes. The conveyance infrastructure that must be built to effect the transfer is likely to built more cost-effectively by the private sector, which has greater incentive to control costs. In addition, the decision to invest in infrastructure is likely to be taken on more rational grounds under a water market regime. Similarly, water user organizations, which have to play an important role under either a administered allocation system or a water market regime will become stronger and better-organized when water rights are well-defined and made transferable. 2.39 The draft law thus handles the above concerns better than the prevailing system of water rights. However, the Government needs to make additional efforts to ensure both a smoother transition to the proposed new system and a well-functioning market for water rights. First, before laws and regulations can be issued, a better public information campaign needs to be carried out to make policy makers and water users aware of the problems with traditional approaches and obtain an understanding of the proposed operation and potential benefits of water markets. This is important to build support for the passage of legislation. An information campaign and 3/ In Chile, where most irrigation systems are not regulated, water rights are typically expressed as a share of the stream flow at the head of a river, with the shares summing to 100 percent.  -26- debate can help ensure that the final design and implementation of the legal framework is done in a transparent and participatory manner. Explaining the draft law officially and publicly, but with a willingness to modify the law, is critical to success. Farmers have to be made aware that their major concerns and objections have been considered and dealt with. The process can also help identify and mobilize groups that stand to benefit the most from the proposed legislation. 2.40 Second, there is a need to establish effective institutions to draft the regulations and to implement the law efficiently and fairly after approval. This requires ensuring that the capacity of the Water Council, the Watershed Directors, the Deputy Watershed Directors, the Watershed Committees and Water User Organizations is adequate to carry out their responsibilities and that sufficient budgetary resources are devoted for their effective functioning. It is encouraging that the Government has already begun efforts to contract for technical assistance to draft the regulations and to strengthen water user associations. It is equally important now to ensure that the members of the Water Council, Watershed Directors and Deputy Watershed Directors are capable, that they fully understand and support the new initiative, and that they are perceived to be honest and unbiased. Presently, only a few policy makers and public officials working in the water sector appear to understand the rationale and proposed operation of a water market. Given their key role in the initial allocation process and subsequent operation of the water market, poorly trained or corrupt Watershed Directors could prevent the market for water rights from ever developing or functioning effectively. There is also a need to establish a water rights registry and to finance technical assistance to WUOs to strengthen them. 2.41 There is a feature in the draft law which could compromise the efficient operation of a water market. The draft law specifies that the Board members of the Water Council and the Watershed Director must have, respectively, at least 15 and 10 years of experience in water activities. This provision is likely to result in current Technical Administrators (Administradores Tecnicos) being appointed to run the new water rights regime. Since they are generally closely associated with the current water rights system and since they lack the confidence of most water users, this could lead to an improper initial allocation of water rights and increase water conflicts. Therefore the criteria for publicly appointed personnel should include only minimal reasonable qualifications for the job and the criteria should be in the regulations to the law rather than the law itself so that they may be changed if they are later found to be inappropriate. 2.42 Third, the Government needs to fully understand and address the implications of various ways of dealing with the issue of third-party  -27- effects. These can occur when the sale or utilization of one person's water rights infringes the rights of another user. This is most likely to occur for agricultural "return flows," which is the water left over after irrigation that was not consumed by crops, but which may have percolated down to an aquifer that later rejoined the river. If a farmer were free to sell his entire irrigation water to users outside the area, users downstream that may have received the farmer's return flows would lose their water without compensation (Figure 1). The return flow problem may exist even when water is non-tradable. For example, in areas with significant return flows, farmers that switched to drip irrigation from traditional irrigation technology in order to expand irrigated area would reduce the water available to downstream farmers. The return flow problem exists whenever the water available in the lower reaches of the river is substantially greater than in the upper reaches (adjusting for any water from tributaries). 2.43 To address the return flow problem, the draft law proposes that the Deputy Watershed Director approve requests for changes of water intake only when third parties are not affected (Articles 27 and 28)