w-rri WORLD BANK TECHNICAL PAPER NUMBER 315 -Deco a I q 4 5 Water Allocation and Water Markets An Analysis of Gains-from-Trade in Chile Robert R. Hearne and K. William Easter FtEE af f lDm~~~~~~ LI Y FOREST P!-__ NlN- ~i IJfl D TENURE-LN ST RECENT WORLD BANK TECHNICAL PAPERS No. 237 Webster, The Emergence of Private Sector Manufacturing in Poland: A Survey of Firms No. 238 Heath, Land Rights in Cote d'lvoire: Survey and Prospectsfor Project Intervention No. 239 Kirmani and Rangeley, Internzational Inland Waters: Conceptsfor a More Active World Bank Role No. 240 Ahmed, Renewable Energy Technologies: A Review of the Status and Costs of Selected Technologies No. 241 Webster, Newly Privatized Russian Enterprises No. 242 Barnes, Openshaw, Smith, and van der Plas, What Makes People Cook with Improved Biomass Stoves? A Comparative International Review of Stove Programs No. 243 Menke and Fazzari, Improving Electric Power Utility Efficiency: Issues and Recommendations No. 244 Liebenthal, Mathur, and Wade, Solar Energy: Lessonsfrom the Pacific Island Experience No. 245 Klein, External Debt Management: An Introduction No. 246 Plusquellec, Burt, and Wolter, Modern Water Control in Irrigation: Concepts, Issues, and Applications No. 247 Ameur, Agricultural Extension: A Step beyond the Next Step No. 248 Malhotra, Koenig, and Sinsukprasert, A Survey of Asia's Energy Prices No. 249 Le Moigne, Easter, Ochs, and Giltner, Water Policy and Water Markets: Selected Papers and Proceedingsfrom the World Bank's Annual Irrigation and Drainage Seminar, Annapolis, Maryland, December 8-10, 1992 No. 250 Rangeley, Thiam, Andersen, and Lyle, International River Basin Organizations in Sub-Saharan Africa No. 251 Sharma, Rietbergen, Heimo, and Patel, A Strategyfor the Forest Sector in Sub-Saharan Africa No. 252 The World Bank/FAO/UNIDO/Industry Fertilizer Working Group, World and Regional Supply and Demand Balancesfor Nitrogen, Phosphate, and Potash, 1992/93-1998/99 No. 253 Jensen and Malter, Protected Agriculture: A Global Review No. 254 Frischtak, Governance Capacity and Economic Reform in Developing Countries No. 255 Mohan, editor, Bibliography of Publications: Technical Department, Africa Region, Iuly 1987 to April 1994 No. 256 Campbell, Design and Operation of Smallholder Irrigation in South Asia No. 258 De Geyndt, Managing the Quality of Health Care in Developing Countries No. 259 Chaudry, Reid, and Malik, editors, Civil Service Reform in Latin America and the Caribbean: Proceedings of a Conference No. 260 Humphrey, Payment Systems: Principles, Practice, and Improvements No. 261 Lynch, Provision for Children with Special Educational Needs in the Asia Region No. 262 Lee and Bobadilla, Health Statisticsfor the Americas No. 263 Le Moigne, Subramanian, Xie, and Giltner, editors, A Guide to the Formulation of Water Resources Strategy No. 264 Miller and Jones, Organic and Compost-Based Growing Media for Tree Seedling Nurseries No. 265 Viswanath, Building Partnershipsfor Poverty Reduction: The Participatory Project Planning Approach of the Women's Enterprise Management Training Outreach Program (WEMTOP) No. 266 Hill and Bender, Developing the Regulatory Environmentfor Competitive Agricultural Markets No. 267 Valdes and Schaeffer, Surveillance of Agricultural Prices and Trade: A Handbookfor the Dominican Republic No. 268 Vald6s and Schaeffer, Surveillance of Agricultural Prices and Trade: A Handbookfor Colombia No. 269 Scheierling, Overcoming Agricultural Pollution of Water: The Challenge of Integrating Agricultural and Environmental Policies in the European Union No. 270 Baneijee, Rehabilitation of Degraded Forests in Asia No. 271 Ahmed, Technological Development and Pollution Abatement: A Study of How Enterprises Are Finding Alternatives to Chlorofluorocarbons No. 272 Greaney and Kellaghan, Equity Issues in Public Examinations in Developing Countries No. 273 Grimshaw and Helfer, editors, Vetiver Grassfor Soil and Water Conservation, Land Rehabilitation, and Embankment Stabilization: A Collection of Papers and Newsletters Compiled by the Vetiver Network No. 274 Govindaraj, Murray, and Chellaraj, Health Expenditures in Latin America (List continues on the inside back cover) WORLD BANK TECHNICAL PAPER NUMBER 315 Water Allocation and Water Markets An Analysis of Gains-from-Trade in Chile Robert R. Hearne K. William Easter The World Bank Washington, D.C. Copyright 1995 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing December 1995 Technical Papers are published to communicate the results of the Bank's work to the development com- munity with the least possible delay. The typescript of this paper therefore has not been prepared in accor- dance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibili- ty for errors. Some sources cited in this paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatso- ever for any consequence of their use. The boundaries, colors, denominations, and other information shown on any map in this volume do not imply on the part of the World Bank Group any judgment on the legal status of any territory or the endorsement or acceptance of such boundaries. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy por- tions for classroom use is granted through the Copyright Clearance Center, Inc., Suite 910, 222 Rosewood Drive, Danvers, Massachusetts 01923, U.S.A. The complete backlist of publications from the World Bank is shown in the annual Index of Publications, which contains an alphabetical title list (with full ordering information) and indexes of sub- jects, authors, and countries and regions. The latest edition is available free of charge from the Distribution Unit, Office of the Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, The World Bank, 66, avenue d'Iena, 75116 Paris, France. ISSN: 0253-7494 Robert Heame is a Research Associate with the International Institute for Environment and Development, London. K. William Easter is a Professor in the Department of Applied Economics, University of Minnesota. Both authors are former consultants to the Agricultural Policies Division of the World Bank. TABLE OF CONTENTS FOREWORD.v ACKNOWLEDGMENTS ...........................vii EXECUTIVE SUMMARY ......................................... ix I. INTRODUCTION ........................................... 1 II. LITERATURE REVIEW ........................................ 4 III. WATER ALLOCATION INSTITUTIONS IN CHILE ...................... 6 IV. LOCAL WATER MARKETS AND GAINS FROM TRADE . 12 A MODEL ....................................12 CALCULATION OF GAINS-FROM-TRADE AND ECONOMIC RENTS ......14 RESULTS ............................................ 27 V. OBSERVATIONS AND CONCLUSIONS ............................. 38 ANNEX I: CASE STUDIES ........................................ 43 THE FIRST SECTION OF THE MAIPO RIVER ...................... 44 THE ELQUI VALLEY . ...................................... 47 THE LIMARI VALLEY . ..................................... 50 THE AZAPA VALLEY . ..................................... 53 ANNEX II: REPRESENTATIVE CROP BUDGET ................57 ANNEX III: SURVEY INSTRUMENT . ................................. 65 REFERENCES ......................................73 MAPS SECTION ..............................................77 .i.i FOREWORD With the publication of its 1993 policy paper Water Resources Management, the World Bank made a commitment to assist developing countries in establishing institutional frameworks and management procedures that would enable countries to utilize their water resources in an economically and environmentally sustainable manner. The impetus for the policy came from the alarming deterioration and increasing scarcity of freshwater resources around the globe, caused mainly by growing population pressures and the failure to properly consider the economic value of water. Usually when water is given little or no economic value, it is misallocated and misused. This paper is intended to reinforce the World Bank's overall effort to improve the management of natural resources and to highlight the importance of water resources in particular. The Bank's 1993 Water Policy Paper and the 1994 technical paper, A Guide to the For,nulation of a Water Resources Strategy, were the first steps in this process. The paper focuses on water al0ccat-on problems and the performance of water markets in improving allocation. By examining specific case studies in Chile, one of the few countries that has encouraged markets for water, it demonstrates that water can no longer be treated as a free good and better ways must be found to improve its allocation and use. Efficient water markets are one means of improving water allocation, while at the same time providing a mechanism to directly compensate existing water users. In addition, as the market value of water becomes clear to water users, they will use it more efficiently. The findings in the paper suggest that market transfers of water use rights in the study area produce economic gains both in intersectoral trades and in trades among farmers, and that they produce rents for both buyers and sellers. The extent of trade and the level of gains vary depending on river basin locations, the alternative value of water in present use, water delivery infrastructure and the cost of the transactions. Our hope is that this paper will encourage professionals engaged in water resources management to adopt practices that produce the desired outputs, but do not have unwanted impacts on the environment. It is important that we give users incentives to make better use of water resources. Without such incentives, the misallocation of water resources will continue and future generations will find their opportunities for water use severely restricted. "Alexander F. McCalIa, Director Gobind T. Nankani, Director Agriculture and Natural Resources Department Country Department I Latin America and Caribbean Regional Office v ACKNOWLEDGMENTS Much of this information was obtained during a six month mission to Chile during 1993 and 1994 with funding provided by the World Bank's Agricultural Policies Division. The authors extend thanks for the collaboration of colleagues at The World Bank, University of Minnesota, and the Pontificia Universidad Cat6lica de Chile, especially Gershon Feder, Ariel Dinar, and Guillermo Donoso. A special thanks to all those in the Chilean government and private sector that provided information and insight, especially Fernando Peralta, Samuel del Canto, Gustavo Manriquez, Humberto Pehia, Luis Jorquera, Pablo Anguita, Juan Ignacio Domiguez, and Eduardo Venezian. Also gratitude is extended for the dedication and camaraderie of data collectors Pablo Venezian, Carlos Bonilla, Jose Cancino, Carlos Anis, and Jeronimo Concha. Helpful comments were provided by Terence R. Lee, United Nations; Ronald C. Griffin, Texas A&M University, John Briscoe, World Bank; Marie L. Livingston, University of Northern Colorado; Karin Kemper, World Bank; Mateen Thobani, World Bank; Pablo Anguita Salas, Direcci6n de Riego, Chile. Document typesetting was done by Maggie Wu, Cindy Wong, and Patricia Noel. The task manager for this study was Ariel Dinar. vii EXECUTIVE SUMMARY With the growing concern about the increased scarcity and inefficient allocation and use of water resources, attention has been focused on the use of markets to allocate water. Market- based allocation could secure water supplies for high-value uses in urban and rural areas without the need to develop costly new sources of supply that may be environmentally damaging. Also, by requiring compensation for water transferred away from low valued uses, water markets provide an incentive for more efficient water use in agriculture, industry, and municipalities. Furthermore, if markets work efficiently, price signals can provide the information needed to allocate water more effectively than models developed by a central water resources management agency. Chile is one of the few countries that has encouraged the use of markets in water resource management. The market allocation of water in Chile is possible, in part, because a system of transferable water-use rights was reestablished in 1981. These rights are independent of land use and land ownership. Thus, trades of water rights are not tied to land sales. When combined with flexible irrigation infrastructure such as adjustable gates and effective water user associations (WUAs), these rights can stimulate a relatively active water market in areas of water scarcity. Given the hope that the market allocation of water-use rights can offer a possible solution to the problems of increased scarcity and inefficient allocation of water resources, this study of water markets in Chile was initiated in late 1993. In order to assess the impact of water markets and transactions costs in Chile, four river valleys, the upper Maipo, the Elqui, the Limari, and the Azapa were selected as case studies. The sale of water-use rights in the Elqui and Limarf valleys, during the years 1986 to 1993, was analyzed to determine the gains-from-trade from market transfers. In the upper Maipo valley, transactions were rare and were not included in the analysis. Similarly, in the Azapa valley, only a few transactions were identified, and gains-from- trade were not calculated. In the Elqui valley, transactions were infrequent and constrained by the lack of adjustable gates, but there were significant intersectoral transfers as well as a slow transfer of water-use rights within agriculture. In the Limari valley, with its well-developed system of irrigation infrastructure and well organized WUAs, transactions were relatively frequent. In the analysis of water markets, crop budgets were used to estimate the value of water in agricultural production. The value of water-use rights to urban water supply companies was estimated using the avoided cost of the next best alternative investment. The analysis demonstrated that the market transfer of permanent water-use rights produced substantial economic gains-from-trade in both the Elqui and Limari valleys. These economic gains were present in intersectoral trades and in trades between farmers, and they produced gains for both buyers and sellers. Buyers, especially farmers growing profitable crops who bought water-use rights and individuals buying rights for potable water supply, received higher rents than sellers. Large table grape producers in the Limari valley and individuals buying water for human consumption in the Elqui valley received the highest rents. In the Elqui valley, total and net gains-from-trade per share (average of 0.5 liters/second) were within the range of recent transfer prices of US $1,000. In the Limari valley, gains-from-trade of water-use rights were three times the recent ix transaction prices of US$3,000 for a share (4,250 m3 annually) of water from the Cogoti Reservoir. In the Elqui valley, where intersectoral trading occurred, most of the water-use rights transferred out of agriculture were not used by their owners prior to the sale. This means that there are considerable financial gains from these transactions, while the estimated economic gains to society are relatively modest as someone else is assumed to use the water. It is important to emphasize that in this valley, the intersectoral transfer of water involved sales by individuals who were not actively using the water in farming, rather than sales by active farmers selling marginal quantities of water. These case studies demonstrate the diversity of water allocation systems and water management practices in northern and central Chile. In areas where trading was active, especially in the Limari valley, transactions costs have not presented an appreciable barrier to trading. Nonetheless, in the large canal systems with fixed flow dividers, such as those found in the Elqui and Maipo valleys, there have been very few transactions. Various factors facilitate trading, but the absence of trading in these large canal systems highlights the costs of modifying fixed infrastructure, especially for trades between farmers. Water user associations play an important role in facilitating the market reallocation of water, especially in the Limari valley where trading is active and in the Elqui valley where intersectoral trading occurs. In the Limari valley, where reservoir storage, adjustable gates with flow meters, and well organized WUAs helped lower transactions costs, the water market is active and gains-from-trade are substantial. This study has several important implications for other countries faced with water scarcity: * First, there are significant gains-from-trade that can be realized by fostering water markets. These gains occur in both intersectoral trades and trades between farmers. * Second, transferable water-use rights are essential for efficient water markets. These rights can be stipulated by volume or by percentage of river or canal flow. But in areas where water supplies are highly variable, it is necessary to agree on how the rights will be altered during times of scarcity. * Third, great care should be exercised in the initial allocation of water-use rights among users in order to make sure that all the rights are not captured by a few individuals. If the water is to be used for irrigation and an equitable distribution of land and water already exists, a good strategy is to distribute water-use rights to the owners of land on which the water is being used. * Fourth, technology such as adjustable gates and institutional arrangements that encourage the formation of active water user's associations can substantially reduce transactions costs and facilitate market trading. x * Fifth, the presence of privately held water rights does not necessarily make it more difficult to reach environmental quality objectives for rivers. Water quality regulations need to be established and enforced irrespective of the water allocation system. In Chile, where river valleys are relatively short, the quantity and quality of return flows may be less problematic than in other countries. * Finally, within a decentralized system of water resource management, there is a continuing role for water management authorities in enforcing rights and resolving conflicts. Yet if institutional arrangements are established that allow water users to resolve conflicts among themselves, they can avoid the need for further government intervention. Further considerations should be given to: * Land use patterns * Urban water use patterns * Irrigated vs. dry land agriculture * Water conveyance technologies as constraint on water market transactions * Irrigation technologies and their role in a water market setup - Supporting legal and institutional mechanisms to regulate water market activities * Role of incentive systems for water savings to enhance water market transactions * Environmental consequences - negative externalities from water transfer (water quality, soil erosion) - Third party effects - regional employment and welfare xi I. INTRODUCTION With the growing concern about the increased scarcity and inefficient allocation and use of water resources, much attention has been focused on the use of markets in water allocation. This market based allocation could secure water supplies for high-value uses in urban and rural areas without the need to develop costly, new sources of supply that may be environmentally damaging. Also by securing compensation for water transferred away from low valued uses, water markets provide an incentive for more efficient water use in agricultural, industrial, and municipal uses. Furthermore, if markets work properly, price signals can provide the information needed for efficient water allocation more effectively than models developed by a central water resources management agency. The effectiveness of water markets is constrained by the ability of buyers and sellers to measure and transport water, to legalize and enforce transactions, to account for water quality, and to protect the rights of other water users. Thus, the effect of transaction costs and the infrastructure and institutions that reduce these transaction costs are critical to the effectiveness of water markets. In addition, the unconstrained movement of water via private exchanges can produce negative external effects on third party users. There is also the fear that the free exchange of water may disadvantage poor people. Chile is one of the few countries that has encouraged the use of markets in water resource management. Market allocation in Chile is possible, in part, because a system of transferable water-use rights was reestablished in 1981. These rights are independent of land use and land ownership, thus trade of water rights is fairly unrestricted. The codification of these water-use rights coincided with a series of reforms in the Chilean economy including privatized land rights and liberalized trade. Although Chile is unique in its water rights system, the challenges that face its system of water resource management are very similar to those that face other countries. The increased population and income in Chile's urban areas is creating an increase in the urban demand for water. In addition, industrial and residential pollution are overburdening the assimilative capacity of many of Chile's rivers. This paper presents a description and an analysis of water allocation and water markets in Chile. The first part of the paper provides a brief introduction and theoretical framework. The second part of the paper reviews relevant literature. The third section offers background information on water allocation institutions in Chile. The fourth part of the paper presents a brief description of local water markets as well as an analysis of gains-from-trade. The last section provides conclusions as well as general observations on water policy in Chile. Annexed to this paper is a more thorough description of four river basins used as case studies in this analysis. A second annex presents maps of the valleys discussed. A third annex presents crop budgets used in valuing water for the gains-from-trade analysis. A final annex displays the questionnaire used in the survey of farmers who have participated in market exchanges. 1 Markets for Water and Water-Use Rights As an introduction to a discussion of water markets it is useful to distinguish between the exchange of water and the exchange of water-use rights. This is appropriate because the distinction highlights the importance of the institutional environment for water resource allocation. The former - sometimes referred to as a "spot market" - occurs when the owner of a legal or prescriptive right to a certain volume or flow of water sells a portion of that water, sometimes outside of legal sanction, to a neighbor in a simple transaction. These exchanges are for a finite period of time - sometimes only a few hours of irrigation. Although the unit of sales may not be metered volumetrically, both buyer and seller have good information on the volume involved. A more permanent transaction involves the exchange of the water-use right itself. This generally requires legal sanction to maintain the security of the right after the transfer. These transfers are generally permanent, but can be for a finite, but extended period of time - at least one irrigation season. And the burden of uncertain supply will fall on the purchaser of the right. Water Markets, Transaction Costs, and Institutions Where water is scarce and legally defined transferable property rights exist, market trades can be expected when the difference in the value of water between two uses is greater than the costs of transferring the water. In the absence of transaction costs economic incentives would induce water users to trade water-use rights until the marginal value of these rights was equal across all users. Of course, transaction costs do limit the movement of water and the transfer of water-use rights. Transaction costs for water market transfers include: i) the cost of the physical infrastructure needed to measure and transport water, including the evaporation and filtration losses during conveyance; ii) the cost of searching and finding willing buyers and sellers, and negotiating a contract; and iii) the cost of validating legal ownership of the water-use right, legalizing the contract, enforcing contract provisions, and acquiring necessary permission from regulatory authorities to transfer water. Because these transaction costs can be large, the number of potential buyers and sellers may be limited - which may result in non-competitive pricing. In order to reduce the burden of these transaction costs, public organizations can be established to construct water delivery infrastructure, to modify and monitor the distribution of river and canal water, to expedite the dissemination of market information, to maintain public records of water-use rights, and to protect the rights of third parties affected by a transfer of water. These services may be provided by a central government, local governments, or community organizations. Many governments have made large investments in irrigation infrastructure and water management authorities. Although these irrigation systems were not generally designed to facilitate market transfers, the presence of flexible infrastructure should reduce the transactions costs of market exchanges. Also, user groups, especially the water user associations (WUAs) that manage and maintain canals, can play important roles in facilitating and monitoring trades. Because water use is characterized by a high degree of interdependence, individuals may want to restrict the amount and types of transfers that occur. Changes in the allocation of upstream water and irrigation practices can impose a negative externality on downstream users. The transfer of water away from a canal can increase the percentage of water lost in conduction 2 and evaporation in that canal. An increase in the number of water-use rights flowing through a canal can reduce the amount of water received per water-use right in the canal during times when river levels are high. This is because during high water unlimited withdrawals from the river are permitted, and the only constraint on water delivery is the carrying capacity of the canal. Also, changes in water use can significantly effect water quality. Thus both government authorities and WUAs may want to regulate water transactions to ensure that they are beneficial to the community of water users and society. 3 II. LITERATURE REVIEW The market exchange of water and water-use rights is a relatively rare phenomenon, and thus the economic analyses of actual exchanges are scarce. Still, there exists a wide range of literature which addresses issues pertinent to water markets. Unfortunately, much of the relevant literature discusses the water markets of the western U.S.A., but more recently studies have been completed for other areas. There is also a lack of economic research on proper regulatory strategies, methods of reducing transaction costs, and the effectiveness of water markets outside the U.S.A.. Included in the literature is a large number of articles which take an interdisciplinary approach to study the institutions that provide the environment for market allocation. In general these articles review how the prior appropriation doctrine in the western U.S.A. has been modified to provide a well regulated framework for market transfers (Clyde, 1989; Anderson and Leal, 1989; Griffin and Boadu, 1992; Harper and Griffin, 1988; and Schupe et al., 1989). In a 1990 volume pertinent to Chile, Lee traces the development of water distribution systems in Latin America and suggests that water management has been dominated by single purpose government agencies concentrating on the development of large supply projects. He then features four South American case studies, including Chile's Limari River valley, to describe the organizational difficulties of large multipurpose water systems. In comparison, the case study from Chile is fairly favorable, although the discussion focuses more on the physical characteristics of the system and gives little attention to the institutional concerns in Chile or to market trading. Along with the literature that explains water institutions, there exists some, mostly non- empirical, economic discussions of water markets and their policy implications. This literature provides a good background of the issues involved with water markets in the western U.S.A. These issues include: transaction costs, hydrological uncertainty, the non-pecuniary value of water, and the "community" value of water (Brajer and Martin, 1989 and 1990); the characteristics of water markets that achieve social welfare (Howe et al., 1986); water quality (Colby Saliba and Bush, 1987; Colby Saliba, 1987); the shadow value of water (Easter and Tsur, 1995); and the rent accruing to water "owners" (Bowen et al, 1991). A few papers have generalized beyond the experience in the western U.S.A.. Brajer et al. (1989) describes market allocation for the benefit of non-economists, featuring a good discussion of transaction costs and how market imperfection can lead to economic rents. Rosegrant and Binswanger (1992) examine alternative policies to improve water use and environmental management in irrigation and suggest that a market for water-use rights would function well, once water-use rights are established. The authors present a institutional innovation approach to the development of markets - as water prices rise, the institutions requisite for water reallocation will follow. There are a few empirical studies concerned with the effects of water markets. These studies mainly focus on the benefits of market and administrative transfers of water. In a large study of interregional water transfers within California, Vaux and Howitt (1984) estimate potential annual gains-from-trade for 1980 ($67 million), 1995 ($156 million), and 2020 ($219 4 million). Chang and Griffin review the water institutions of the lower Rio Grande valley (U.S. side) and estimated the gains-from-trade (ranging from $3,000 to $16,000 per 1000 m3 ) for transfers from agriculture to municipal water supply. Dinar and Letey (1991) use a micro-level production model for the central valley of California, and suggest that water markets increase farmers' profits, reduce farmers' use of water - thus reducing the salinity and selenium buildup, and increase farmers' investment in water conserving technologies. Whittelesey and Houston (1984) simulate diversions of water from irrigation to hydrogeneration in the Pacific Northwest and show that the value of water, welfare, and farm income all increase. Hamilton et al. discuss the welfare gains and policy implications of a transfer of provisional water rights from agriculture to a hydroelectric utility in the Pacific Northwest. Maass and Anderson (1978) evaluate the water market in Alicante, Spain, and found that the market system produced greater net increases in regional income than the rotation systems used in neighboring communities. Another set of articles address water markets and focus on specific policy implications of water trading. Charney and Woodard (1990), and Howe et al. (1990) estimate the effects of rural-to-urban water transfers on the agricultural areas losing water and show that the indirect upstream and downstream effects on rural commercial activity can be significant in highly localized areas. In a simulation of potential California water trades Weinberg et al. (1993) show that although the effect of increased water prices on salinity and selenium accumulation is noteworthy, water markets may not serve as well as a set of well formulated Pigouvian' taxes in reducing negative environmental externalities. Colby (1990) investigates the transaction costs required to obtain approval of water transfers and the litigation costs of third party challenges to transfers. She suggests that these institutional constraints can be used as Pigouvian taxes to protect against the negative externalities of water transfers. Colby's estimates of transactions costs average 6% of purchaser's costs for transfers in 4 western states and this is considered not to be a burden. In a fairly detailed analysis, Rosen and Sexton (1993) revisit the transfer of water from the Imperial Irrigation District to the Southern California Metropolitan Water District, and conclude with suggestions of policy reforms that decentralize control over water-use rights. There is also a limited literature on current water allocation in Chile, although without economic analysis. Two volumes of reports prepared for the National Irrigators Conventions of 1986 and 1989 contain a lot of general information on contemporary water issues as well as many reports on water-use from various river valleys. Gazmuri (1994) provides a good, and very optimistic, review of the primary features of the 1981 Water Code. In a institutional review of the roles of property rights, markets, and the government in water-use in Chile, Bauer suggests that the 1981 Water Code has worked well in the agricultural sector but not so well in intersectoral water allocation. He asserts that private property rights to water have served agriculture well, despite the fact that poor infrastructure, incomplete archives, and a cultural resistance to water sales has limited use of market mechanisms to transfer water. Nonetheless, Bauer (1994) argues that the Water Code has serious flaws in its approach to conflict management, non-consumptive water-use rights, and water quality. Finally, Donoso (1994) reviews the negative economic effects of incentives that could be generated by the proposed use- it-or-loose-it rule on water-use rights. ' A.C. Pigou argued that externalities produce a difference between social and private returns, and a system of taxes and subsidies could be used to internalize these externalities. 5 III. WATER ALLOCATION INSTITUTIONS IN CHILE In order for market mechanisms to efficiently allocate water between competing uses, institutions must be in place that enable both buyer and seller to obtain fair value in a trade. The most critical prerequisite for markets is a system of transferable water-use rights. Ideally, these rights should specify the quantity, quality, and timing of water delivery. Water-use rights should also specify the criteria by which water is rationed in times of drought. Also, a system of regulations or taxes may be required to protect third parties from damage and to enforce the privileges and restrictions placed on these rights. Finally, a mechanism of resolving conflicts between water users is necessary. Water-Use Rights Chile has a tradition of private development of water resources and private rights to shares of river and canal flows that dates to the colonial era. This tradition is maintained in the National Water Code of 19812 which allows private transferable property rights for water use. This water law reversed the 1969 water law, written during a period of land reform, which tied irrigation water to the land and mandated state control over water resources. The 1981 water law stipulates that water is a national resource for public use but that permanent and transferable rights to utilize water can be granted to individuals in accordance with the law. Water-use rights can be granted by the government upon petition, can be purchased from an individual owner, or can be retained based on traditional use. Currently, there is no stipulation that water-use rights must be utilized in order to be retained. Rights can be defined as permanent or contingent. Permanent rights are granted for use on unexhausted sources of supply. In most of Chile's river basins, especially in the north and in the central valley, all permanent use rights have already been assigned. Contingent rights are granted for surplus water, that is water flows that exceed those demanded by permanent rights holders during times of high water. Reservoir or lake water is not subject to contingent rights since, under most climatic conditions, the regulation of water flow is sufficient to nullify the chance of excess flows of water. Rights are also designated as consumptive and non-consumptive. Consumptive rights entitle the user to completely consume the water without any obligation to return it. Non- consumptive rights grant the owner the use of the water as long as it is returned to its source at a specified quality, and does not interfere with consumptive use rights. The law stipulates that rights are to be specified by volume of flow per unit of time. But in reality rights are defined as a share of stream flow, because the high variability of natural river flows prohibits volumetric specification. In order to resolve this inconsistency, natural rivers are divided into sections, and each canal, intake, and withdrawal point receives a percentage of the water in that section of the river. Volumetric equivalencies of river shares are stipulated for the flow of the river that occurs in 85 out of 100 years. When river flows are insufficient to meet volumetric specifications, 2 D.F.L. Number 1122, published in the Official Journal, Santiago, October 29, 1981. 6 water flows into the different intakes are reduced proportionally. However, rights on some rivers have been over allocated so that water flows will be sufficient to meet volumetric specifications in far less than 85 % of years. Since consumptive use rights are granted for the full use of all the water stipulated in the right, downstream users do not have any right to return flows generated from upstream users. Of course, this has little effect on the first section of a river since return flows mostly augment downstream sections. Water users in downstream sections of a river divide water that enters through springs, rainfall, and return flows. These rights holders are not protected by law from any change in upstream water use that significantly reduces return flows. There is also no restriction on the transfer of upstream water to another basin. Water-use rights are required for groundwater exploitation. Individuals can request from the Direcci6n General de Aguas (General Directorate of Water) a right to groundwater, once they have confirmed the existence of a certain yield, at a certain depth. The groundwater-use right is accompanied by a prohibition on other groundwater withdrawals in the protective area specified in the right. Any party with legally entitled rights to water that may be adversely affected by the granting of new groundwater-use rights, can oppose the grant, by informing the regional Direcci6n General de Aguas office within 30 days of the publication of the entitlement in the Official Journal'. If a petition is opposed, the Regional Director of the Direcci6n General de Aguas can either grant or permit the new water-use right. There is no property tax on water-use rights. But land is taxed according to its productive value, which accounts for irrigation. There are seven different categories of agricultural land for tax purposes. These range from high quality, irrigated, central valley land to non-irrigated land. Thus the concept of separating land and water has not reached the tax code. There is no sales tax on the transfer of water-use rights, but there are fees paid to lawyers, notaries, and the Real Estate Registry, conservador de bienes raices (CBR). Water User Associations and Irrigation Development Historically, the development of irrigation in Chile has been dominated by the private sector. Over one million ha. have been developed for irrigation with private investment (Gazmuri, 1994). These were mostly small run-of-the-river systems. Starting around 1930, the government began developing major irrigation infrastructure. Many of these investments were never completed, and since 1945 only one major irrigation system (the Paloma Reservoir system in the Limari Valley) was built by the government. Also, long-term contracts to recover costs from users were denominated in local currency which has since lost value drastically. In the 1970s both private and public investment in irrigation was absent due to the uncertainty of agrarian reform and government austerity. With more secure land and water rights, and liberalized agricultural policy, private investment in irrigation for high-valued fruit and vegetable crops rapidly expanded in the 1980s. 3 Owners of rights to surface water and water user associations have opposed groundwater exploitation near rivers and canals. 7 All privately developed irrigation systems and many of those developed by the state are owned and controlled by independent water user associations (WUAs). These WUAs are owned and operated by their members, and charge fees based on the their capital and operating costs. The WUAs maintain the canal systems, keep records of rights holders, apportion water to individual rights holders according to their recorded shares, and enforce water rights. The 1981 Water Code specifies rules for the formation, governance, and obligations of these WUAs. There are three different types of WUAs that are recognized in Chile. A water community, communidad de agua, consists of any water users that share a common source of water. They can be chartered and recognized, with formal procedures, but many communidades de agua are not. Irrigation Associations, asociaciones de canalistas, serve irrigators that share a common canal, and have a legal status which allows them to enter into contracts and receive financing. Juntas de vigilancia (JDVs), made up of all users and user associations on a common stream or section of a river, are responsible for administering water use in the river. JDVs control the canal intakes that flow from the river. Some JDVs administer dams for storage of irrigation water. At the national level the Confederation of Canal Operators (CCC) is legally recognized as the representative of most WUAs. According to the CCC, about one half of all WUAs are legally registered with the Direcci6n General de Aguas. The Ministry of Public Works (MOP), which includes the Direcci6n General de Aguas, has played an important in water management in Chile. Its Direcci6n de Riego (Directorate of Irrigation) is responsible for planning, supervising construction, and operation of public sector irrigation infrastructure. The National Irrigation Commission (CNR) is an interministerial committee chaired by the Minister of Economy with the membership of the Ministries of Finance, Public Works, Agriculture, and Planning. The CNR is the major government entity which determines irrigation policy. Except for recent initiatives in the construction of several large schemes, there has been no public investment in large irrigation projects for the last 15 years (Gazmuri, 1993). Water Supply and Sanitation Chile has traditionally had a high level of water and sewerage coverage: 98% of urban and 75% of rural households have had household access to piped water, and 80% of urban households are connected to central sewerage systems. However, in the past fifteen years the water and sanitation services have undergone a major transformation. In 1990, the regulatory functions of the former national water supply and sanitation service, SENDOS, were transferred to the newly created Superintendency of Sanitary Services (SSS). In addition, SENDOS was decentralized into 11 separate, autonomous, regional water supply and sanitation (WSS) companies, along the lines of the Metropolitan WSS Company of Santiago (EMOS) and the WSS Company of Valparaiso (ESVAL). Stock is currently held by the government and CORFO, a publicly owned corporation. These independent water companies are obligated by law to provide water and sanitation services to the large municipal areas. They are required to deliver full water supply to their concessions 95% of the time. Water rates are based on delivery costs, with a fair return on capital, and reviewed every five years by SSS. A premium is charged in the summer months in order to manage demand during periods of high use. These water companies have inherited the 8 water-use rights that were traditionally held by the municipalities that they serve and later held by SENDOS. Some of these rights are considered to be priority rights, and consequently, such volumetric withdrawals are not reduced proportionally during times of water scarcity as is the case of other shares. There is a government program, operated through the municipalities, to subsidize potable water supply and sanitation service charges to low-income households. Environment The Chilean constitution guarantees citizens the right to live in a pollution free environment. Water quality standards for both agricultural use and potable water have been adopted. But generalized water quality standards that limit the effluent that can be placed in a river or stream have not yet been adopted. The 1993 proposed amendments to the Water Code would have empowered the Direcci6n General de Aguas to set these standards. Currently the Direcci6n General de Aguas's Environmental Department and the Superintendency of Sanitary Services (SSS) are coordinating to inventory all the polluters in the country4. Information on the actual amounts of each contaminant is collected in areas where SSS is monitoring drinking water, but this information is not part of the inventory. Based on this inventory, SSS is developing basic standards to ensure the enforcement of a 1916 law concerning industrial effluents5. A major source of water quality problems is the natural sediment flowing from the highly eroded Andes mountain range. This is a young mountain range which has been deforested for centuries, and reforestation efforts are only marginal. Both the mining and industrial sectors are known to produce harmful effluents. Most large mining operations have made efforts to control pollution in order not to disturb the environmental sensibilities of stockholders and the public at large. In southern Chile, a large percentage of water quality problems result from large pulp mills. In the Santiago area, there is one particular water course - used for irrigation - that has been until recently, an open sewer. Outbreaks of cholera and typhoid have led to a controversy over the reuse of municipal wastewater in the irrigation of food crops. Historically, regulations made by public health authorities on the type of cultivation allowable with wastewater irrigation have been enforced inconsistently. However, new vigilance, including efforts to limit the sale of certain horticultural products - as well as the increased demand for produce that is irrigated with safe water - have decreased the risk of food born diseases considerably. Formal institutions responsible for environmental management and protection are new to Chile. Two commnissions, the Environmental Legislation Committee (COLMA) and the National Ecological Commission (CONADE), were formed in the early 1980s but remained relatively inactive. In 1990, The National Environmental Commission (CONAMA) was formed. The central role of CONAMA in formulating an environmental agenda for the government was reaffirmed by President Aylwin in May 1992. In early 1994, the government passed legislation to require environmental impact studies for new construction and development projects. 4 Interview with Carlos Salazar, DGA, June 1993. 5 Memo from Terence Lee, CEPAL, March 1995. 9 Regulation and Conflict Resolution The Direcci6n General de Aguas, which is part of the Ministry of Public Works (MOP) is responsible for water resources planning and the development and exploitation of water resources. It collects data on water resources, maintains cadastres of water use, and has limnited authority to intervene in conflicts between water users. The Direcci6n General de Aguas also grants, upon petition, water-use rights for surface and groundwater that is not already claimed. These grants are a matter of public record, and are published in the Official Journal in both Santiago and the regional capitals. During times of drought the Direcci6n General de Aguas can impound water, with compensation to water-use rights holders. Currently, the Direcci6n General de Aguas is involved in a project to "regularize" water rights. Regularization entails the creation of an inventory of all rights holders, including rights that have not been formally titled. This process includes the specification of water flows and the water-use rights of irrigators holding traditional water-use rights. Regularization does not involve actual titling, which is done at the local real estate registry at the owners expense. Although the Direcci6n General de Aguas does have broad authority in water resources management, much of the actual control over river flows is exercised by the local river monitoring authorities the juntas de vigilancia (JDV). These JDV manage and maintain storage reservoirs, and are responsible for reducing the flow of water to canals during times of water scarcity. These JDV are controlled by the river's water-use rights holders on a particular section of a river. The votes each rights holder possesses depends on the number of shares each owns. Thus in some river basins, irrigators, mining companies, and water supply companies are part of the same JDV. Currently, the court system is the final arbiter of water use conflicts. But the effectiveness of the courts in conflict management has been limited by judicial restraint and an over-emphasis on formalism6 (Bauer, 1993). Recent Events in Water Resources Management The ability of current institutions to effectively resolve conflicts is being challenged, especially in the BioBfo basin in south-central Chile. In this area, irrigation, hydroelectric power generation, industry, municipal water supply, and recreation all compete for use and control of the river waters. Much of the conflict in this area results from the 1981 Water Code's stipulation of non-consumptive rights (Bauer, 1993). Non-consumptive rights were originally designed to encourage the development of hydroelectric generation. The Water Code specifies that these rights shall not interfere with other water-use rights. But it is quite possible that the hydroelectric generation company will schedule water releases from the Pangue Dam on the BioBfo River (under construction) in a manner that would harm downstream irrigators with consumptive use rights. This dispute was initially resolved by a Court of Appeals of Concepci6n ruling that protected downstream users from the unwarranted use of non-consumptive rights in the regulation 6 This is a general characteristic of the Chilean legal system, which has not demonstrated the institutional capacity to creatively resolve conflicts. 10 of river flows. However, this ruling was overturned by the Supreme Court which ruled that non- consumptive rights did include a right to regulate stream flows. The Supreme Court ruling does allow for damaged parties to seek ex-post compensation, but does not grant irrigators ex-ante participation in determining the timing of water flows. This ruling was clearly unfavorable to irrigators since it transfers some of their property rights to a large hydroelectric company, and would require them to absorb the high cost of litigation in order to seek compensation for losses. In 1993 the Direcci6n General de Aguas presented to Congress several amendments to the 1981 Water Code. These modifications would: i) return to the state water-use rights that have been granted by the state but not used for productive ends for a period of five years; ii) strengthen the role of the Direcci6n General de Aguas in monitoring and regulating water quality; iii) establish autonomous river management corporations; and iv) require state approval of new water uses in the extreme northern regions of the country. Although the five year use-it-or-lose- it rule was designed to correct the initial allocation of a large number of rights to ENDESA, Chile's largest electric generation company, most water-use rights holders considered it to be a threat to their right to private property. These modifications have yet to be passed in Congress, and a modified proposal by the Direcci6n General de Aguas is expected. In recommending river management corporations, the Direcci6n General de Aguas was motivated by the principle of decentralized management. Also, the government hoped that the river basin corporations would accept responsibility for watershed management, coordinate water use, and mediate disputes between users. But there was little mention in the proposed legislation of how these corporations should be constituted and how power would be shared with existing JDVs. This lack of specifics led to some confusion and distrust among irrigators who feared that these new authorities would take away power from the JDVs that have been traditionally controlled by irrigators. 11 IV. LOCAL WATER MARKETS AND GAINS FROM TRADE Intersectoral and Intrasectoral Trade in Water-Use Rights Chile has an environment where water use is limited by both resource scarcity and legally defined property rights and where the value of water in varies between different uses. Thus, trade for water-use rights, both between farmers and urban users and amongst farmers, is expected. Because of the relatively low value of water in traditional agriculture, the seller of water-use rights would most likely be a farmer. A farmer would be better off selling water-use rights if the present value of these rights - calculated as the expected value of the discounted marginal product of water - is less than the price offered by a buyer. Of course, the farmer may be paid more than this present value. In this case the difference between the present value of the water-use right in agriculture and the price that is determined by the supply and demand of water would be an economic rent that would accrue to the farmer as owner of the property right. Similarly, the difference between the present value of a water-use right to a buyer and the purchase price of the right is an economic rent that accrues to the buyer. The difference between the value of the water-use right to the buyer and its value to the seller is society's gains-from-trade. A MODEL These concepts can be formalized using the following equations. Assume, for simplicity, that each farmer plants one crop in a particular location every year. Let farmer i, have a profit function: I,(X,y (W,i,L/,P,R)) = PF(X:*(W,i,LtL,P,R)) - RXt* (1) where: tli(-) = farmer i's profit function in year t; F(.) = a well-behaved, increasing, concave, continuously differentiable production function; Xi, = a vector representing the profit maximizing allocation of inputs; Wit = pre-determined quantity of water-use rights available to farmer i in year t; Lit = pre-determined quantity of land available to farmer i in year t; R = a vector of exogenously determined input prices; and P = the exogenously determined output price, and t = (1,2,3...). Since the production function is well behaved, the profit function can be denoted in its reduced form rIt(W,,L't,R,P). Thus, the value to a farmer of an endowment of land and water-use rights Vt(W,,L') is the present value of a discounted stream of profits that can be earned with the use of the land and water, as shown: Vi(W,L,Lti) = E (t( ,,L' ,R,P)), t = (1,2,3,...). (2) 12 And after accounting for the value of land, the value of water V,(IV,) can be derived. Of course, if a farmer does not have positive profits, then the value of water will be negative. The demand for water in urban areas is the sum of: 1) household demand for water in residential use; 2) industrial demand, based on the value of the marginal product of water in industry; and 3) public demand for water for recreational, amenity, environmental, and public health uses. All individuals and firms in an urban area would solve different constrained maximization problems and develop input demands for water use that are dependent on prices, incomes, and production technology. Thus, an urban water company's (UWC) demand for water is a derived demand, in which population and income are key parameters. In Chile, UWCs are profit maximizing firms, but are regulated and required by law to meet this urban demand for water, Dt'. Because rates are determined administratively, these companies will have incentives to minimize the discounted costs, PtXX + Ptwwt (3) COST=Et (1 +r)t subject to Dt* Y(WtU,Xt) (4) and WtU =wt + wt-I (5) where: WCL = the UWC's stock of available water rights in year t, Xt = a vector of non-water inputs into the UWC's production function, Wt = water rights purchased or obtained through alternative means by the UWC in year t, D * = the amount of water that the UWC is required to supply to its customers in year t, ptx = a vector of input prices, ptw = the least cost input price of water rights that are either purchased or obtained through alternative means, Y(WUt,XX) = the UWC's well-behaved, increasing, concave, continuously differentiable production function for water supply, and t = (1,2,3,...). If the constraints are binding, than an UWC's indirect demand for water-use rights in year t can be defined as wt*(Ptw,Ptx, Wu l,D,'). Given an alternative cost of acquiring these rights, ptW(A), the UWC's maximum willingness-to-pay for water-use rights, IAu(w*) can be determined as shown: V (W *) = p "w *(PX,P,,,W,-,,D, )(6 13 For simplicity, assume that an urban water company is buying water-use rights w from farmer i. The net gains-from-trade to society from this transaction is the difference between the value of the water to the buyer and seller less the transactions costs, as shown: GFTtui(w) = [VtU(w) - VA(w(WA1 - w))] - TCu i(w) (7) where: GFTiu,(w) = the gains-from-trade to society of a transfer of w from farmer i to the urban water supply company, TCuYi(w) = the total transactions costs of a transfer of w from farmer i to the urban water supply company, and Vt(w(lt, I- w) = the value of water-use rights w, as a percentage of the average value of total post-trade water-use rights (Wt,, - w). The net rents to farmer i of the trade can be calculated as shown: NRiL(w) = Pw(u9l)w - [V'(w(KA1 - W)) + TCO(W)] (8) where: pw(ui) = the transfer price of w negotiated between farmer i and the urban water supply company, NR'L(w) = the net rent to farmer i of a sale of w to the urban water supply company, and TC(w) = farmer i's transactions costs. Similarly, the net rents to the buyer can be calculated as shown: NRt U(w) = Vt U(w) _ [PW(U'w + TC U(w)] (9) CALCULATION OF GAINS-FROM-TRADE AND ECONOMIC RENTS In order to assess the impact of water markets and transaction costs in Chile, four river valleys, the upper Maipo in Chile's central valley, and the Elqui and Limari in north-central Chile, and the Azapa in the far north of Chile were selected as case studies (see ANNEX I and Maps Section). These valleys were identified based on prior information which suggested that there was - or should be - active trading in the area. Attempts were made to identify all transactions of water-use rights in these areas for the years 86-93. These years were chosen because of reports of more active trading in recent years, because they include both wet and dry years, and because farmers were expected to provide more accurate information on recent transactions. Water transactions in conjunction with land transactions were generally not considered, because they usually do not represent a change in water use patterns. A survey instrument was developed to solicit information from farmers who have participated in water market transactions (see ANNEX III). This information included: i) individual characteristics of buyers and sellers; ii) water transactions, including price, quantity, and transaction costs; iii) farmer's investments in irrigation technology, groundwater, small dams, and canal operation and maintenance; and iv) farmer's crop yields. The survey instrument was pretested with farmers in the Maipo Valley who had not participated in market transactions. 14 Individuals making transactions were identified from records of WUAs, from local Conservadores de Bienes Raices, and from other transacting parties. Agronomists from the Universidad Cat6lica of Santiago were recruited as enumerators. Enumerators attempted to contact all individuals who had made transactions in the areas selected for analysis. Managers of urban water supply companies and industrial users of water were also interviewed. Crop Budget Analysis in Valuing Water Rights The net returns calculated using farm budget analyses is an upper bound to farmer's willingness to pay for water in irrigation. This is not a marginal analysis, but an estimate of the average value of water. The net returns to a farmer is the maximum that a farmer could pay for water and land, and still break even. In this study, farm budgets are used for both an economic and financial analysis of water markets7. Net returns are divided between land and water according to relative market values (solicited in this study using a survey of farmers, see ANNEX III) of land and water used in crop production. A summary of all the assumptions used in the gains-from-trade analysis is presented, along with summary data in Table 1. An example of a representative crop budget is shown in Table 2. Other crop budgets used are presented in ANNEX II. The following paragraphs explain this procedure and the assumptions used. Yields and Production Costs Farmer's per hectare yields are based on survey data. In the case where farmers did not respond to this question, average yields from other surveyed farmers were used. Output price and cost of production information were taken from representative crop budgets used by the Direcci6n de Riego and the Department of Economics of The Universidad Cat6lica of Chile in the financial and economic appraisal of irrigation projects in the Elqui and Limari valleys, dated 1991 and 19928. In situations where either representative crop budgets or yield data were not available for a particular crop grown by an individual farmer, the average value of water for the farmer's other crops was substituted. ' The financial analysis is performed to determine individual incentives to participate in water markets. In contrast, the economic analysis provides the net return to society of market transfers. In an economic analysis, transfer payments are excluded and prices are adjusted for economic values (explained in more detail later in this chapter). 8 Crop budget data was dated 1990 and 1991. 15 Table 1: Assumptions Used in Gains-From-Trade Calculation CROP BUDGETS PARAMETER ASSUMPTION 1. Input Prices Taken from representative crop budgets used in evaluation of irrigation projects in Region IV. INPUT UNIT PRICE 1991 CH $a Labor Daily Rate 1,500 Urea Kg 101 Fosfato Diamonico Kg 107 Sulfato Potasio Kg 116 Citrolliv Liter 513 Parathioin Liter 3,000 2. Output Prices Taken from representative crop budgets used in evaluation of irrigation projects in Region IV. OUTPUT UNIT PRICE 1991 CH $b/ Table Grapes Kg 218 Pisco Grapes Kg 80 Tomatoes Kg 52 Chile Pepper Kg 500 Wheat Kg 79 Corn Kg 500 3. Yields (Kilograms/Ha) Taken from survey data. CROP Ns' MEAN MINIMUM MAXIMUM STANDARD DEVIATION Table Grapes 27 14,900 8,000 22,500 4,445 Pisco Grapes 25 31,484 3,640 80,000 14,251 Field Tomatoes 15 17,230 5,850 30,000 6,999 Greenhouse Tomatoes 6 57,570 37,800 97,200 8,856 Chile Pepper 4 1,600 900 3,500 1,270 Wheat 6 2,380 1,500 3,200 830 Corn 5 2,700 1,700 5,000 1,200 a/ Prices are later inflated to account for 1993 prices. b/ Prices are later inflated to account for 1993 prices. c/ Many farmers could not give yield estimates. 16 Table 1: Continued 4. Costs of Water Delivery Taken from survey data.4' Elqui Valley Total Cost of Water Delivery 1993 $ Ch per Share/Year N MEAN MINIMUM MAXIMUM STANDARD DEVIATION 28 14,464 0.00 ' 124,000 24,066 Limari Valley Total Cost of Water Delivery 1993 $ Ch per M3/Year N MEAN MINIMUM MAXIMUM STANDARD DEVIATION 65 3.77 0.00 18.37 3.09 5. Maragement Premium 7%. 6. Commercial Loan Rate 15%. 7. Interseasonal Finance 50% of input costs for 6 month. 8. Discount Rate 12%. 9. Grape Production Full production after 6 years, which continues for 25 years, no production after the 25 years. 10. Crop Development Loan 3 year disbursement, 3 year grace, and 19 year payment. 11. Net Returns to Land Calculation based on survey responses and and Water farmers' perceived value of land and water-use rights. d/ Farmer's annual payments to water user associations and for additional costs for cleaning canals that are outside a farm gate. s' Some individuals had purchased rights, but not yet transferred water. 17 Table 1: Continued 12. Ratios of Value of Water to Values of Water and Land N"' MEAN MINIMUM MAXIMUM STANDARD DEVIATION Elgui Valley 20 0.43 0.02 0.88 0.28 Limari Valley 120 0.55 0.05 0.93 0.24 13. Term 50 years. 14. Farm Sizes Taken from survey data. 15. Hectares in Production N MEAN MINIMUM&' MAXIMUM STANDARD DEVIATION Elgui Valley 29 17 0 172 36 Limari Valley 69 53 0 1,200 167 16. Number of Interviews BUYERS SELLERS Elgui Valley 18 14 Limari Valley 19 37 POTABLE WATER 17. Growth Rate of Water Demand 2.1% in Coquimbo/La Serena Area 18. Cost of Alternative Water US $23,500,000 (1992) Source in Elqui Valley 19. Value of Elqui River Water see Table 4. for Potable Water Supply " Ratio of the value of water to the value of water and land for each crop for each farmer. &' Both individuals that had left farming and individuals that had purchased water for small country houses had zero hectares in production. 18 Table 2: Representative Crop Budget (Table Grapes) Field TypO' - 50 Ha. Table Grapes, Drip Irrigation - Monte Patria, Limari, IV Region, Chile - with 100 Shares of Water from Cogoti Reservoir Thousand $ ChW' Total Revenue 16,400 kg/ha * 218 $/kg * 50 ha 178,760 Costs' Machinery 48,000 $/ha Labor 126.5 days * 1,500 $/day 189,750 $/ha Inputs 985,961 $/ha Total 50 ha * 1,223,711 $/ha (61,186) Cost of Water Deliveryq' (837) Intraseasonal Finance 50% of costs * 6 months @ 15% (2,335) Net Revenue 114,402 minus 7% management (7,991) 106,411 minus land tax 10% of 50% stated values' (2,500) 103,911 minus debt service for crop development and drip irrigation installation 3 year disbursement, 3 year grace, 19 year payment @ 15% (60,559) 43,352 Net Returns to Land and Water 43,352 minus 29% relative value of land (12,572) 30,800 Net Return per Water Right 308 Present Value of Water Rights @ 12% interest, 50 year term 2,558 a Farm, yield, water, and land information taken from survey data. More than one crop may be grown on these farms. b/ The June 1993 exchange rate of US $ 1.00 = Ch $403 was used. C/ Input and output prices are taken from representative crop budgets used in Direcci6n de Riego analyses. Crop budgets reflect 1991 prices that are later inflated to represent June 1993 values. d/ Survey data on fees paid to WUAs for water delivery and labor assigned to clean exterior canals. Figures reported in 1993 terms were initially deflated to correspond to the rest of the cost information. f Land taxes are included in the financial analyses and not in the economic analyses. Output and Input Prices Because a time series of input prices was not available, relative input and output prices are assumed to remain constant. Output prices for export fruit crops, especially table grapes, can be highly volatile. Because the export price of table grapes is volatile, and has decreased since the dates of the representative crop budgets, the analyses for the Limari Valley was also performed with an alternative price for export grapes of 25% below the reported crop budget price. Most 19 table grape production is assumed to be sold on the high-priced international market, but 25 percent of total output is sold on the national market at a lesser price. Output prices for pisco grapes are determined by the two local pisco production cooperatives and are fairly stable. Cost of Water Delivery Farmer's cost in supplying labor to clean common canals and water fees paid to WUAs are included as production costs, because they account solely for the cost of water delivery. WUAs are non-profit organizations, and they determine their own fees in accordance with their capital and operating costs. Some farmers in the Limari Valley receive water delivery services from the government-owned Paloma Reservoir, but they do not pay for this service. The cost of labor to clean canals within the farm are included in the labor cost of the crop budget. Farmers' total expenditure for water is multiplied by the percentage of water used for each crop as determined in the interview. Debt Service The cost of servicing a debt to cover the costs of initial crop development is included for all permanent crops. These costs are taken from the same crop budget information. For grape vines, partial yields are present in the third year, but at 16% of production this is not sufficient to cover operating costs. Thus the crop budget is representative of a year of full production, with debt service for the first three years of negative net returns, with three more years of grace to allow for full production before debt servicing. The fruitful life span of a grape vine is assumed to be 25 years, after which, yields decline dramatically. Management and Risk Net revenues are discounted by 7 percent in order to account for management services. Seven percent is considered a standard charge for management 9. In order to account for the risk absorbed by a farmer, loan rates used for both interseasonal finance and crop development were set at the 15% commercial rate, which would include a risk premium. Land Taxes In Chile agricultural taxes are fairly complex. A simple option for farmers is to pay a land tax of 10% of an assessed value. The stated assessed value is probably much less than the market value reported in the survey, and for this reason reported market values are discounted by 50%. Land taxes payments were deducted from net revenues in the financial analysis of rents to buyers and sellers. These payments were not deducted from net revenues in the economic analysis because they are transfer payments that do not reflect a net loss to society. 9Alkire (1990) uses this number as a standard, Young and Gray (1972) suggest 6 - 10 % (page 112). 20 Net Returns to Land The returns to land are derived in order to determine the residual value that can be attributed to water. One method of determining the value of land that is independent of the value of water is to use the value of dryland. Computationally this methodology would be attractive because dryland in this area is used only in very low-intensive goat herding and has a value that approaches zero. However, this methodology would fail to incorporate the locational attributes of land that is adjacent to canals or rivers. This land is valuable because it can be irrigated without the cost of constructing canals, because it often can be irrigated without permanent water- use rights during times of high-water, and because irrigated land also has greater accessibility to roads and other infrastructure. Thus, in this study the value of land was determined by the relative value of land and water as revealed in the survey of farmers in which they were asked to estimate the value of their land without water-use rights as well as the value of their water. Value of Water for Potable Water Supply in the Elqui Valley The value of water-use rights of the Elqui River to ESSCO can be determined from ESSCO's ability to use these rights to forego the need to invest in more costly alternatives to obtain water supplies. Since the demand for water in this service area will continue to rise, the least cost alternative for meeting this demand - in the absence of effective water markets - is the appropriate methodology for assessing the value of water. Groundwater in this area is very deep, and of poor quality, and would not be a cost-effective alternative, except for emergency supplies in summer months. Another alternative would be to construct a water storage project in the Elqui River. In fact, ESSCO has opted out of a proposed joint venture with Direcci6n de Riego to construct the Puclaro Dam on the Rio Elqui. At a cost of US $23,500,000, ESSCO was offered sufficient capacity in the proposed reservoir to meet its needs for secure water supplies well into the next centuryl0. If this dam were in place, ESSCO could fill its part of the reservoir during times of high water flows. ESSCO's refusal of this offer coincided with its current program to purchase water-use rights on the Elqui River. Many of the water-use rights transactions in the Elqui valley involved the transfer of water from agriculture to potable water uses for residences outside of ESSCO's command area. A few of these transfers involved purchases of a small quantity of water by small investors interested in developing plots of land along the river - and adjacent highway - for country homes and small tourist cabins. Other transactions involved the same type of investment but in the upper reaches of the Elqui valley. Another purchaser was interested in developing a large tourist and residential area on the beach north of the city of La Serena. Since this area is outside ESSCO's service area, and an agreement to purchase crude water from ESSCO was not guaranteed, this developer began to purchase water-use rights on the Elqui River. '1 Conversation with Direcci6n de Riego, Santiago, June 1992. The Direcci6n de Riego also offered reservoir water at a lower price but without the security of delivery that ESSCO is required to have. 21 In placing a value on water-use rights for these developers it is important to note their possible alternative water sources. Groundwater pumping along the Elqui River is opposed by the Junta de Vigilancia which represents water users. Outside of the immediate valley, groundwater is both expensive to exploit and of poor quality for potable water. Thus, the most likely alternative source of crude water supplies for all of these tourist developments is the same source as ESSCO's. Therefore, the cost of water to ESSCO is also the value of a water-use right to others needing potable water supplies downstream of the proposed Puclaro Dam project. Two necessary assumptions will be made in using this procedure. First it is assumed that these smaller entities can receive crude water at ESSCO's cost, and no additional transactions costs are involved. Another assumption is that this additional water demand does not affect the timetable of water demand and value as shown in Table 3. Consequently, these estimates may undervalue water for smaller developers who may actually have to purchase more expensive water. Table 3: Value of Water Rights for ESSCO's Purchases Year of Year of Demand Number of Value of a Share f Purchase Shares I 1992' 2010-' 45.3 $ 2,561,220 1992 2011 21.9 $ 2,204,690 1993 2011 25.0 $2,475,505 1993 2012 47.7 $ 2,157,305 1993 2013 48.5 $ 1,881,798 1993 2014 36.6 $ 1,643,589 a/ ESSCO purchased 67.2 shares in 1992 and 157.8 shares in 1993. 2/ Projected demand based on current growth in water use. Following a procedure similar to that outlined in Moncur and Pollack (1988), the value of water-use rights is estimated as the cost avoided by not investing in the Puclaro Dam (see Figures 6.1). In year T, ESSCO's current supply of water, provided at a cost of Cl, will be insufficient to meet a growing demand for water. At that time a new water supply, at a cost of C2, must be made available. If, however, the need for this new costly water supply can be postponed, this postponement has a value, dIx, which is equal to the total value of all crude water supplies needed to effect the postponement. Information for the calculation of bt is taken from a 1991 CIAPEP feasibility study of the Puclaro project and from conversations with ESSCO's planning department. To service its demand, ESSCO had in 1991 a total supply of 27.36 million m3 per year, which includes a few wells and drains, as well as 550 shares of the Elqui River, with each share representing 1 22 liter/second". This CIAPEP calculation appears to presuppose that the yearly demand for water occurs at the same time as the yearly supply of water, and that this supply is secure in dry years. This is somewhat of a weak assumption given the very large sumrner population of this beach resort area and the variable flows in the Elqui river. But, river flows during the January- February tourist season are relatively high, and ESSCO enjoys a priority in the delivery of water during dry years, especially for its long-held traditional rights. Furthermore, ESSCO is currently investing in groundwater, largely for the purpose of ensuring emergency water supplies during peak summer periods. According to the CIAPEP study, the demand for water in the La Serena/Coquimbo service area is growing at a rate of 2.1 percent per annum. At this rate ESSCO's current supply of water will be insufficient to meet its needs by the year 2011 (which also is a reasonable date to have expected - in 1992 - the full completion of the Puclaro Dam project12). The value of a water-use right is equal to the costs saved in delaying the need for introducing a more expensive water supply. This value, kt, is the 1992 value of the 570,000 m3 of water needed in 2010 to cover the increased demand for water. As shown, c-2C1 (10) e r(T-t) where: r = .12; T = 2010; t = 1992; and C2 - C, = CH $1,006,054,000 = the yearly payment for a loan of US $23,500,000, at 12% interest for 35 years, at a 1992 exchange rate of CH $350 = US $1.00. Thus, bt = CH $116,023,303 which is the 1992 value of the water needed to fulfill ESSCO's unmet water demand in 2010. One interesting feature of this calculation is that the volume of water needed in 2011 is not a factor in the determination of its value. This is because the Puclaro reservoir would be large enough to meet any reasonable increase in demand. Once the value of the quantity of water needed to delay the investment in the more expensive technology is determined, it is necessary to put this value into a per-share basis. If each right of Elqui River water continues to guarantee 1 liter/second of water, then this amount, 4), is divided by the 18.11 shares that would be required to meet the 2010 demand. However, although " It has been the policy of the Junta de Vigilancia of the Rio Elqui (JDVRE) to supply ESSCO with 1 liter/second of water per share, even when the water delivered to other shareholders is reduced because of low river flows. I2 Fifteen years was needed for full completion of the Paloma Dam project. 23 Figure 1: Calculation of Avoided Cost of Alternative Supply of Water" MARGINAL EXTRACTION COST ($) C2 - d Cl a 0 T YEAR FIGURE 1-1 EXTRACTION COST PATH MARGINAL COST ($) cl I a f l I 0 T T + I YEAR FIGURE 1-2 EFFICIENCY PRICE PATH CONSTANT MC at Source: Moncur and Pollack, 1988. 24 ESSCO's traditional shares carry a priority in water delivery, this priority cannot be guaranteed for future shares. For this reason, it is assumed that 2.5 shares of the Elqui River are needed to ensure each liter/second, because in 95 of 100 months, the lower Elqui River delivers at least 0.4 liters/second per water-use right"3. Thus, 4t = CH $116,023,303 is divided by 45.3 shares to reach a value of CH $2,561,220 for each of the first 45.3 shares purchased in 1992. The value of all 217.8 net shares of Elqui water purchased by ESSCO in 1992-3 are listed in Table 3. Economic Analysis of Gains-From-Trade Financial prices taken from representative crop budgets were adjusted to present economic values. Information from the 1991 InterAmerican Course in Preparation and Evaluation of Projects (CIAPEP) study of the Puclaro Dam was used in this adjustment because it contains both economic and financial prices. Because Chile has a relatively open economy with low tariffs, free exchange rates, and open markets, the difference between economic and financial prices are small. For exportable goods, such as table grapes, a percentage of the exporters commission is added to the farmgate price in order to account for this transfer payment. For imported inputs, financial prices are adjusted for both import tariffs (11 %) and the foreign exchange premium (10%). Adjustment factors for financial prices to reflect economic values are shown below in Table 4. Table 4: Adjustment Factors for Economic Prices E Crop and Price Adjustment Factor Table Grapes, Output Price 1.084 Table Grapes, Production Costs 0.995 Pisco Grapes, Production Costs 0.997 Corn, Production Costs 1.062 Potato, Production Costs 1.020 Wheat, Production Costs 1.050 Tomato, Production Costs 1.012 Alfalfa, Production Costs N/A Chile Pepper, Production Costs N/A 13 Municipal water supply companies are, by law, required to have 95 % security in water delivery. The estimate of 2.5 shares per liter /second, is taken from a review of the JDVRE's records of water flow in the river, this also corresponds with ESSCO's stated plan to purchase 1300 additional shares of the Elqui River by 2020. 25 In both the economic and financial analyses, the avoided cost of investing in the Puclaro Reservoir is used to value water for urban water supply in the Elqui Valley. An adjustment was not made to differentiate between the economic and financial costs of constructing the Puclaro Reservoir. The CIAPEP study does provide a discount factor of .991 for social prices to account for the tariffs paid for imported goods. But this report does not detail the percentage of the construction costs which will be dedicated to imported goods. Because the discount factor is so close to 1.00, and the percentage of imported goods is assumed to be relatively small, the financial and economic costs are assumed to be equal. Gross gains-from-trade were calculated by subtracting the value of water to the seller before a sale from that of the buyer after a purchase. Efforts were made to identify and interview both buyer and seller of each transaction. When only one party of a transaction was interviewed, the value of water from buyers or sellers on the same canal or area for the same year was used to substitute for the other party. In cases where the seller of a water-use right was known not to have used the water prior to the transaction, the right was valued at the weighted average value-of-water to neighboring farmers. This is because unused water is generally distributed to other water users, along the same canal or river. Because of the interconnections in the Limari Valley, average values of water for above and below the Paloma Reservoir were used. In the Elqui Valley water was divided between users of the Rio Claro and Rio Cochiguaz, users of the Rio Elqui, and users of the Canal Herradura. In the Limari Valley, cubic meters of water sold on the spot market, were valued at the weighted average value-of-water to neighboring farmers. All values were adjusted to June 1993 values using Chile's consumer price index. Transaction costs were obtained in the survey. These costs include: fees for attorneys, notaries, and engineers; payments for modifications of canals or gates; the costs of soliciting information; and the value of the time involved in the process. A summary of total transactions costs for the Elqui and Limari Valleys is presented in Table 5. Net gains-from-trade were calculated by subtracting the transaction costs to both buyers and sellers from gross gains-from-trade. Financial Analysis of Economic Rents to Buyers and Sellers Financial prices were employed in the crop budgets used in the analysis of individual economic rents to buyers and sellers. In this analysis, a water-use right that was not used by the seller prior to the transaction was valued at zero. The seller's net rent is the sale price, adjusted to 1993 values using Chile's consumer price index'4, less the value of the water to the seller and less the seller's transactions cost. To the buyer, the net rent is the difference between the value of water to the buyer and the sum of the buyer's purchase price and transactions cost. As reported, the sum of rents to buyers and sellers does not equal the calculated gains-from-trade. "4Prices for each year were multiplied by the following factors: 1986, 3.19; 1987, 2.68; 1988,2.34; 1989, 2.00; 1990, 1.60; 1991, 1.29; 1992, 1.03. 26 Table 5: Transactions Costs Buyers Sellers Buyers Sellers Elqui Elqui Limari Limari l _______________________________________ Valley Valley Valley Valley Total Transactions Costs as a Percentage of Transaction Price 0.21 0.02 0.05 0.02 Costs of Attorney's, Notaries, and Obtaining Legal Inscription of Rights as 0.59 0.79 0.16 0.34 Percentage of Total Transactions Costs Costs of Engineering and Modifying Canal Infrastructure as a Percentage of 0.20 0.03 0.64 0.62 Total Transactions Costs'l Opportunity Cost of Time Invested as a Percentage of Total Transactions Costs 0.20 0.18 0.18 0.00 Costs of Gathering Information on Buyers and Sellers as a Percentage of 0.00 0.00 0.02 0.04 Total Transactions Costs al Estimated expenditures for engineering services, modification of gates and canals, and indemnities for transferring water through canals. Includes expenditures through 1993 and does not include planned expenditures for modifications needed in the future. This is because economic gains differ from financial rents. Also, for some transactions both buyer and seller were not interviewed, and financial rents are calculated and reported only for interviewed parties. RESULTS Only the transactions from the Elqui and Limari valleys were analyzed to determine the gains- from-trade from market transfers. In the upper Maipo valley transactions were rare - except for water-use rights ceded to municipal water companies from developers of urbanized land - and were not included in the analysis. Similarly, in the Azapa valley only a few transactions were identified and gains-from-trade were not calculated. In the Elqui valley transactions are infrequent, but there is significant intersectoral transfer as well as a slow transfer of water-use rights within agriculture. In the Limari Valley, with its well developed system of irrigation infrastructure and well organized WUAs, transactions are fairly frequent. Net Returns per Share As a bases for comparison to demonstrate the difference in the value of water in producing alternative crops, average net returns per share were calculated by crop combining data from the 27 Elqui and Limari Valleys. The average value-of-water in alternative crop production is shown in Table 6'5. For the purposes of calculating average values for this table, water-use rights or shares are standardized across valleys, although as discussed below this is not done in the gains- from-trade analyses. As shown, water used in the production of pisco grapes is very valuable. This may be because pisco production is limited by the major pisco distilling cooperatives, and the value of these purchasing agreements is not included in the analysis. There is a lot of variation in these calculated values, especially for table grape and tomato production. Most of this variability is due to the variability in reported yield estimates. Except for pisco grapes and paprika, the average estimated present value of a share of water in crop production is well below reported market prices for the water-use rights. The Elqui Valley The Elqui Valley in Chile's Region IV supplies water for 18,700 ha of farmland as well as potable water for a medium-sized city (250,000 inhabitants) with a very large summer population (see ANNEX I and Map 2 in the Maps Section). A small tourist industry in several small communities in the scenic valley compliments the large coastal tourism boom. Major crops include table grapes, pisco grapes, other fruit crops, potatoes, and pasture. One small mine in the upper reaches of the basin utilizes water from a tributary. Rainfall is scarce in this region with average yearly precipitation less than 120 mm. Wet and dry years follow a cyclical pattern, with two and three year droughts common. In the Elqui Valley a limited population of 47 permanent transactions for the period of 1986-1993 were identified with the assistance of the JDV and other WUAs in the Elqui Valley. For 41 of these transactions representing a total of 712 shares of the river16" 7, either the buyer or seller or both was interviewed. In total, 14 individuals selling 491 shares and 18 buyers purchasing 467 shares were interviewed'". 15 For farmers who sold water, total net revenue was divided by the aniount of water owned before the transaction. For farmers who bought water, total net revenue was divided by the amount of water owned after the transaction. Net returns are adjusted from the year of the crop budget to 1993 by using Chile's consumer price index. In tabulating averages, all individual farm observations are weighted equally, and not weighted by farm size nor by water use. 16 There are 25,000 total shares in the Rio Elqui each share is supposed to deliver I liter/second in a good year, although 0.5 liters/second is generally considered closer to average. "7 Shares of the Estero Derecho of the Rio Claro, which is not controlled by the JDVRE were considered equivalent to shares of the rest of the river, as suggested by local irrigators. 18 Some individuals were involved in more than one transaction. 28 Table 6: 1993 Value of Water-Use Right Shares in Crop Production, Chile's IV Regional Crop Number Mean Annual Standard Mean Present of Value of Deviation Value of Farms Water Right Water Right l j1993 $ Ch 1993_$ Ch Table Grapes 32 $ 189,259 291,738 $ 1,571,700 constant price l Table Grapes 32 3,920 19,501 32,555 25% reduction in international price Pisco Grapes 31 348,894 248,710 2,897,393 Tomatoes 25 16,304 198,622 135,397 Alfalfa 24 33,014 31,174 274,168 Wheat 9 50,467 42,915 419,109 Maize 5 - 45,964 22,906 - 318,713 Chile Pepper 5 127,934 70,307 1,062,432 Potatoes 5 133,583 65,430 1,109,345 Others 6 231,050 181,338 1,918,753 Paprika, Morron a/ Values are in 1993 Chilean pesos. Water-use rights in both the Elqui and the Limari are included. Ninety percent of the shares sold in this valley have not been used by their sellers in recent years. Some of these sellers had surplus water, others had rights to water along canals that did not conduct water efficiently, others had quit farming, and still other sellers owned rights to water but did not own land'9. Purchase prices of shares of the Rio Elqui are quite variable, reflecting differences in transaction costs and in many cases the particular conditions of buyers and sellers. ESSCO's average purchase price in 1992-93 was near US $1,100 per share. During the same period, small plot developers in the lower Elqui Valley were paying US $2,500 per share for water in a conveniently located canal. Exchange prices between individual irrigators ranged from US $250 to US $1,000. Often this disparity in prices reflects the individual circumstance of the seller. 19 In general water that is not used by its owner is used by neighboring farmers in the same canal or section of the river. 29 Transaction costs for most transfers are low because this is a narrow valley with many short canals flowing directly from the Rio Elqui and its tributaries (see Table 5). Results of the economic analysis of gains-from-trade in the Elqui Valley are shown in Table 7. The average gross gains from trade for the 712 shares were US $84620 per share. With an average transaction cost of US $56 per share, net gains-from-trade are US $790 per share. Table 7: Economic Analysis of Gains-from-Trade: Elqui Valley Elqui Valley | Number of Shares [_U.S. $ per shar&' Total Gains-From-Trade 712 846 Net Gains-From-Trade 712 790 Trades with ESSCO 298 675 Total Gains-From-Trade Trades with ESSCO 298 658 Net Gains-From-Trade Other Intersectoral Trades 63 1160 Total Gains-From-Trade Other Intersectoral Trades 63 1139 Net Gains-From-Trade Intra-Agricultural Trades 351 934 Total Gains-From-Trade Intra-Agricultural Trades 351 839 Net Gains-From-Trade a" The June 1993 exchange rate of Ch $ 403 = U.S. $ 1.00 was used. ESSCO has purchased 292 shares of the river, which amounts to 28% of its current water-use rights. The estimated net economic gains from these transfers (US $658 per share) are relatively modest. One reason for this is that much of the water sold to ESSCO was not used by their previous owners, and the procedure of using the value of water to neighboring farmers may be biased to more productive farmers. This bias would result from the fact that a weighted average of water values is greatly influenced by the value of water to a few profitable water-use rights buyers. 20 The average exchange rate reported by the Central Bank of Chile for June 1993 was used. This rate is Ch $403 = US $1.00. 30 There are some small transfers of water to developers of small residential and tourist housing along the valley and on the coast north of La Serena. The gains-from-trade from these transfers are higher than those for transfers to ESSCO. This reflects the fact that the value-of- water to the sellers of these shares is relatively low. The large gains-from-trade in intrasectoral trades reflects the high value of water to a few profitable farmers buying water-use rights. Results of the financial analysis of individual rents in the Elqui Valley are presented in Table 8. The net rent to 14 individuals selling 491 shares of Elqui water averaged US $ 1,156 per share. In contrast, net rent to 18 buyers, purchasing 452 shares, averaged US $ 3,047 per share. In general individuals and firms buying water-use rights received larger net rents than water sellers. Intersectoral trades produced higher net rents than trades between farmers. The fact that many of the water-use rights sold in this valley were valued at zero because they were not employed by their owners at the time of sale, is a large factor in these high economic rents. Table 8: Financial Analysis of Economic Rents to Buyers and Sellers: Elqui Valley Elqui Valley [ Number of | Number of Mean Net Buyers or Sellers Shares Rent per Share U.S. $V' All Sellers 14 491 1,156 All Buyers 18 452 3,047 Individuals 4 292 1,071 Selling to ESSCO ESSCO 1 218h' 3,104 Other Sellers of 3 32 1,041 Intersectoral Transfers Other Buyers of 6 40 4,890 Intersectoral Transfers Other Sellers 7 167 1,327 Other Buyers 11 194 2,603 a, The June 1993 exchange rate of Ch $ 403 = U.S. $ 1.00 was used. hi Some of ESSCO's shares are reduced upon purchase by the Junta De Vigilancia de Rio Elqui in order to account for conduction losses. 31 The Limari Valley South of the Elqui Valley in Region IV is the productive Limari Valley, which contains the Rio Limari and its tributaries the Cogoti, Grande, Guatalarne, Hurtado, and Rapel. This valley has 50,000 ha of irrigated farmland producing table grapes, pisco grapes, horticultural products, basic grains, and pasture (see ANNEX I and Map 2 in the Maps Section). A central feature of this valley is the presence of a large interconnected system of three interseasonal storage reservoirs: Paloma (750 million m3), Cogoti (150 million m3), and Recoleta (100 million m3). This storage along with flexible gates and well organized WUAs allow for volumetric specification of water-use rights. This is a dry area with mean annual precipitation of 140 mm. There is one small city, Ovalle with a population of 80,000 that draws water from the Rio Limari. In the Limari Valley, a population of all water transactions was not developed. Instead, data collection was concentrated on areas with frequent transactions and individuals with many transactions. Attempts were made to balance interviews in these areas with other areas of less market activity. In the areas of the Rio Hurtado and Rio Grande which irrigate a total of 5400 ha above the Recoleta and Paloma Dams respectively, only a handful of transactions were identified. All other identified transactions were in areas below the reservoirs. In this valley there is volumetric denomination of water-use rights based on the amount of water stored in the three main reservoirs2". Each WUA assigns a different amount of water to each share. Because of this all water-use rights were converted to an average volumetric denomination, based on survey results. These volumetric denominations are presented in Table 9. The low transaction costs and frequent trades in the valley can be attributed to both modern infrastructure and well-developed WUAs. Because of reservoir storage and flexible gates, water is delivered to farmers on demand and a water transfer is almost costless. Thus, the frequency of transfer is high. Consequently, the market for water-use rights in this valley is active. Individuals easily separate water from land, and farmers make marginal water use decisions. In total, 37 farmers selling water-use rights accounting for 2 million m3/year were interviewed along with 19 buyers purchasing 7.2 million m3/year. Another set of 16 farmers who purchased 965,220 m3 of water on the spot market, (mostly during the 89-91 drought) were also interviewed. 21 Because of the volumetric denomination, it is dangerous to directly compare the volume of water rights in the Elqui River to those of the Limari Valley. In the former, 15,750 m3 are delivered in an average year, but without regard to the time of delivery. In the latter, water is delivered to farmers at times that they request water. 32 Table 9: Estimated Volumetric Equivalencies of Shares: Limari Valley Water User Association | M3 Per Year Per Share AC Canal Camarico 3,000 AC Cogoti Reservoir 4,250 AC Canal Palqui-Maurat-Semita 45,000 JDV Rio Limari y Grande 7,000 JDV Rio Guatalame 5,500 Ac Recoleta Reservoir 3,600 Results of the economic analysis of gains-from-trade in the Limari Valley are presented in Tables 10 and 11. An average gross gains-from-trade of US $2.47 for each m3/year transferred was estimated. After subtracting transaction costs the average net gains-from-trade is US $2.40 for each m3/year. Net gains-from-trade in the spot market for volumetric purchases of water were US $ - 0.05. This is because water sold on the spot market is generally not used by its owner, and unused water is divided up among other water users22. Thus, on average, the value-of-water to farmers who purchased water on the spot-market is not much different from the weighted average of the value-of-water to other farmers in their part of the valley. Trades that involved the transfer of water to large grape producers produced higher net returns than other trades. But as shown in Table 13 these gains-from-trade are reduced considerably with a reduction in the export price of table grapes. Results of the financial analysis of individual rents in the Limari Valley are presented in Tables 12 and 13. Net rents to 37 individuals selling water rights, amounting to 1.7 million m3/year averaged US $ 0.00 per m3/year23. Net rents to 19 individuals purchasing water-use rights that deliver 7.5 million m3 averaged US $1.78 m3/year. The zero net rents to individuals selling water reflects the need of many farmers to sell their water-use rights to help make debt payments. The spot market for a specific volume of water is active during dry years. According to WUA managers, the spot market is generally supplied by owners of shares of water not used in irrigation. This water is valued at zero to the seller. Net rent for 16 individuals purchasing 965,220 m3 of water, mostly during the 89-91 drought, was US $ 0.08 per mi3. 22 WUAs in the Limari Valley generally do not allow farmers to save the water they receive for their water-use rights from one year to the next. 23 The actual figure is - 0.0023. 33 Table 10: Economic Analysis of Gains-From-Trade: Limari Valley Constant Table Grape Prices [ Limari Valley Cubic Meters per year U.S. $a, Total Gains-From-Trade 9,218,000 2.47 Net Gain-From-Trade 9,218,000 2.40 Net Gains-From-Trade 965,2200' - .05 on Spot-Market l Total Gains-From-Trade 5,834,000 2.85 Purchases of Large Table Grape ProducersS' Net Gains-From-Trade 5,834,000 2.84 Purchases of Large Table Grape Producers Total Gains-From-Trade 3,384,000 1.81 Other Purchases Net Gains-From-Trade 3,384,000 1.65 Other Purchases a/ The June, 1993 exchange rate of Ch $ 403 = U.S. $ 1.00 was used. b The spot market is for cubic meters of water not m3 per year. c, Table grape producers with over 100 ha in production. Table 11: Economic Analysis of Gains-From-Trade: Limari Valley Export Price of Table Grapes Reduced 25 % Limari Valley Cubic Meters per year_I U.S. $a l Total Gains-From-Trade 9,218,000 0.58V' Net Gain-From-Trade 9,218,000 0. 58s' Net Gains-From-Trade 965,22P' - .05 on Spot-Market . al The June, 1993 exchange rate of Ch $ 403 U.S. $ 1.00 was used. - U.S. $ 0.584. Cl U.S. $0.576. f The spot market is for cubic meters of water not m3 per year. 34 Table 12: Financial Analysis of Economic Rents to Buyers and Sellers: Limari Valley Constant Table Grape Prices Limari Valley Number of Cubic Mean Net Rent Buyers Meters per m3/Year or Sellers per Year U.S. $ All Sellers 37 1,708,000 0.00 All Buyers 19 7,510,000 1.78 Individuals Selling to Large 20 857,000 0.08 Table Grape Producers Large Table Grape 3 5,834,000 2.05 Producers Other Sellers 17 851,000 - 0.08 Other Buyers 16 1,677,000 0.84 Buyers in Spot Market 16 965,000 0.04 ' The June 1993 exchange rate of Ch $ 403 = U.S. $ 1.00 was used. The spot market is for one time purchases of m3 not for m3 per year. Table 13: Financial Analysis of Economic Rents to Buyers and Sellers: Limari Valley Export Price of Table Grapes Reduced 25 % Limari Valley | Number of Cubic Mean Net Rent per Buyers Meters m3/Year U.S. $ or Sellers per Year All Sellers 37 1,708,000 0.02 All Buyers 19 7,510,000 0.24 Individuals Selling to Large 20 857,000 0.12 Table Grape Producers Large Table Grape 3 5,834,000 0.27 Producers Other Sellers 17 851,000 - 0.08 Other Buyers 16 1,677,000 0.16 |Buyers in Spot Market 16 965,000 0.06 t' The June 1993 exchange rate of Ch $ 403 = U.S. $ 1.00 was used. The spot market is for one time purchases of M3 not for M3 per year. 35 Rents to buyers are generally higher than those of sellers. And rents to large table grape producers - purchasing 78% of the water-use rights included in this analysis - are high relative to all sellers. This reflects the high value-of-water to these profitable farms. Although a 25%reduction of the export price of table grapes does reduce considerably the net rents to buyers of water, the rents received by large table grape producers is still relatively large. Much of this difference in individual rents from market transactions is due to the market power enjoyed by buyers. Since transactions prices are generally well known in these valleys, little advantage is derived from asymetric information. Prices range from US $3,000 for a right with an average volume of 4,500 m3/year in the table grape producing area above the Paloma Reservoir to US $500 for a share with the same volume below Paloma. This difference in price reflects both the premium placed on water in the hot, dry sunny uplands, and the prohibition on transferring water-use rights from below Paloma Reservoir to cana's above the dam24. Also, the value of reservoir storage is demonstrated by the fact that a water-use right in the Limari Valley is worth more than a water-use right in the Elqui Valley that delivers five times as much water. Many of the permanent transfers of water-use rights in this area involves large acquisitions of both land and water by a few large table grape exporters. The land and transactions are generally reported separately because of the mobility of water rights in this valley. Some of these transactions entail a shift away from traditional crops to higher valued fruit crops. Also, in the last few years, many small and medium sized farmers have forfeited land and water rights to fruit exporting companies in lieu of debts owed to the companies. The Maipo Valley The first section of the Maipo River supplies water to 4.5 million people in the Santiago area as well as irrigates over 30,000 ha (see ANNEX I and Map 4 in the Maps Section). The river is divided into 8133 shares, each representing 8 liters/second, 85% of the time. In this first section of the river, there have been very few transactions of water-use rights in the past eight years (see ANNEX I). The Metropolitan Sanitation Works Company (EMOS) has contracted a team of lawyers to purchase rights but has purchased only 33 shares of the upper Maipo in the last eight years, with prices averaging US $10,000 per share. The only industrial concern to purchase water rights was a paper mill which made two purchases totaling 4.5 shares. There are also very few permanent trades between farmers. In the large canal systems of five WUAs, distributing 65 % of the irrigation water in this section of the river, there were only a handful of trades between farmers. All of the canals serving these five WUAs have fixed flow dividers regulating the distribution of water. 24 Without this prohibition it would be relatively inexpensive to pump water directly from the Paloma Reservoir to the adjacent grape producing areas irrigated by upstream canals. 36 The Azapa Valley The Azapa Valley, which is located in the downstream section of the San Jose River Basin in Chile's far north Region I, supports 3,280 acres of irrigated farmland and the city of Arica (population 160,000). This valley, bordering the extremely arid Atacama Desert, is very dry and rainfall in the lower reaches is negligible (see ANNEX I and Map 3 in the Maps Section). The Water Supply and Sanitation Company of Tarapaca (ESSAT), which supplies water to Arica, has been able to use rental agreements to meet the short term crude water needs of the city of Arica. ESSAT is renting wells from owners of groundwater rights. Various government agencies, responding to a presidential mandate to give Arica's water supply a priority in government action, have assisted in the process of renting wells. In the negotiations process, ESSAT was able to both invoke the government's appeal to farmers to release water for Arica and the possibility that the state could impound water during periods of severe water shortage. ESSAT has not needed to purchase water from users of a surface canal in the Azapa valley, because the groundwater supplies were made available. Thus the recent additions to Arica's water supply is much more a result of government action than of market activity. In these arrangements ESSAT digs or rehabilitates a well, installs and operates a pump, delivers free-of-charge a few hours of water per week to the rights owner, and pays a rental fee for the use of the well. ESSAT will then pump water from the well continuously, whereas an irrigator would probably only use the water a few hours per week. This rental arrangement is quite lucrative to the individual irrigator, who does not absorb the negative externality of a depleted aquifer, at least in the short run. Since these rental contracts can be terminated by either party, it is probable that a few will be terminated after ESSAT receives water from the wells that are currently being rehabilitated in the Andean highlands. 37 V. OBSERVATIONS AND CONCLUSIONS Water allocation institutions in Chile have a crucial role in the development of markets for water-use rights in local areas. Chile's heritage of privately developed irrigation, and its traditional allocation of river water by shares, has created a favorable climate for the establishment of transferable water-use rights. These rights have secured water supply to irrigators and thus provided an enabling environment for investment in permanent fruit crops and improved irrigation technology. The overall growth in the value of Chile's agricultural output during the last decade can be attributed to various reforms in both input and output markets. The effects of the 1981 Water Code on the agricultural sector cannot be fully separated from the effects of liberalized trade and secured land rights. But, the agricultural sector that is highly dependent on irrigation has expanded without new investments in irrigation infrastructure. Chilean irrigators are also generally content with the codification of their traditional water-use rights. And since water-use rights are a tangible asset, which do not currently face a property tax, irrigators benefit from ownership of property rights even when the market for these rights is inactive. This research has demonstrated that the market transfer of water-use rights does produce substantial economic gains-from-trade in both the Elqui and Lirnari Valleys in north-central Chile. These economic gains occur in intersectoral trades and in trades between farmers, and they produce rents for both buyers and sellers. But buyers, especially large table grape producers in the Limari Valley and individuals buying water-use rights for potable water supply in the Elqui Valley, receive higher rents then sellers. In the Elqui Valley total and net gains- from-trade per share were within the range of recent transfer prices of US $1,000 per share. In the Limari Valley, gains-from-trade from shares of water-use rights are three times the recent transaction prices of US $3,000 for a share of water from the Cogoti Reservoir (one share delivers 4,250 m3 in an average year). One of the most interesting result of this analysis is the relatively modest economic gains from intersectoral trade in the Elqui Valley. Although, the value of water in municipal water supply is high, the value of water to profitable farmers is also high. When water is transferred away from these profitable farmers, the economic gains from this reallocation are small. Even if water is not used by its owner, it is generally used by other farmers. If these farmers are profitable, than the economic gains of the reallocation are small, even though the financial gain to the seller is large25. These four case studies (see ANNEX 1) demonstrate the diversity of water allocation and water management in northern and central Chile. In three of the four areas studied, especially where large canal systems use fixed flow dividers, market transaL,ions were uncommon26. Despite the fact that these valleys were chosen for analysis because of expectations that they had 25 In most years, water is used by someone in the valley and is not wasted. 26 These four valleys were selected for analysis because of prior information that there are active water markets. 38 relatively active local water markets, only the Limari Valley had active trading. In this valley, transactions costs are low and trade between farmers are facilitated by the presence of reservoir storage, adjustable canal gates with flow meters, and well organized WUAs. In the Elqui Valley, the presence of many short canals flowing from the river also reduce the costs of physically changing the flow of water. In large canal systems with fixed flow dividers, the cost of changing stream flows might be prohibitive especially for trades among farmers, as suggested by the scarcity of trading in the Canal Bellavista system in the Elqui Valley and the large canal systems in the upper Maipo Valley27. Indeed, outside of the Limari basin farmers do not seem to make marginal decisions with water-use rights. Sales of marginal amounts of surplus water are far from typical. Investment decisions and crop selection have been based on the joint land/water allocation that resulted from the final stages of land reform28. Furthermore, WUAs play an important role in facilitating the market reallocation of water, especially in the Limari Valley where trading is active and in the Elqui Valley where intersectoral trading occurs. Many of these WUAs have been able to adapt to the needs of their members for services that facilitate or impede transactions. One possible explanation for the lack of intersectoral trading in the Rio Maipo is the failure of EMOS, the local municipal water supply company, to gain the cooperation of the Sociedad del Canal del Maipo, in its plans to purchase water-use rights. The proposed Puclaro Dam project, on the Elqui River upstream of La Serena, is an example of how the presence of a market alternative to water allocation may reduce political pressure to invest in large water storage projects. This project has been proposed in different forms since 1956. As recent as 1989, potable water was considered to be the most important benefit of this dam. But when ESSCO did not agree to collaborate with the Directorate of Irrigation in paying for part of dam construction, the political importance of the dam declined. This research demonstrated the economic and financial gains from using markets to reallocate water in the Elqui Valley instead of investing in a large water storage project. But it also showed how public investrnents in water storage and delivery systems in the Limari Valley have the external benefit of reducing the transactions costs involved in market trading. In the presence of a system of transferable water-use rights further analysis of large water storage projects should consider the value of storage and adjustable gates in facilitating the reallocation of water. 27 When transactions occur in canals that are divided using fixed flow dividers (marco partiadores), it is necessary to measure and modify water flows through many of the individual fixed flow dividers within a canal system. For instance, if a tertiary canal delivered water to six farms (1,2,...,6) in succession, and farmer 2 sold water to farmer 6, than fixed flow dividers to farms 2 through 6 would have to be modified and recalibrated by an engineer. If trades occur outside the tertiary canal the modifications become more complicated. 281In the final stages of land reform, land expropriated prior to 1974 was divided into family farms equivalent in value to 8 irrigated ha. Water rights traditionally assigned to the land were distributed to the land owners. 39 There is little danger that the current allocation of water-use rights will lead to the exploitation of urban water consumers. Indeed in the valleys of Elqui, Limari, and Azapa, there is a general deference to urban water supply. Potential disputes are resolved in favor of municipal water supply without much conflict. One important reason for this is that in these areas there is little rural-urban competition. The urban areas of La Serena/Coquimbo, Ovalle, and Arica serve as service centers for the local agricultural areas. Most large irrigators and leaders of WUAs have houses and businesses in these communities, and don't want them to be short of water. Thus, in the Elqui Valley, ESSCO receives a priority on water delivery during periods when the Rio Elqui is low. Also, in the Limari Valley, ESSCO has received groundwater rights directly from the banks of the Rio Limari near Ovalle, without significant opposition from the JDV that manages this river. In the Azapa Valley, the aquifer is being rapidly depleted in order to supply water to the city of Arica. Despite the fact that water users have been able to avoid serious conflicts in these valleys, the absence of institutions for intersectoral discussion and conflict resolution is apparent. In the Maipo, Elqui, and Limari valleys, there was no indication that market allocation of water has contributed to environmental degradation, reduced water quality, or caused a decline in the use of return flows. In the Azapa valley, ESSAT's increased access to groundwater rights has led to an accelerated decline in aquifer storage. Despite this, there is little indication - from these case studies - that the 1981 Water Code needs to be modified in order to preserve the environment. However, accompanying Chile's economic growth is an increasing demand for environmental amenities, including fresh-water based tourism and recreation. Since many of these environmental amenities are public goods that are not well defined in a system of private property rights, efforts by either the DGA or the proposed river basin management corporations to protect the environment should prove to be beneficial. The current 1981 Water Code does include some mechanisms to control certain negative externalities. The water code does not recognize the rights of users of return flows, and thus significantly endangers downstream users. But the DGA has the limnited authority to deny permission to change river intakes. It could use this authority to limit any significant change in the location of water use that would adversely affect downstream users. Also, when aquifers are being drained by over pumping, the DGA has the authority to restrict the right to withdraw groundwater. It has done this in the Copiapo Valley, but not in the Azapa Valley due to the government's stated mandate of securing water for Arica. One modification to the 1981 water code that could have a significant economic benefit is a property tax on water-use rights. Such a property tax on water-use rights could provide an incentive to increase the efficiency of water use, and would stimulate the reallocation of water in areas where individuals are holding rights for speculation or in a relatively low valued use. Also, this proposed tax would give the government the incentive to rapidly regularize all water- use rights and designate ownership of all shares. The need for a flexible approach in the development of river management corporations is demonstrated in the Limari Valley where WUAs are acting independently to try and form a river management company that will allow them to take control of the Paloma Reservoir. These WUAs have experience managing the Cogoti and Recoleta reservoirs and have demonstrated that they have the ability to manage large canal systems. But these WUAs have yet to accept the 40 challenge of developing the hydroelectric generation capacity of their dams. This suggests that current WUAs, designed specifically to address the needs of irrigators, may lack the incentives needed to ensure sustainable intersectoral development, especially in the presence of interdependencies in consumptive and non-consumptive use. Implications for Other Countries This study has several important implications for other countries faced with water scarcity. First, there are significant gains-from-trade that can be realized by fostering water markets. These gains occur with both intersectoral trades and trades between farmers. Second, transferable water-use rights that are separate from land are an important requirement for water markets. These rights can be stipulated by volume or by percentage of river or canal flow. But in areas where water supplies are highly variable, it is necessary to designate how water is allocated during times of scarcity. Third, great care should be exercised in the initial allocation of water-use rights among users in order to make sure that all the rights are not captured by a few individuals. If an equitable distribution of land and water already exists, it is relatively easy to distribute water-use rights to the owners of land on which the water has been used. Fourth, proper technology and institutions such as adjustable gates and effective water user associations can substantially reduce transactions costs and facilitate market trading. Fifth, the presence of privately held water-user rights does not necessarily reduce the possibility of proper environmental management of rivers. Water quality regulations need to be established and enforced irrespective of the water allocation system. In Chile, where river valleys are relatively short, the quantity and quality of return flows may be less problematic than in other countries. Finally, within a decentralized system of water resource management there is a continuing role for water management authorities in enforcing rights and resolving conflicts. Yet if water-users and local authorities are able to resolve conflicts among themselves, they can avoid the need for intervention from an outside government entity. Suggestions for Further Research An analysis of intersectoral transfer of water-use rights should incorporate a study of land use patterns. This is especially true in expanding metropolitan areas, with significant suburban growth, where transfers of water-use rights from agriculture to the city tends to correlate with changes in land use. Also, changing land use patters may lead to situations, as in Santiago, where urban and suburban land continue to have irrigation wate