WATER GLOBAL PRACTICE DISCUSSION PAPER JUNE 2019 Dustin Garrick, Dividing the Water, Lucia De Stefano, Laura Turley, Isabel Jorgensen, Sharing the Benefits Ismael ­Aguilar-Barajas, Barbara Schreiner, Renata de Souza Leão, Lessons from Rural-to-Urban Erin O’Donnell, and Avril Horne Water Reallocation About the Water Global Practice Launched in 2014, the World Bank Group’s Water Global Practice brings together financing, knowledge, and implementation in one platform. By combining the Bank’s global knowledge with country investments, this model generates more firepower for transformational solutions to help countries grow sustainably. Please visit us at www.worldbank.org/water or follow us on Twitter at @WorldBankWater. About GWSP This publication received the support of the Global Water Security & Sanitation Partnership (GWSP). GWSP is a multidonor trust fund administered by the World Bank’s Water Global Practice and supported by Australia’s Department of Foreign Affairs and Trade; the Bill & Melinda Gates Foundation; The Netherlands’ Ministry of Foreign Trade and Development Cooperation; Norway’s Ministry of Foreign Affairs; the Rockefeller Foundation; the Swedish International Development Cooperation Agency; Switzerland’s State Secretariat for Economic Affairs; the Swiss Agency for Development and Cooperation; Irish Aid; and the U.K. Department for International Development. Please visit us at www.worldbank.org/gwsp or follow us on Twitter #gwsp. Dividing the Water, Sharing the Benefits Lessons from Rural-to-Urban ­Water  Reallocation Dustin Garrick, Lucia De Stefano, Laura Turley, Isabel Jorgensen, Ismael Aguilar-Barajas, Barbara Schreiner, Renata de Souza Leão, Erin O’Donnell, and Avril Horne © 2019 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org This work is a product of the staff of the World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of the World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of the World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because the World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for non-commercial purposes as long as full attribution to this work is given. Please cite the work as follows: Garrick, Dustin, Lucia De Stefano, Laura Turley, Isabel Jorgensen, Ismael Aguilar-Barajas, Barbara Schreiner, Renata de Souza Leão, Erin O’Donnell, and Avril Horne. 2019. “Dividing the Water, Sharing the Benefits: Lessons from Rural-to-Urban Water Reallocation.” World Bank, Washington, DC. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Cover design: CheryRamalho / Shutterstock Cover photo: Jean Franz, Franz & Company, Inc. Contents Acknowledgments iv Summary 1 Water Reallocation in an Urban World 4 Reallocation Defined 9 Methodology 12 Understanding Reallocation 15 What Drives Reallocation? 15 How Is Water Reallocated in Practice? 18 Navigating the Controversies of Reallocation 23 Can Water Reallocation Deliver Efficient and Equitable Outcomes? 24 Key Lessons Learned 32 Notes 33 References 34 Boxes 1.  Assessment Questions 14 2.  Reallocation through North to South Pipeline in Melbourne 19 3.  Cycles of Reallocation Amid Change in São Paulo 22 4.  Narrow Compensation Approach in Mokopane 27 5.  Portfolio Approach to Compensate Region, not Individual, in Monterrey 29 Figures 1.  Types of Water Reallocation 9 2.  Moving from Compensation to Benefit Sharing 26 Maps 1.  Urban Extent of the Monterrey Metropolitan Region 5 2.  Expansion of the Regional Water Supply System, Including El Cuchillo Reservoir 6 3.  Rural-to-Urban Water Reallocation, following Garrick et al. (2019) 12 Tables 1.  Selected Attributes of the Water Reallocation Case Studies 13 2.  Trends and Drivers for Rural-to-Urban Water Reallocation 15 3.  Factors Driving Administrative Reallocation from Agriculture to Urban Water Uses 16 4.  Benefit Sharing: Form, Amount, Recipients, and Financing 28 Dividing the Water, Sharing the Benefits iii Acknowledgments The authors of the report gratefully acknowledge the initiative and support of Winston Yu at the World Bank, as well as a wider set of feedback provided during a brown bag lecture and associated visit on September 16–17, 2017, and a decision meeting in late October 2017. The authors are grateful to the 60 experts and stakeholders who shared their knowledge and perspectives during interviews in Monterrey, Melbourne, Mokopane, and São Paulo. For their valuable contributions to the final report, we thank Kaye LaFond (mapping), Codi Kozacek (managing editor), Celeste Robitaille (editorial support), Jesper Svensson (research support), Daniel Chico and Charles Wight (mapping support), Abigail Juárez-Rivero (data processing and research support), José Polendo-Garza (data processing and research support), and Erin Barrett (production support). ­ iv Dividing the Water, Sharing the Benefits Summary Global urban populations are expected to grow by approximately 2.5 billion people by 2050, with 90% of this growth occurring in Africa and Asia. Rural regions, including agriculture, are often seen as key sources of urban water supply, creating pressure for reallocation and agriculture. How effective potential hotspots of competition for water between cities and ­ and equitable is reallocation from rural to urban regions, and what have we learned from the experience. It examines global experience? This synthesis report draws lessons from global ­ the drivers, processes, politics, and outcomes of reallocation based on a review of the litera- ture and insights from four in-depth case studies where governments have reallocated rela- tively large volumes of water from rural to urban regions: Melbourne, Australia; Mokopane, Africa; Monterrey, Mexico; and São Paulo, Brazil. South ­ The findings suggest that water reallocation can play an important role in regional devel- opment. However, reallocation projects have also been controversial because of distribu- tional conflicts regarding who wins and loses. The sources of disputes depend on local conditions. Despite these differences, the literature and case studies reveal several common challenges and flashpoints of conflict: (i) disputes about water rights and who owns the water, (ii) the treatment of informal water users, (iii) the level of water use efficiency (and perceptions of waste) in both the rural and urban regions, (iv) the arrangements for monitoring and level of confidence and trust in the data, and most important, (v) the type, ­ amount, source, and recipients of compensation for the donor region. The concept of benefit sharing, long applied to transboundary river basin management, offers a framework for designing effective and equitable reallocation processes, shifting the focus from dividing the water to sharing the benefits among rural and urban regions. The report identifies seven key lessons for realizing the potential of reallocation, and limiting the risks: ­ 1. Benefit sharing requires a systems perspective. Cities are part of urban-rural systems. Reallocation, therefore, involves multiple economic, social, and physical linkages between rural and urban regions, requiring a systems perspective to understand the changing pat- terns of water use and the associated redistribution of benefits and costs over multiple time scales. A systems perspective can improve reallocation processes in all phases, including the design of compensation mechanisms to share benefits with the donor region. 2. Get the numbers—and the water balance—right. Reallocation is an inherently political deci- sion, but should be informed by reliable data, robust accounting, and sound modeling. This includes tracking changes in water supply, water use, and water use efficiency; designing alternatives for allocation rules; and understanding and measuring costs and benefits of reallocation and their distribution. A sound understanding of the water bal- ance underpins effective and equitable reallocation, which in turn depends on robust accounting that prioritizes the data, monitoring systems and evaluation processes needed for designing and delivering reallocation projects. Dividing the Water, Sharing the Benefits 1 3. Water sharing agreements should be backed by sound governance. Reallocation from rural to urban regions involves a form of water sharing agreement across sectors and some- times political borders. Like other transboundary agreements, rules should guide the content, and also the governance processes. Clear rules should define the volume of water transferred, the timing and location of deliveries, monitoring and compliance, compensations, and the processes and parties involved in both planning and decision mechanisms and making. Moreover, effective agreements require conflict resolution ­ agreements are com- well-functioning water governance. Coordination will ensure that ­ frameworks, and other local patible and complementary with national laws, regulatory ­ and regional institutional arrangements. Regional authorities and river basin organiza- tions can offer the platform for effective negotiation, conflict resolution, and agreements. 4. Inclusion and transparency preempt future troubles. Any durable reallocation requires effective engagement with the donor region and ensuring inclusion of key actors. This may require taking into account informal or illegal water users. Failing to do so during the initial planning stage can breed resentment, distrust, and even technical problems that will likely affect later stages of negotiation and implementation. For this, communication and transparency are essential at the outset and throughout the process. This can require having a good communication strategy in place and well-defined venues to ensure engage- ment. The process and manner for sharing the benefits and spreading the costs of reallo- cation should be deemed fair by the key actors to get buy in and broad support. Misinformation, distrust, rumors, and false myths can cost time and resources in the long run. 5. Droughts and crises can tip the balance, but rushing carries risks. Urbanization and economic development are the ultimate drivers for reallocation from rural to urban areas, but real- location rarely happens until there is alignment of enabling conditions (physical, finan- cial, institutional, social) and key triggers. For example, droughts often present windows of opportunity but raise the risk of rushed decisions. In other circumstances, the catalyst comes from political changes or financing. Seizing the opportunities, and limiting the risks, requires preparation of the technical analysis and engagement before these win- dows open. 6. Benefit sharing can transform conflict into cooperation. Broader approaches to compensa- tion can help reduce conflicts and support regional development opportunities for cities and rural livelihoods. The scope and scale of compensation—who receives it, who pays, and how much—should account for the needs of the donor region as a whole, not only the individuals, to capture regional economic impacts of reallocation. Forms of compensation include financial payments, investment in infrastructure, and alternative water supplies. These should be guided by efficiency, fairness, and inclusiveness. It is also critical to mea- sure and understand the socioeconomic impacts and implications of water reallocation 2 Dividing the Water, Sharing the Benefits (avoiding or minimizing preventable negative impacts on poor households, which often suffer most). 7. Be ready to learn and adapt. Even permanent reallocation requires adaptation to changing circumstances and unintended consequences, both in the donor and the recipient regions. Adjustments to reallocation agreements may be necessary due to population and eco- nomic growth in the donor region, changing legal and institutional conditions, and the emergence of new needs. To the extent possible, periodic reviews should be built into the process to avoid crisis-driven responses. Sound accounting, monitoring, and evaluation are needed to make adjustments to the original agreement. Dividing the Water, Sharing the Benefits 3 Water Reallocation in an Urban World water. Urbanization, population growth, and rising incomes are intensifying competition for ­ Approximately 2.5 billion additional people are expected to live in cities by 2050, with 90 per- ­ sia. Urban demand is projected to increase by up to 80 percent by 2050, and cent in Africa and A many cities already face supply-demand imbalances and shortages during drought (Flörke, Schneider, and McDonald 2018­). These trends create a pressing need for policies, incentives, and ­ ommunities (Civitelli and infrastructure to share water between cities and surrounding rural c Gruère 2017). There is a growing recognition that doing so will require a “fundamental ­re-thinking of water rights and appropriate governance mechanisms” (World Bank 2016, 20­ ). ­eallocation— For rapidly urbanizing regions experiencing water scarcity, water resource r a change to the volume, timing, location, or quality of water delivered under formal or informal ­mportant (Hommes et al. 2019). River basin closure— water rights—is becoming increasingly i when downstream water needs are unmet—and groundwater overdraft are signs of an imbal- ­ emand (Wester et al. 2008). In the context of growing water stress, ance between supply and d reallocation presents a management option to change patterns of water supply in response to demands. Reallocation occurs and is enabled by water supply augmentation and shifting ­ ). As lower cost options for demand management, and is often pursued with them (Molle 2006­ dealing with water scarcity are exhausted, cities and countries may explore large interbasin transfers (as with Monterrey, Mexico; maps 1 and 2) or pursue reallocation using existing ­ infrastructure. For example, the city of Monterrey (maps 1 and 2) depicts a common trajectory of water development. The metropolitan area has expanded rapidly in response to the city’s emergence as a hub for economic development and international trade. The sprawling urban area has relied on water reallocation from surrounding regions, expanding the regional water supply network far beyond the metropolitan borders. Water reallocation is therefore central to ­ enerate. the emerging paradigm for sharing water resources and the benefits they g agriculture. The era of reallocation is likely to involve increasing linkages between cities and ­ cities. In places and eras ranging from ancient Water has always moved from rural regions to ­ Rome and Los Angeles in the early 20th century to contemporary initiatives from Mexico City to ­ ecurity. Yet the Kathmandu, Nepal, imported water supplies have offered a path to urban water s scale, intensity, and dynamics of the water imports are rapidly ­changing. Urban water infrastruc- ture now moves approximately 500 billion liters per day over a distance of 27,000 kilometers; 12 percent of cities over 750,000 people rely on interbasin transfers, many drawing water from rural regions (McDonald et ­ ). A systematic review of the global experience identified al. 2014­ 103 reallocation projects reallocating approximately 16 billion m3 of water per year from rural to urban regions, serving 69 cities with a 2015 population of almost 400 million (Garrick et al. 2019). increase. Almost half the Competition between cities and agriculture is projected to ­ world’s surface water dependent cities with a population over 750,000 are expected to com- pete with agricultural regions for water by 2050. Even where water is abundant, the failure of local utilities to deliver safe drinking water prompts informal water vendors to fill the gap 4 Dividing the Water, Sharing the Benefits MAP 1. Urban Extent of the Monterrey Metropolitan Region 04.59 18 27 36 Kilometers N 1970 1985 2010 2014 25ºN 100ºW al. 2016. Source: Adapted from Sisto et ­ by pumping water from agricultural tube wells for growing slums and cities (Venkatachalam ). Water reallocation from agriculture to urban uses can also occur implicitly as land use 2015­ changes. The conversion of prime cropland is expected to cause a 3.2 percent reduction in ­ cropland by 2030; local hotspots are much higher than the global average, such as in Vietnam percent). In addition, urban food sys- (10 percent) and the Arab Republic of Egypt (over 30 ­ ­ reating tems are embedding agricultural water use within the city and its supply systems, c ). novel allocation and management challenges (d’Amour 2017­ economic. First, the economic productivity of The standard justification for reallocation is ­ water varies across competing uses, and there is potential to increase social welfare by ­ uses. The marginal benefit moving water from relatively lower valued uses to higher valued ­ agriculture. Second, of supplying water to a city often far exceeds the value of that water in ­ agriculture uses the most water in arid and semi-arid regions often with low water use ­ efficiency. Recent modeling projects that a 10 percent increase in irrigation water use effi- ciency over 50 years would reduce anticipated urban water deficits by over 2,600 cubic megameters and support more than 230 million people (Flörke, Schneider, and McDonald Dividing the Water, Sharing the Benefits 5 MAP 2. Expansion of the Regional Water Supply System, Including El Cuchillo Reservoir State/province boundaries Lake/reservoir River Canal/pipeline Source: Adapted from Sisto et al. 2016. Note: The population of the metropolitan area of Monterrey increased from approximately 270,000 in 1950 to almost 3 million ). The regional in 1990, when El Cuchillo Dam was approved, expanding the spatial extent of the metropolitan area (map 1­ water supply system expanded as a result to draw water from greater distances, including the El Cuchillo project, which diverts water historically used by downstream irrigators. ). However, the potential for water savings needs to be considered carefully given the 2018­ paradox of irrigation efficiency, whereby farm-level efficiency gains translate into increased al. 2018­ consumption at the basin scale (Grafton et ­ ). Increasing irrigation efficiency involves technical, financial, and political challenges to ensure water accounting, reduce water ­ eallocation. extractions, generate sufficient capital, and address the distributional impacts of r The practical and political difficulties of reallocation have meant that the economic ratio- nale has not been sufficient due to the distance, cost, and contentious politics involved reallocation. The differences in marginal productivity may miss the system-level with ­ ­ ulnerable. value of water in agriculture and equity impacts on the most marginal and v 6 Dividing the Water, Sharing the Benefits Further, realizing gains in water use efficiency creates profound technical, political, and challenges. As a consequence, water planners follow the path of least resistance, economic ­ in which political, social, and economic costs are considered together, rather than assess- ). These politi- ing only financial criteria and cost-benefit analysis (Molle and Berkoff 2009­ cal costs of reallocation can be formidable and have stymied or delayed projects that are alone. justified on economic grounds ­ The growing global experience with reallocation highlights the importance of political economy issues: the distribution of benefits and costs and the winners and losers associated decisions. Reallocation from rural regions to cities is expected to reduce agri- with allocation ­ declines. This creates a potential zero-sum game: cultural production as water availability ­ cities benefit from the additional water supplies, while reduced water supplies for the agri- ­obs. In this scenario, water security cultural region threatens agricultural production and j ­ griculture. Owens Valley of California for the city comes at the expense of water security for a approach. Land and water rights purchases in the is frequently cited as an example of this ­ early 20th century led to the retirement of agricultural lands to supply water for Los Angeles and Southern ­California. Owens Valley benefited from rising land values, which increased by a factor of 11 compared with a doubling of land values in a comparable rural county during the same ­period. However, the lion’s share of benefits were captured by Los Angeles, ­creating the perception of an unfair deal and lingering resentment that has constrained reallocation ). from agricultural to urban uses today (Libecap 2008­ As pressure for reallocation from rural to urban areas grows, so have fears about the households (Birkenholtz 2016). Prior reviews note the poten- consequences for poor rural ­ tially adverse consequences for equity, environmental sustainability, and the livelihoods populations. Coimbatore, India, offers an illustration of this threat: adminis- of poor rural ­ trative reallocation projects since 1960, coupled with informal groundwater markets, ­ ses. Reallocation have steadily reallocated water from agricultural to urban and industrial u has reduced farm income by 50 percent at the tail end of irrigation systems in the Bhavani River basin and has increased poverty in farm households by 12 percent; groundwater depletion to meet urban and industrial needs has reduced water levels (Meinzen-Dick and ). Reallocation in Coimbatore has contributed to economic diversification Ringler 2008­ and urban water security, but it has raised the specter of deepening inequalities across the ­ ehind. rural-urban divide that could leave the poorest households b More recent experiences with reallocation have strived to share the benefits of reallocation as equitably as possible, including a negotiation process seen as legitimate and fair by the region. In these circumstances, compensation and benefit sharing measures are donor ­ increasingly used to offset the negative impacts of reallocation, and often leave the donor off (Dai et al. 2017). For example, financial payments to farmers or irrigation region better ­ districts can offset lost production or facilitate farm-level efficiency improvements to main- productivity. Other forms of compensation are proliferating: investments in tain or increase ­ irrigation modernization, alternative water supplies, and other benefit sharing programs in Dividing the Water, Sharing the Benefits 7 the donor region are means of maintaining or enhancing agricultural production in the region. These opportunities for win-win outcomes depend on the type and level of donor ­ negotiation. compensation and ­ Despite these concerns, experience with water reallocation is growing, leading to new thinking about the drivers, types, and effectiveness of reallocation, as well as lessons learned ­ rojects. It has been 20 years since World about how to design and implement reallocation p Bank research framed opportunities and surveyed innovations in water reallocation and conservation (Bhatia, Cestti, and Winpenny 1995), and 10 years since Molle and Berkoff (2009) conducted a broad review of reallocation as a means of sharing water between agri- ­ ities. Garrick et al. (2019) contributes to this research with a systematic review culture and c of the global experience with reallocation from rural to urban regions. We explore whether ­ hocks. reallocation can be an effective and equitable response to resource scarcity and s 8 Dividing the Water, Sharing the Benefits Reallocation Defined Water reallocation changes the prevailing pattern of water use and can be defined as change to the volume, timing, and location or quality of water delivery, intended to deliver specific ­goals (Grafton, Garrick, and Horne 2017). Reallocation occurs within a legal and institutional framework of formal or informal water rights governing who gets water, how much, when, and where, as well as who is involved in decisions about water allocation and r ­ eallocation. Economic theory distinguishes an initial allocation from reallocation because the two governance. An ‘initial’ allocation establishes processes may involve different modes of ­ water rights, while subsequent changes are governed by rules that determine when and redistributed. An initial allocation implies a formal process by the state how water can be ­ or community to define rights or licences to access, withdraw, use, manage and trade water, and is often accompanied by an assessment of the renewable supplies under different systems. Rarely is there a blank ­ infrastructure ­ slate. Informal water claims almost always ­ ormally. Therefore we predate efforts by governments and communities to allocate water f argue that allocation almost always involves reallocation, either from informal users, the environment, or ­ both. Reallocation from rural regions to cities involves changes in the type and/or place of water ­ urposes when the existing allocation is deemed physically impos- use from rural to urban p sible, economically inefficient or socially unacceptable (Marston and Cai 2016). This occurs when urbanizing regions approach or exceed full development of local water supplies, ­ egions. The nature of water use requiring additional, higher cost water from surrounding r ­ griculture. A substantial portion of water varies between urban and rural types, particularly a is consumed by crops in agriculture in contrast to drinking water and industrial uses, which quality. produce wastewater and changes in water ­ Water reallocation takes many forms: administrative agreements, negotiated settlements, market-based transactions, court decrees, and implicit processes of water reallocation ­ hange (Hooper and Lankford 2018). In practice these different approaches through land use c are often combined or occur in sequence, and they are contingent upon the institutional resolution. framework for water planning, water rights, and conflict ­ Two interrelated characteristics shape the political economy of reallocation from rural regions to cities: (i) to what degree the reallocation decision is voluntary (figure 1), and water. (ii)  whether and how existing users are compensated for the loss of ­ FIGURE 1. Types of Water Reallocation Voluntary Involuntary th fer ts e it d e tiv re te ic ke t or ans pl ec tia ef ra ar Im td ist go M r lt in ur Ne a m eg Co Ad Ill Dividing the Water, Sharing the Benefits 9 ­ egotiation. Voluntary reallocation is associated with market-based transfers or collective n Voluntary approaches require participation by existing users in the reallocation decision, and, typically, a compensation scheme deemed fair by the existing water users and other affected ­stakeholders. By contrast, some court decisions or administrative decrees involve involuntary reallocation by suspending or terminating existing water rights based on legal ­ rust. Involuntary real- prerogatives of the granting authority, such as to serve the public t location processes may trigger negotiation to resolve the grievances raised by those decision. Water reallocation can occur implicitly, such as land excluded from the initial ­ use conversion, a gradual shift in the operating rules for a reservoir storage, or informal ­ wellers. water vendors moving water from rural regions to supply urban and peri-urban d Hermosillo, Sonora, in Mexico, illustrates how multiple forms of reallocation can occur in the same place as cities grow and the institutional arrangements governing water sharing shift with time: market-oriented transfers from agricultural groundwater wells in 2004 were followed by negotiated transfers in 2005 and construction of an aqueduct (involving ). an administrative decision) in 2009 (Scott and Pablos 2011­ Political and technical challenges associated with water reallocation increase with dura- sectors. First, reallocation can be temporary or tion, distance, or involvement of multiple ­ perpetuity. Permanent reallocation permanent, ranging from part of an irrigation season to ­ typically involves more conflict than a temporary agreement due to the distributional issues ­rreversible. In Southern involved and the perception that the change in water use is i California, permanent acquisitions from Owens Valley created significant disputes over val- uation, negotiation, and enforcement compared with the contemporary dry-year agree- ments used to reduce agricultural water use temporarily to buffer shortages during sustained ­droughts. boundaries. Reallocation Second, reallocation can occur locally or span political and basin ­ over longer distances creates greater potential for conflict due to the cascading impacts in ­ ystem. the donor region and the impacts on the local water balance as water departs the s Rural-to-urban reallocation projects range from a few kilometers for tanker trucks operating at the peri-urban fringe to interbasin transfers moving water over great distances, such as ­ roject. Costs, infrastructure, and China’s 4,000-kilometer South-to-North Water Diversion P ­ istances. disparities in costs and benefits are expected to increase as water moves longer d ­ ses. Reallocation across water uses— Finally, reallocation can occur within or between u such as from rural to urban—involves more substantial risk of distributional conflicts than use. The reduction of water in the area of origin involves indirect economic within a given ­ and social impacts, which are less prevalent when water stays within the same group of ­ users. The capacity to address the increasing complexity and conflict associated with longer duration and distances and the involvement of multiple uses is a function of the infrastruc- ­ nvolved. ture, water accounting, water rights, and wider institutional arrangements i A political economy perspective provides criteria for examining reallocation based on efficiency. Water reallocation involves questions of justice, effectiveness, equity, and ­ 10 Dividing the Water, Sharing the Benefits generated. It also affects current and fairness, and the distribution of resources and benefits ­ ­fficiency. Economic efficiency implies the allocation of scarce long-term economic e resources to their highest valued economic uses, defined as a situation that maximizes the uses. Equity and fairness are difficult to marginal productivity of water across its competing ­ define and measure by comparison; these terms are aspirational, contextual, and c ­ ontingent. Further, there are trade-offs between efficiency and equity, as when efficiency dictates the reallocation of water from agriculture to cities, but equity merits consideration of the amount and pace of reallocation, and the compensation and benefit sharing arrangements in ­place. We focus on the drivers, process, and outcomes of reallocation from rural to urban regions according to these criteria, exploring whether the goals of the reallocation are met as well as ­nvolved. This requires an understanding of costs and benefits of the winners and losers i ­ ities. reallocation, as well as their distribution between and within rural regions and c Dividing the Water, Sharing the Benefits 11 Methodology ­ ities. The This report synthesizes the global evidence on reallocation from rural regions to c methodological approach includes two steps. The first step involves a systematic review of the global experience with rural-to-urban water reallocation (map 3) (Garrick et al. 2019).1,2 The review examines the supply and demand drivers of reallocation, the reallocation pro- cess, and impacts. Here we emphasize three of the key findings that shaped this report’s methodology. First, reallocation is driven by the interaction of supply and demand drivers. Urban population growth and economic development leads to increases in water demand in a context of limited water supplies, climate variability, or water quality problems. Second, collective negotiation and administrative decisions are the most prevalent form of realloca- tion decision making. The administrative agreements that govern the negotiation and imple- mentation of reallocation projects often involve the construction of new infrastructure and include provisions for compensation. Finally, the evidence on the costs and benefits of these projects and their distribution is sparse, and virtually uncharted for informal reallocation processes (e.g., via informal markets or theft). In this context, there is increasing need to assess the evolution of water reallocation projects and their political economy. The second step involved the development of in-depth case studies, employing a political economy framework to assess the drivers, processes, and outcomes of water reallocation MAP 3. Rural-to-Urban Water Reallocation, following Garrick et al. (2019) Monterrey São Paulo Mokopane Melbourne Global review Case studies IBRD 43703 | JUNE 2019 This map was produced by the Cartography Unit of the World Bank Group. The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of the World Bank Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. 12 Dividing the Water, Sharing the Benefits TABLE 1. Selected Attributes of the Water Reallocation Case Studies Melbourne, Australia Mokopane, South Africa Monterrey, Mexico São Paulo, Brazil Year (agreements) 2007, 2010 1993 1989, 1995, 1996 1974, 2004, 2017 Primary Goals • Irrigation modernization • Municipal water supplies • Municipal water supplies • Municipal water supplies • Municipal water supplies • Farmer debt relief • Flood control • Environmental water recovery Distance (km) 70 ~25 108 80 Volume (m3/y) Up to 75 million Up to 2.4 million Up to 157 million Up to 900 million Capital costa $784 Pipeline onlyb ­n.a. $893–$1,236 $516 Phase 1 only (millions, US$ 2019) Financing Melbourne Water (state-owned National government Nuevo Leon state Various sources water utility) government, Monterrey Water Authority, national government; IDB Donor region Goulburn Valley, Murray River Sterk River Irrigation Irrigation District 026, PCJ rivers (Goulburn-Murray Irrigation District) Scheme Lower San Juan Donor characteristics 258,000 ha of irrigation 1,600 ha of irrigation 70,000–100,000 ha of PCJ headwaters irrigation (~5,000 farms) Sources: Aguilar-Barajas and Garrick 2019; de Souza Leão and De Stefano in press; O’Donnell, Garrick, and Horne 2019. n.a. = not applicable; PCJ = Piracicaba, Capivari, and ­ Note: IDB = Inter-American Development Bank; ­ Jundiaí. a All dollar figures in this report are in 2019 US dollars (US$). Foreign currency values were first converted to contemporary US$ based on the annual average currency exchange rate during the year in which the costs were incurred and were then adjusted for inflation using the US Consumer Price Index to obtain the 2019 US$. b Capital investments in irrigation efficiency to generate the water “savings” cost an additional US$972.5 million and were funded by the Victoria state government. projects from a political economy perspective. four administrative reallocation projects from rural regions to cities (table 1): Melbourne, Australia; Mokopane, South Africa; Monterrey, ­ razil. The case studies are examples in which the national or state Mexico; and São Paulo, B (e.g., inter- governments have reallocated large volumes of water from one place to another ­ basin or interregion ­ transfers). Case study selection criteria include the following: ­ ransactions. • A focus on administrative reallocation projects, rather than market-based t • Cases that have not received extensive coverage in academic literature to date, or have completed. only received coverage when first negotiated and ­ • Data availability for in-depth analysis, including primary data on political, social, environ- mental, and economic drivers and impacts of the project; technical information regarding the change in water balance; and information on the costs and benefits of the project and their distribution. • Participation of local partners and experts, including personnel with institutional mem- ory of the project, such as government officials and key informants from the donor and recipient ­regions. Three of the four case studies are from countries with reallocation at key stages in the cities. Whereas the Monterrey and São Paolo cases involve development path of growing ­ Dividing the Water, Sharing the Benefits 13 large cities, the case of Mokopane provides insights from a rapidly urbanizing township that exemplifies the growing challenges faced by towns and small to intermediate-sized cities. The fourth case study from Australia involves the reallocation of water “saved” through irri- gation efficiency; it offers an important example given the growing push to use irrigation cities. efficiency to quench the growing water demands of ­ Drawing on multiple sources of data, the case studies assess the evolution and effectiveness criteria. Data include archival information (agree- of reallocation projects across multiple ­ ments, meeting minutes), interviews with decision makers and stakeholders, and available financial and socioeconomic ­ data. A set of assessment questions explore the political econ- ). omy of reallocation, examining the history, negotiation, and impacts of the project (box 1­ An international team conducted 60 interviews during a two-month period (February and ). Interviewees included key actors from the rural communities (donor or source March 2017­ region), city representatives (recipient region), water resource agencies, and other relevant ­ stakeholders. They included three former state governors, a deputy state government pre- mier, four heads of national water agencies (past and present), directors of major urban water utilities (past and present), as well as irrigation managers and farmers, city water managers, leaders. However, the distributional impacts on water members of civil society, and political ­ ­ nalysis, which was ­limited. reallocation require high resolution and localized socioeconomic a BOX 1. Assessment Questions • Drivers and ­ goals. What drove the reallocation, and what were the goals of the project? ­ • Mechanisms. What processes were used for the reallocation? What alternatives were considered? ­ • Efficacy. Did the reallocation achieve its intended goals? What were the unintended consequences? • Efficiency and ­ equity. What were social, economic, and environmental costs and benefits of reallocation, and how were they distributed within and between the donor and recipient region? How did this change over the short term and long term?­ • Compensation. How were existing users (donor region) and third parties impacted by the project? How was the donor region ­ compensated? • Dynamics. How has the reallocation project affected the future prospects for reallocation in the ­ region? • Lessons. What lessons can be learned about the design, implementation, and rene- gotiation of the project? What are the perceptions of the project across different actors? 14 Dividing the Water, Sharing the Benefits Understanding Reallocation ­ uestions. We distil key ideas and lessons This section summarizes the responses to these q for decision makers facing the challenge of scoping, designing, or implementing a water areas. Both the litera- reallocation agreement to quench the thirst of fast developing urban ­ ture review and the in-depth case studies were guided by five primary questions: • Drivers: What drives reallocation? • Process: How is water reallocated in practice? • Politics: What are the political controversies associated with reallocation? • Outcomes: How can water reallocation deliver more efficient and equitable outcomes? • Lessons: What lessons can be learned? What Drives Reallocation? Reallocation from agriculture to cities is the product of two fundamental drivers: growing ). However, supply-demand imbal- demand and inadequate supply (Marston and Cai 2016­ ances are not sufficient on their own to overcome the political, financial, and technical barriers, which require additional enabling conditions or t ­ ­ riggers. We provide an overview of the different demand and supply drivers and triggers based on the broader literature ). (table 2) and for the four case studies (table 3­ Demand ­drivers. Population growth; the rising economic value of water for urban water uses; and the technical, political, and economic feasibility of increasing urban water use regions. efficiency represent the key variables driving reallocation from rural to urban ­ Demand increases with population growth until urban water use efficiency TABLE 2. Trends and Drivers for Rural-to-Urban Water Reallocation Trends Drivers Changing demand patterns • Exhaustion of local surface and groundwater supplies • Increasing competition between agriculture, peri-urban and urban uses • Changes in the timing, location, and quality of water Changing supply patterns • Impacts of drought on cities and agricultural water availability • Inadequate water supply infrastructure for storage and piped connections • Improved wastewater collection, treatment, and reuse for agriculture • Climate variability and change affecting water reliability Declining water quality • Impacts of agricultural and industrial pollution on urban water supplies • Local contamination due to inadequate wastewater treatment Increasing water use efficiency • Increased irrigation efficiency generates water savings subject to proper accounting of return flows Source: Adapted from OECD 2015. Dividing the Water, Sharing the Benefits 15 TABLE 3. Factors Driving Administrative Reallocation from Agriculture to Urban Water Uses Population growth Intensive use Pollution of local Improved irrigation and economic of local water Prolonged drought Wastewater reuse resources efficiency development resources Melbourne, Y Y n.a. Y n.a. Y Australia Mokopane, Y n.a. n.a. n.a. n.a. n.a. South Africa Monterrey, Y Y n.a. n.a. Y n.a. Mexico São Paulo, Brazil Y Y Y n.a. n.a. n.a. Source: Author elaboration. Note: Y= yes; n.a. = not ­ applicable. enables  decoupling. The marginal economic value of water in cities is often higher than agriculture. Finally, urban water use efficiency provides a potential substitute for realloca- ­ ­ esistance. tion, but can prove costly or encounter political and cultural r A conventional trajectory involves population growth leading to the exhaustion and deg- ­ l. 2013­ radation of local water supplies (Richter et a ). The economic importance of cities and industry creates political will and financial capital for reallocation when local supplies prove ­ inadequate. For example, Monterrey grew from 270,000 in 1950 to almost 3 million by 1990, becoming one of the largest cities in Mexico and a hub of international trade and economic growth. This growth led to an increase of its water consumption, which outstripped water ­ ­ eservoirs. Similar trajectories are occurring across available from local aquifers and nearby r Kenya. the global south, from Hyderabad, India, to Nairobi, ­ reallocation. In São Paulo, pollution Water quality problems in the city may also lead to ­ due to heavy industrial development and poor wastewater collection and treatment has degraded local surface and groundwater resources. In some regions where water stress is already high, population growth has decoupled from water demand through increasing ­ eeds. In urban water use efficiency, but this may not be enough to meet urban water n Albuquerque, New Mexico, severe water imbalances and plummeting groundwater levels have led to major water efficiency efforts since 1995. The city has achieved a reduction of half of the daily per capita water consumption in 20 years: from 251 to 127 gallons per capita day. Despite these efforts, Albuquerque has needed to pursue new surface water per ­ resources, both from agricultural farmers in the surrounding counties and through an inter- transfer. basin water ­ Supply ­drivers. Legal restrictions on new supplies, infrastructure development, and regions. Legal droughts are key supply variables driving reallocation from rural to urban ­ and technical constraints on new infrastructure development have provided a primary driver for reallocation, when new demands depend on reallocation from existing water ­ ources. However, infrastructure development supplies rather than the importation of new s or modernization can enable large-scale reallocation projects to move water from historic uses. to contemporary ­ 16 Dividing the Water, Sharing the Benefits Droughts are critical factors driving reallocation. Even in cities with sufficient water sup- plies to meet demands in average years, temporary deficits associated with droughts can spur reallocation to meet chronic and acute shortages ­alike. In Australia, the reallocation to ­ rought. In Melbourne was accelerated by the growing threats of water shortages due to d Taiwan, rainfall deficits from 2002 to 2006 led to reductions in paddy growing, following provisions for emergency transfers to support urban demand under the 1942 Taiwan Water ­ Law. In Spain, prolonged drought has led to administrative decisions to temporary reallo- ­ reas. Other regions in the Western United States have cate irrigation water to urban a entered into interruptible contracts, or dry-year options, in which irrigation districts fallow cities. lands during droughts to buffer water supplies for ­ improvements. Investments in irrigation Reallocation is linked with irrigation efficiency ­ efficiency can create opportunities for reallocation when agricultural regions reduce water ­consumption. Although improved irrigation efficiency offers a potential source of water sup- ply for growing cities, it also raises challenges for water accounting to document and verify water savings and understand the implications for downstream water users and the environment (Wester et al. 2008). If designed and monitored carefully, irrigation districts ­ and urban areas can establish a win-win situation, by which cities can benefit from water savings due irrigation modernization and farmers can obtain financing from cities to con- systems. In this regard, irrigation efficiency tribute to the modernization of their irrigation ­ ­ river. In the first instance, irrigation effi- can act as both a response to reallocation and a d ciency increase can be a strategy by the donor region to manage reductions in water avail- ­ eallocation. In the second instance, the need for capital investment ability associated with r in irrigation modernization schemes may drive reallocation when agricultural regions financing. In northern Victoria, Melbourne and the state gov- release water in exchange for ­ ernment invested in a US$972.5 million irrigation modernization program to address short- age risks associated with the Millennium drought and expand the supply portfolio for the growth. Similar programs have been developed in Southern California and city’s long-term ­ Japan. For example, in Japan, the Saitama Prefecture paid for the rehabilitation of irrigation ­ infrastructure and reallocated over 7 cubic meters per second of saved water for domestic al. 2007­ use over a series of incremental projects from 1968 to 2003 (Matsuno et ­ ). Such modernization as a form of ­ approaches involve irrigation ­ compensation. Finally, the collection, treatment, and reuse of urban wastewater present another growing reallocation. Improved sanitation infrastructure provides cities with large opportunity for ­ volumes of water that can be offered to irrigators in exchange for water from conventional sources. This involves reallocation of high-quality water from agricultural uses to cit- water ­ ies in exchange for treated wastewater that meets quality requirements for irrigated agriculture. In Monterrey, for example, the development of the El Cuchillo agreement came ­ with a subsequent water sharing agreement committing the city to deliver 6 cubic meters district. per second of treated wastewater to the downstream irrigation ­ Dividing the Water, Sharing the Benefits 17 External ­Triggers. Many water stressed cities struggle to reallocate water from surrounding regions. For instance, in Cape Town, South Africa, reallocation from agriculture agricultural ­ has been politically infeasible despite being routinely identified as an important manage- security. This illustrates that the economic and tech- ment option for the city’s urban water ­ own. Reallocation depends on factors or levers nical drivers are not sufficient on their ­ coming together to overcome political resistance, economic crises, financing challenges, problems. and technical ­ Even when water demand for cities outstrips available supplies, reallocation may not ­ ction. These triggers can take multiple forms occur until triggers create a window for a scales. Political windows may arise during and operate at a range of spatial and temporal ­ region. changes in government and the emergence of new leadership from the affected ­ For example, approval and implementation of the Telugu project to deliver water from the Krishna River Basin to Chennai, India, became possible because of a temporary polit- ical alignment between the Union government (national government) and state govern- project. Across the case ments enabled by the electoral gains brought about by the ­ ­eallocation. studies, political transitions created opportunities and constraints for r ­ lection of President Carlos Salinas de Gortari brought new In Monterrey, the national e studies. The impetus to the El Cuchillo project nearly 70 years after the first planning ­ president made modernization a plank of his presidency and had family from the region, which made the El Cuchillo project an attractive target due to its wider economic and strategic ­significance. Financing represents another important trigger or constraint on reallocation, particularly when capital investments in infrastructure are required for interbasin transfers or irrigation modernisation. Australia presents a clear example of financing enabling rural to urban real- ­ ). The existence of a budget surplus and location from Northern Victoria to Melbourne (box 2­ willingness of the state government to invest in irrigation water savings helped to tip the balance for the project; moreover, the need for external investment to modernize irrigation made the agricultural community and leaders initially more amenable to reallocation in exchange for ­financing. Conversely, the lack of finance can hinder projects even when the addressed. In benefits far exceed the costs and political and technical challenges have been ­ Kathmandu, the US$689 million Melamchi Water Supply project can be traced to the 1973 al. 2017­ Water Master Plan (Gurung et ­ ). An infusion of funding from the Asian Development ­ roject. As a condition for the funding, the Bank (ADB) in 2000 spurred development of the p ADB mandated institutional reforms to include private operations and other measures; restructuring. these proved challenging and led to ­ How Is Water Reallocated in Practice? ­ ecades. The case studies illustrate that reallocation is a process that can span multiple d Although reallocation is shaped by local history and hydrology, administrative approaches ­ ontexts. The global experiences to reallocation involve broad similarities across diverse c 18 Dividing the Water, Sharing the Benefits BOX 2. Reallocation through North to South Pipeline in Melbourne The North-South (or Sugarloaf) Pipeline was completed in 2010 to deliver up to 75 cubic megameters per year to Melbourne through a interbasin transfer from the River. The US$729 million, 70-kilometer Goulburn River, a tributary of the Murray ­ pipeline was announced in June 2007 during the depths of the 13-year Millennium ). Reservoir storage for Melbourne’s 4 million residents dipped drought (1997–2009­ below 30 percent in May 2007, prompting a US$4.8 billion plan to address Victoria’s security. Even though Northern Victoria and the Murray-Darling Basin were water ­ also experiencing severe drought, the Sugarloaf Pipeline was considered the ­ lowest-cost source of new water for Melbourne and the most rapid of the options available, to be further supplemented in the longer term by a proposed desalina- plant. The water from the Goulburn River would be generated through water tion ­ savings from a US$972.5 million irrigation infrastructure upgrade (the Food Bowl Modernization scheme) championed by a prominent group of agricultural and busi- region. ness leaders from the donor ­ More than 40 years after former Victorian premier Sir Henry Bolte proclaimed that not a single drop of water would leave the Murray-Darling Basin for Melbourne, a opened. A unique convergence of four factors drove once in a 100-year opportunity ­ the project: (i) the threat drought posed to Melbourne’s water security; (ii) the l ­ onger term needs of a growing city; (iii) the push for irrigation modernization—and the associated need for financing—to drive economic growth in Northern Victoria; and environment. The drought “insur- (iv) growing commitments to secure water for the ­ ance” offered by the project for Melbourne’s water supply was a critical justification and helped to tip the scales for its rapid approval, despite the legacy of r ­ esistance. The Sugarloaf Pipeline was complete, and delivering small volumes of water to Melbourne, in late 2010. However, in November 2010, a new coalition government was elected in Victoria, bringing a new water minister, Peter Walsh, who opposed the opposition. The incoming government project and was elected in part due to his vocal ­ ­ ipeline. immediately halted deliveries of water to Melbourne from the Sugarloaf P Since late 2011, new rules of operation prevent the use of the Sugarloaf Pipeline to deliver water to Melbourne unless the city’s water reservoirs drop below 30 percent by November 30. This is an unprecedented threshold for Melbourne (even during the Millennium drought, storages remained above 30 percent by November 30), and means that the Sugarloaf Pipeline will rarely (if ever) be used under current r ­ ules. The Sugarloaf Pipeline is an example of rural-to-urban reallocation in the con- text of irrigation modernization, in which the savings generated by irrigation effi- ciency provided the water supply for urban, additional agriculture, and environmental restoration ­ ­ purposes. The reallocation experience in Melbourne and the Goulburn box continues next page Dividing the Water, Sharing the Benefits 19 BOX 2. continued region illustrates the political and technical challenges of reallocation using water generated through irrigation efficiency projects due to the water accounting needed to ­ ­ savings. Other important lessons include drought’s status as a document and validate ­ double-edged sword: it can facilitate investments in infrastructure projects that would be ­ difficult under normal circumstances, but it can also accelerate projects with limited time ­ ­ echnical study, engagement, and political negotiation needed to sustain the ­ for the t reform. Source: O’Donnell, Garrick, and Horne 2019 with administrative approaches to rural-to-urban reallocation illustrate the existence of four phases—drivers, diagnosis, design, and delivery—although these are often hard to dis- ­ rocess. The supply and demand drivers create tinguish in practice and rarely part of a linear p pressure for reallocation as cities grow, and can be triggered by the political and financial previously. When cities pursue more water, reallocation from rural regions factors noted ­ ­ e.g., groundwater devel- constitutes only one option among a range of other infrastructure ( opment, desalination) and demand management approaches, and is assessed in relation to these alternatives as a complement or ­ substitute. Even when reallocation presents the most efficient option, design and delivery challenges often arise due to technical and political previously. Changes in patterns of water allocation and use involve economy issues noted ­ unintended consequences that create the need for adjustment, and, sometimes, renegotiation. Reallocation projects involve an assessment (diagnosis) to evaluate realloca- ­ tion against other alternatives and to design the project based on technical, economic efficiency, and equity ­ ­ considerations. Diagnosis involves technical studies and stakeholder consultations, which can vary tre- mendously in their rigor, depending on capacity for water accounting and budgeting, stake- studies. A water holder mapping, communication and engagement efforts, and feasibility ­ balance and water accounting can track the supply and demand from a system, and the boundaries of the system become a crucial determination, as illustrated by the expansion of São Paulo Metropolitan Region water system to include the (Piracicaba, Capivari, and basins. The system boundaries can guide mapping of the relevant stakeholders Jundiaí) PCJ ­ from the source and destination regions, offering a baseline and benchmark for assessing the reallocation. In Mokopane, farm owners were included in the mapping of stake- impacts of ­ ­ ot. Effective communication and engagement with affected holders, but farm workers were n stakeholders become important at this phase, requiring strategies to reduce or preempt process. backlash later in the ­ Design determines whether new or modernized infrastructure is needed for r ­ eallocation. Design is institutional and involves the development of agreements regarding financing 20 Dividing the Water, Sharing the Benefits compensation. The design phase of administrative reallo- arrangements, water sharing, and ­ cation often involves extensive negotiation in all these areas, which builds on the stake- scoping. These processes holder mapping and engagement addressed during diagnosis and ­ will involve different agencies or stakeholders based on the institutional framework govern- ing allocation and reallocation, ranging from highly centralized and formal processes to arrangements. The design of administrative agreements—water permits, opera- informal ­ ­ eallocation. tional rules and related instruments—proves critical for effective and equitable r In this regard, reallocation agreements bear striking resemblance to international trans- boundary water agreements, which establish rules, rights, and responsibilities and ensure resolution. For example, the Monterrey project provisions for adjustment and conflict ­ involved three agreements, including an operational agreement in 1996 with detailed crite- ria and conditions for water sharing, compensation, and coordination by a river basin council. The political feasibility and socioeconomic impacts of the project are heavily influ- ­ enced by compensation agreements, which take diverse forms, sources, amounts, and recip- ­ ouseholds. ients based on the nature of the impacts and focus on poor h ­ esign. The city The experience in Hyderabad, illustrates the importance of agreement d draws water from the Krishna and Manjira rivers with a sequence of government orders used each. Although politically contentious in both to enact administrative reallocation in ­ instances, reallocation from the Krishna occurred without compensation and has met with resistance. Conversely, an intersectoral water sharing agreement for the Manjira sustained ­ transfer defined water volumes to support irrigation projects when sufficient water is reserved to meet Hyderabad’s allocation, mitigating impacts on the agricultural region ). (Celio, Scott, and Giordano 2010­ Delivery involves the construction and operation of the agreement, as well as the monitor- ­ mpacts. The unintended consequences of reallocation and the lessons ing and evaluation of i renegotiation. from operating experience provide the basis for adjustment and may lead to ­ Periodic renewal processes can build predictable openings for adjustment, as experienced with São Paulo in which the initial water permit in 1974 involved a 30-year term, followed by 10-year agreements in 2004 and 2017. These renegotiation opportunities allow adjustments to shifting drivers and the diagnosis and response to new risks and opportunities, particu- ­ rows. Although the temporary nature of such agreements affords larly as the donor region g several advantages and may make the parties more likely to agree to a reallocation, droughts or other social, economic, and environmental changes can make renegotiation more diffi- ). cult, or delayed, as occurred in 2014 due to the severe drought affecting São Paulo (box 3­ Delivery requires appropriate tools for evaluating and tracking impacts, including capacity ­eallocation. For to disaggregate the impacts to understand the distributional impacts of r example, socioeconomic indicators, agricultural productivity, water use efficiency, and water accounting all prove critical for understanding and addressing the negative impacts of ­reallocation. Dividing the Water, Sharing the Benefits 21 BOX 3. Cycles of Reallocation Amid Change in São Paulo In 1974 the Cantareira system, a network of reservoirs located in the inland region of the State of São Paulo, started operations to reallocate water from the rural headwa- (SPMR). Currently, up to ters of the PCJ basins to the São Paulo Metropolitan Region ­ 900 million cubic meters per year of water travel over 80 kilometers to provide drink- ing water to about 8 million people in the SPMR, the most important financial and eco- people. Since 1974, the operating nomic center in Brazil and home to over 21 million ­ rules of the Cantareira system have been revised and renegotiated twice, once in 2004 and once in 2017, leading to the revision of the water allocation shares between the ­ eallocation. donor and the recipient regions, as well as the conditions regulating the r The case study demonstrates that reallocation is dynamic and needs to be revis- time. The project was originally designed and developed over 50 years ited over ­ ago under a military regime to bring water from a sparsely populated, rural region region. Sustained urban and industrial growth to a densely inhabited, fast-growing ­ in the donor region, the democratization of Brazil, and the implementation of a new water management model have created a radically new context that required stake- holders to revisit the original goals and rules of the ­ reallocation. Since the start of operations, the Cantareira project has been effective in achieving its ­ opulation. main goal of meeting the water needs of an important share of the SPMR p Over time, however, it became evident that the reallocation could not be managed unilaterally by the SPMR utility company (Sabesp), but had to consider the needs region. Recurrent floods and droughts showed the weaknesses of the of the donor ­ reallocation agreement and the need for more flexible solutions to increase water regions. The expiration of the water reallocation permit in 2004 security in both ­ and in 2014 allowed the government to revise the reallocation rules to increase its ability to deal with extreme events and to slightly increase the share of benefits of region. In both renewals, droughts catalyzed the Cantareira system with the donor ­ ­ esign. change and adaptation, as they revealed the weaknesses of the reallocation d The 2003–04 and 2014 droughts pushed decision makers to look for more sophisti- ­ esign. cated and flexible solutions in the revision of the reallocation d During negotiations, strong technical advice is key, but politics and perceptions also matter. The evolution of the water permit over time shows that the rules for the alloca- ­ tion have become increasingly more technical and complex to improve the resilience of system. In the design of both permits, in 2004 and 2017, negotiators the water supply ­ were sophisticated in defining operating rules, terms, and conditions ( ­ e.g., risk aversion curves, water banks, and new location of the allocation control p ­ oints). Reallocation design was rooted in sound technical studies, but the negotiation was also inevitably a issue. In particular, the political constellation at federal, state, and munic- very political ­ ipal levels highly influenced the negotiations, as did the electoral c ­ alendar. Source: de Souza Leão and De Stefano, in press. 22 Dividing the Water, Sharing the Benefits Navigating the Controversies of Reallocation ­ rrangements. Administrative reallocation occurs across a range of contexts and institutional a Regardless of the differences, some common issues contribute to misperceptions and tensions. The literature and case studies reveal challenges associated with diverging per- ­ spectives about the ownership of water, informal water users, water use efficiency, monitor- compensation. ing, and ­ Controversies over water rights affect water reallocation by creating uncertainty about ­ ater. Uncertainty can stem from legal disputes associated with who owns and should control w water licencing and the rules and priorities governing water allocation between cities and agriculture. Both cities and agricultural regions cite such controversies to promote realloca- ­ respectively. Cities may claim water that “never really belonged” to agricul- tion or resist it, ­ claims. tural regions, while agricultural regions often claim the priority of their rights and ­ In Monterrey, for example, downstream farmers asserted strong claims regarding their own- ership over water, invoking a 1952 Presidential Accord that had been superseded by national licences. The Accord remains a basis for water laws leading to the regularization of water ­ validity. In São Paulo’s case, PCJ stakeholders complaints by irrigators, regardless of its legal ­ stress that water transferred to the capital originates in their headwaters, and, therefore, used. they should have a say about how it is shared and ­ Informal water use is prevalent in many agricultural regions, and it can hinder efforts to ­ echanisms. Agricultural regions may reallocate water or implement agreed benefit sharing m lobby against reallocation projects that threaten informal water users or bring added scru- tiny and metering that could lead to tighter regulation e ­ nforcement. Informal, unlicensed, or illegal water uses can complicate the implementation of agreements when informal water users intercept water intended for a recipient region, or capture the return flows from the agriculture. This was the case in Monterrey, in which treated city meant for downstream ­ wastewater was intercepted by informal water users before reaching the downstream irriga- compensation. tion district, requiring ­ Water use efficiency has emerged as a common source of disputes in reallocation proj- efficiency. ects, as both sides—rural and urban—claim the other can increase water use ­ This can cause “finger pointing” regarding how efficiently the donor and the recipient regions manage their ­water. In São Paulo, Monterrey, and Mokopane, rural stakeholders have complained about low water use efficiency in the recipient region and have argued reallocation. In São that water conservation could delay or even preempt the need for ­ high. As the Paulo Metropolitan Region, water losses from distribution networks are ­ donor region developed and stakeholders asserted claims on the water that been trans- ferred to São Paulo, new commitments were negotiated to improve water use efficiency Paulo. The 2004 water permit required Sabesp to maintain permanent programs in São ­ for leakage control, rational water use, water waste reduction, and incentives for water reuse. Conversely, cities increasingly see irrigation water use efficiency as a source of ­ new water, illustrated by experiences across Australia, the Western United States, Dividing the Water, Sharing the Benefits 23 ­ hina. The Melbourne case study illustrates how the perception of water Japan, and C powerful stimulus for urban investment in irrigation modernization, savings can be a ­ droughts. particularly in the context of ­ Disputes related to water use efficiency are linked to challenges with monitoring, which reallocation. The lack can breed mistrust regarding water use, deliveries, and impacts of ­ of measurements and monitoring of changing water availability and use across the donor and recipient region can lead to claims from both sides that the reallocation project is implemented. In Monterrey and Melbourne, disputes regarding either unfair or poorly ­ monitoring have led to doubts about the validity of the compensation mechanisms (Monterrey) and the water savings ­(Melbourne). Monterrey lacked an effective mechanism for detecting the quantities of treated wastewater arriving at the downstream irrigation tributary. In district and separating these flows from other water sources conveyed by the ­ Melbourne, irrigation modernization has raised technical accounting challenges associ- ated with verifying water savings and addressing the negative impacts on those farmers systems. who lost their connections to irrigation ­ The perceptions of fairness hinge not only on the issues of ownership, water use efficiency, ­ ompensation. In Monterrey, it was not suffi- and measurement but also on the source of the c cient for the farmers to receive compensation: they also believed that the city, rather than the national government, should pay for ­ it. Melbourne experienced a different challenge: the city paid for irrigation modernization but the irrigation community split its support ­ thers. Some irrigators also rejected out- when some irrigation regions benefited more than o right the idea of an exchange, arguing that the state government should fund irrigation mod- ­ eturn. ernization without requiring any water transfer to Melbourne in r The diverging perceptions and lingering political tensions have underscored the need for ­ rocesses. creative approaches to benefit sharing, supported by transparent and inclusive p ­ utcomes. They also support the need to reconcile perceptions of reallocation with the actual o Doing so requires measuring and understanding the impacts on the donor and recipient term. regions over the short and long ­ Can Water Reallocation Deliver Efficient and Equitable Outcomes? The concept of benefit sharing offers a useful way to think about and evaluate reallocation ­ projects. Although the term has typically been used in the context of transboundary rivers, regions. Benefit sharing requires it also applies to water sharing between rural and urban ­ accounting for the full range of benefits and costs associated with reallocation of water, not ­ se. This creates opportunities to improve outcomes for only the changing patterns of water u ­ egion. both the donor and recipient region, or at least minimize the costs borne by the donor r Reallocation generates a stream of multiple types of benefits and costs, which are distrib- region. The outcomes of reallocation pro- uted between and within the city and the donor ­ ­ istribution. First, an cesses can be assessed according to the benefits, costs, and their d 24 Dividing the Water, Sharing the Benefits ­ osts. The project’s benefits efficient reallocation project requires that benefits exceed the c will enhance the economic productivity of the water for the city, which can generate positive spillovers because urban water security fosters economic development, trade, and ­ industry. However, benefits may also accrue to the donor region additional investment by ­ water. For example, when reallocation requires increased river flow despite the loss of ­ regulation, the reallocation project may provide flood control and regulation benefits, as ­ Paulo. The city or financiers will have financial con- occurred in both Monterrey and São ­ struction costs and opportunity costs of the capital investment, which could have been ded- icated to alternative water supply and demand management options or investments outside altogether. The alternative water supplies for the city may involve lower the water sector ­ (e.g., water use efficiency) but carry other risks or political or technical c costs ­ ­ oncerns. For rural regions, the project costs often exceed the ­benefits. They include the opportunity costs of the water and lost economic opportunities, such as the reduction in irrigated agri- culture, which may be partially offset by the productivity improvements associated with deliveries. The global review and the in-depth upstream regulation and more reliable water ­ cases illustrate opportunity costs associated with lost agriculture, as well as the countervail- ­ upplies. These ing productivity improvements as farmers attempt to adjust to limited water s productivity gains were achieved in Monterrey and north of Melbourne, but the improve- ments were uneven, illustrating the distributional impacts and implications, over time and ­ space. The social and economic costs also extend beyond the direct negative impacts on water users through “third-party effects,” which are indirect negative impacts, or ripple effects, of lost agricultural and rural economic activity felt by associated industries and labour. Third-party effects are often ignored or underestimated in the design and delivery of ­ projects. They led to concerns about the viability and stability of agricultural production in ­ Monterrey. the donor regions in Mokopane and ­ Benefit sharing can make reallocation fairer by accounting for the direct and indirect costs ). The equity impacts of reallocation depend on the type, borne by rural regions (figure 2­ ­egions. amount, and form of compensation to offset the costs concentrated in the rural r Conventional approaches to compensation consider the opportunity costs faced by the farmers and design monetary payments based on the lost acreage, productivity, and com- modity ­ values. These approaches address the direct costs borne by individuals but ignore the wider, indirect costs associated with the third-party effects noted previously, such as lost employment and industries or other longer term consequences on society and the environment. Conventional approaches encounter several practical and political difficulties ­ associated with monitoring changes in acreage and water application, estimation of produc- ­ alues. A shift from narrow tivity rates, and assumptions about crop choices and commodity v compensation to benefit sharing establishes a premium over the opportunity costs and con- siders the system-level costs—not only the impacts felt by the individuals—to ensure a broader portfolio of benefits that may include regional infrastructure development, alterna- ­ evelopment. Benefit sharing may allow tive water supplies, and investment in community d Dividing the Water, Sharing the Benefits 25 FIGURE 2. Moving from Compensation to Benefit Sharing Reallocation Reallocation Reallocation (before compensation) (with compensation) (bene t sharing) Increased water security Investments in rural well-being Compensation to farmers Positive side e ects Bene ts Farmers’ Costs opportunity Reallocation costs direct costs Third-party e ects City City City Donor Donor Donor region region region both urban and rural regions to gain from the project, or at least sustain rural well-being by project. The outlook of a reallocation project changes with offsetting the wider costs of the ­ time, because the costs are localized and immediate, while benefits are delayed and more diffuse. Thus, benefit sharing also takes into account mid- and long-term effects of ­ ­reallocation. This shift from compensation to benefit sharing involves three main implications: the need to (i) account for the full range of costs and benefits over space and time; (ii) understand the distributional issues, particularly the consequences for poor populations (see example ­ ndividual. The full range of costs and ben- in box 4), and (iii) compensate the region, not the i efits includes the direct and short-term costs and benefits, including the opportunity costs  for the donor region—the forgone agricultural production and future economic development—and the benefits accruing to the c ­ ­ ity. The indirect and longer term costs and benefits should also be considered, accounting for the regional impact of the reallocation for the donor region and the ancillary benefits generated by the project for the regional economy that can ripple out and reach the rural region over the longer ­term. Designing and ­ onsiderations. implementing benefit sharing strategies requires the following four main c sharing. The global experience suggests three primary forms of Forms or types of benefit ­ compensation for the donor region: (i) monetary payments based on opportunity costs (compensation in its narrow sense); (ii) infrastructure modernization and other material ­ upplies. Table 4 summarizes bene- investments; and (iii) the provision of alternative water s fit sharing in the case studies, illustrating the wide range of options and the movement from narrow compensation based on opportunity costs of agriculture in Mokopane to the broader Monterrey. portfolio of measures adopted in ­ 26 Dividing the Water, Sharing the Benefits BOX 4. Narrow Compensation Approach in Mokopane In the latter part of the 20th century, the town of Potgietersrus and its black town- ­ apidly. The 1991 ship Mahwelereng, in the Limpopo province of South Africa, grew r Census indicated a growth rate of 4.5 percent, and the 1996 Census indicated a ­ population of 281,285. Increasing water demands from the urban area resulted in various options being considered for meeting the new demand, including the transfer ­ River. In 1993, however, in the face of a severe drought, the of water from the Vaal ­ state bought water rights from the adjacent irrigation scheme to provide water for town. The reallocation of water occurred from the Sterk River Irrigation Scheme the ­ Mokopane). This was not, (SRIS), served by the Doorndraai Dam to Potgietersrus (now ­ however, a market-based purchase of water rights, but an administrative process, driven by the state, in which struggling farmers were offered a set price (US$3,752 per hectare, 2019) for their water rights, in a context in which they had few other debt. It was a process with a dual aim: farmer relief and options to get out of deep ­ use. Many farmers on the irrigation scheme, water transfer from agriculture to urban ­ established in the 1950s, were struggling due to lack of water and poor soils, and saw the buyout as a possible solution, although in many cases the payout was not suffi- financially. cient to rescue them ­ This occurred in the final days of apartheid, just prior to the first democratic elections of April 1994. It also took place in the context of a severe drought in 1992–93. The project provided water to the municipality to meet growing water needs, and served drought. However, the project was as a relatively quick solution in the face of a major ­ part of a context in which poor water use regulation had already impacted the down- stream irrigation farmers, with significant negative financial and economic impacts area. The approach did not resolve the issues in the catchment arising from in the ­ poor water regulation, and interviews with affected farmers revealed an ongoing bit- terness about the handling of the ­ project. The experience in Mokopane suggests lessons relevant for rapidly growing towns in which a narrow, unilateral approach to compensation can engender lingering resent- community. First, in areas of high cli- ment by overlooking the broader needs of the ­ matic variability, longer term scenarios of water availability for the sectors must be considered. This approach is particularly important to prevent expedient decision ­ making that may have unintended long-term implications and leave limited time for negotiation. Second, the broader impacts on the remaining farm- engagement and ­ ers need to be considered if only a portion of the water is taken from an irrigation scheme to avoid placing the full burden of operations and maintenance on a smaller infrastructure. Finally, the group, leading to the steady degradation of irrigation ­ narrow and rushed approach to compensation meant that farm owners did not have box continues next page Dividing the Water, Sharing the Benefits 27 BOX 4. continued an opportunity to negotiate or even to examine or question the methodology used by the price. Although farmers were compensated for the loss of government to calculate the ­ offered. The wider commu- their water rights, many were not satisfied with the amount ­ ignored. Farm workers do not appear to have been factored into the nity was also largely ­ compensation. As a result of these steps, reallocation is seen as a politically infeasible in ­ ­ frica. the South African context, with a chilling influence across Sub-Saharan A TABLE 4. Benefit Sharing: Form, Amount, Recipients, and Financing Case Types Benefits for donor region Recipient Financing Mokopane, South Financial Transfer payments to farmers 29 irrigation Department of Africa US$3,752 per ha, 2019 farmers on 30 Water Affairs portions of land São Paulo, Brazil River flow regulation Flood control downstream PCJ Basins Sabesp Share of regulated flow Monterrey, Mexico Financial Transfer payments to farmers Farmers in Monterrey Water Alternative water 6 m /s of treated wastewater 3 Irrigation Authority, Nuevo supplies Districts 026 and Leon state Share of regulated flow and 031 government, Infrastructure flood protection and federal (irrigation government modernisation and compensation dam) Melbourne, Infrastructure US$972.5 million irrigation Irrigators in Victoria Australia (irrigation infrastructure upgrade for Goulburn River government and modernization) Stage 1 Valley Melbourne Water (state-owned water utility) Jundiaí. Note: PCJ = Piracicaba, Capivari, and ­ Recipients. Recipients range from water users in the donor region to the associated ­ systems. The São Paulo experience districts and wider region through infrastructure ­ suggests that benefit sharing must account for future urban growth and the associated ­ users. The Monterrey case shows that compensations only to farmers ignore the water ­ damages to the economy associated with the production ­(e.g., irrigation material suppliers, seed providers, fuel suppliers, and seasonal harvest labor) and the postproduction (e.g.,  produce distribution, canning industry) p ­ ­ rocesses. As a result, there is increasing focus on regional development and rural livelihoods in benefit sharing arrangements in which water districts and rural communities receive infrastructure, training, and other economies. ­ means of transitioning agricultural and rural ­ 28 Dividing the Water, Sharing the Benefits Financing. This is inextricably linked with the amount of compensation: the question of ­ ays. The how much to pay in the absence of a market is bound up with the question of who p options range from the direct beneficiary—the city and its ratepayers—or the wider national financing. The case studies bring evidence of governments, international donors, or private ­ blended sources, as illustrated by the experience of Monterrey (box 5­ ). Amount. This represents one of the most contentious issues, because it requires sharing the benefits in a manner deemed fair by the donor region and acceptable by the recipient ­ ntities. The starting point for deter- region and the associated financial and government e mining the amount of compensation has been the opportunity costs of current water use: that is, the forgone agricultural and economic productivity associated with the loss or reduc- region. This is not always a straightforward matter in tion of water availability in the donor ­ right. It involves contestable assumptions about the levels of historic water use, the its own ­ ­ rices. This can also cropping patterns and productivity of farmlands, and the commodity p miss the labor-intensive agricultural economy in many low-income contexts and the eco- nomic activity associated with the distribution and processing of agricultural p ­ roducts. BOX 5. Portfolio Approach to Compensate Region, not Individual, in Monterrey Monterrey is a rapidly growing semi-arid city in Northern Mexico with a territory over 6,000 square kilometers and the third largest metropolitan area in Mexico after the Guadalajara. Population has increased from 300,000 in 1950 to over Mexico Valley and ­ today. The city is strategically important due to its economic significance and 4 million ­ ­ roduct 176 percent role as a hub of international trade, with per capita gross domestic p ). In 1994, former president of Mexico, Carlos Salinas de of the national average (2012­ Gortari, inaugurated El Cuchillo reservoir to transfer water from the Lower San Juan Monterrey. A 108-kilometer aqueduct delivers up to 5 cubic meters per sec- River to ­ ond from El Cuchillo to the Monterrey (MMA), contributing approximately 20 percent metropolis. The US$969 million ­ of the water supply for the ­ project (2019) was financed by the Inter-American Development Bank (IDB) and Mexican government, including sources. Since 1989, three agreements have been developed both state and federal ­ to finance, construct, and operate the El Cuchillo project to benefit Monterrey and compensate the farmers in the downstream irrigation ­ ­ district. El Cuchillo Dam has enhanced urban water security, contributing to an increase of daily water supply from 15 hours per day in 1993 to continuous coverage by 1997, despite a period of intense drought that coincided with its fi ­ lling. Irrigation devel- opment predated the dam, however, requiring a set of agreements to reallocate water from downstream agriculture to the city and to devise compensation mecha- nisms to offset the initial decrease in water supply reliability experienced by agri- communities. Irrigation in District 026 (ID026) has supported as many cultural ­ Dividing the Water, Sharing the Benefits 29 BOX 5. continued as 100,000 hectares of maize and sorghum with 4,800 farms and just under 70,000 hectares in the 2013–14 ­ ­ Census. Benefits from treated wastewater, monetary payments, and a new dam have addressed some of the regional needs, although there are technical and political challenges asso- enforcement. Contrary to expectations that the reallocation ciated with monitoring and ­ of water from the San Juan River to Monterrey would threaten the viability of the down- stream farmers, the project is now seen as integral to the regional water supply for both farmers. The project involved, for the first time in Mexico, an agree- the MMA and the ­ ment between two states and included commitments of treated wastewater for delivery to storage. The project illustrates the multiple flows of water and benefits asso- downstream ­ ciated with water reallocation from agriculture to urban uses, and the need for a regional perspective. and systems ­ In 1996, an operational and benefit sharing agreement formalized requirements for ­ . Gómez (MRG) Monterrey to deliver treated wastewater to the ID026 at the Marte R compensation. These commitments included provisions for the dam as a primary form of ­ following: • Construction of an aqueduct to convey treated water from Monterrey to the irrigation region through the Pesquería River funded by Nuevo León and the federal government (not date). constructed to ­ • Downstream delivery capacity of 8 cubic meters per second to the MRG so a minimum of 6 cubic meters per second (or 189 million cubic meters per year) would reach the reser- voir, which was considered essential for the Tamaulipas government and ID026 to sign the ­accord. • Inclusion of legal water rights in the Pesquería River (used to convey the wastewater from Monterrey to the agricultural region) in a registrar for the users of the San Juan Basin to users. prevent interception by informal ­ ­ ulfilled. The agreement specified compensation if the preceding commitments were not f Every November, a determination of deficits in treated wastewater deliveries is calcu- lated based on data as of October 31. Compensation is based on the opportunity costs for ­ astewater. The farmers by estimating the acreage fallowed due to the deficit in treated w ­ ost. The amount compensation is based on the price of 1 ton of white maize per hectare l ­ ay. Comisión Nacional is paid in two equal instalments, one in February and the other in M del Agua (CONAGUA), the national water agency, determines, based upon the specific con- ditions of each case, the source of compensation with the farmers preferring that the city, pay. the beneficiary, ­ box continues next page 30 Dividing the Water, Sharing the Benefits BOX 5. continued The main lesson from El Cuchillo is that water reallocation needs to be designed and assessed in a regional context using a systems perspective on the costs, benefits, and them. Doing so requires well-designed agreements, including mechanisms for sharing ­ processes for including affected stakeholders; compensation schemes that focus on the region, not only the individual losses; robust and transparent water accounting to inform ­ onflicts. implementation; and appropriate coordination bodies to resolve c Dividing the Water, Sharing the Benefits 31 Key Lessons Learned The cases of reallocation considered in these studies are just a small fraction of a larger inter- national phenomenon that suggests a set of lessons and practical implications for decision makers considering reallocation as part of the water resource management tool ­kit. We argue that cities should be reconceptualized as part of urban-rural systems when we talk about lesson  1­ water allocation (­ ). This has implications for reallocating water in terms of water accounting (what needs to be measured and where) (lesson 2); governance, including who needs to be included and how (lesson 3); institutional mechanisms to facilitate coordination lesson 4); and, finally, to shift from thinking about compensation to and conflict resolution (­ thinking about benefit sharing (lesson 5) in a dynamic context (lessons 6 and 7­ ). 1. Benefit sharing requires a systems perspective. Cities are part of urban-rural systems. Reallocation, therefore, involves multiple economic, social, and physical linkages between rural and urban regions, requiring a systems perspective to understand the changing pat- terns of water use and the associated redistribution of benefits and costs over multiple time scales. A systems perspective can improve reallocation processes in all phases, including the design of compensation mechanisms to share benefits with the donor region. 2. Get the numbers—and the water balance—right. Reallocation is an inherently political ­decision, but should be informed by reliable data, robust accounting, and sound ­modeling. This includes tracking changes in water supply, water use, and water use efficiency; designing alternatives for allocation rules; and understanding and measuring costs and benefits of reallocation and their distribution. A sound understanding of the water bal- ance underpins effective and equitable reallocation, which in turn depends on robust accounting that prioritizes the data, monitoring systems and evaluation processes needed for designing and delivering reallocation projects. 3. Water sharing agreements should be backed by sound governance. Reallocation from rural to urban regions involves a form of water sharing agreement across sectors and some- times political borders. Like other transboundary agreements, rules should guide the ­content, and also the governance processes. Clear rules should define the volume of water transferred, the timing and location of deliveries, monitoring and compliance, compensa- tions, and the processes and parties involved in both planning and decision making. ­ echanisms and well-­ Moreover, effective agreements require conflict resolution m ­ greements are compatible functioning water governance. Coordination will ensure that a ­rameworks, and other local and and complementary with national laws, regulatory f regional institutional arrangements. Regional authorities and river basin organizations can offer the platform for effective negotiation, conflict resolution, and agreements. 4. Inclusion and transparency preempt future troubles. Any durable reallocation requires effective engagement with the donor region and ensuring inclusion of key actors. This may require 32 Dividing the Water, Sharing the Benefits taking into account informal or illegal water users. Failing to do so during the initial planning stage can breed resentment, distrust, and even technical problems that will likely affect later stages of negotiation and implementation. For this, communication and transparency are essential at the outset and throughout the process. This can require having a good communi- cation strategy in place and well-defined venues to ensure engagement. The process and manner for sharing the benefits and spreading the costs of reallocation should be deemed fair by the key actors to get buy in and broad support. Misinformation, distrust, rumors, and false myths can cost time and resources in the long run. 5. Droughts and crises can tip the balance, but rushing carries risks. Urbanization and economic development are the ultimate drivers for reallocation from rural to urban areas, but real- location rarely happens until there is alignment of enabling conditions (physical, finan- cial, institutional, social) and key triggers. For example, droughts often present windows of opportunity but raise the risk of rushed decisions. In other circumstances, the catalyst comes from political changes or financing. Seizing the opportunities, and limiting the risks, requires preparation of the technical analysis and engagement before these win- dows open. 6. Benefit sharing can transform conflict into cooperation. Broader approaches to compensa- tion can help reduce conflicts and support regional development opportunities for cities and rural livelihoods. The scope and scale of compensation—who receives it, who pays, and how much—should account for the needs of the donor region as a whole, not only the individuals, to capture regional economic impacts of reallocation. Forms of compensation include financial payments, investment in infrastructure, and alternative water supplies. These should be guided by efficiency, fairness, and inclusiveness. It is also critical to mea- sure and understand the socioeconomic impacts and implications of water reallocation (avoiding or minimizing preventable negative impacts on poor households, which often suffer most). 7. Be ready to learn and adapt. Even permanent reallocation requires adaptation to changing circumstances and unintended consequences, both in the donor and the recipient regions. Adjustments to reallocation agreements may be necessary due to population and eco- nomic growth in the donor region, changing legal and institutional conditions, and the emergence of new needs. To the extent possible, periodic reviews should be built into the process to avoid crisis-driven responses. Sound accounting, monitoring, and evaluation are needed to make adjustments to the original agreement. Notes ­ eallocation. 1. A dissertation project (Hooper 2015) includes a literature review of agricultural to urban r 2. All case studies in this report were covered by the systematic review focusing on urban agglomerations with populations exceeding 300,000 (Garrick et al. 2019), with the exception of Mokopane. Dividing the Water, Sharing the Benefits 33 References Aguilar-Barajas, I., and D. Garrick. 2019. “Water Reallocation in Northeastern Mexico: A Systems Analysis of El Cuchillo Dam.” Manuscript submitted for publication. Bhatia, ­R., ­R.E. Cestti, and ­J. ­Winpenny. 1995. “Water Conservation and Reallocation: Best Practice Cases in Improving Economic ­English).” Working Paper 15025, Water and Sanitation Currents Series, World Bank, Efficiency and Environmental Quality ( Washington, ­DC. ­http://documents.worldbank.org/curated/en/141221468169758468­/Water-conservation-and-reallocation-best​ -practice​-cases-in-improving-economic-efficiency-and-environmental-quality. Birkenholtz, T. 2016. “Dispossessing Irrigators: Water Grabbing, Supply-Side Growth and Farmer Resistance in India.” Geoforum 69: 94–105. Celio, ­ C. ­ M., ­ A. Scott, and ­ Giordano. 2010. “Urban–Agricultural Water Appropriation: The Hyderabad, India, C M. ­ ­ ase.” Geographical Journal 176 (1): 39–57. Civitelli, F., and G. Gruère. 2017. “Policy Options for Promoting Urban -Rural Cooperation in Water Management: A Review.” International Journal of Water Resources Development 33 (6): 852 -67. d’Amour, ­ B., ­ C. ­ G. Baiocchi, ­ F. Reitsma, ­ B. Güneralp, et ­ S. Barthel, ­ al. 2017. “Future Urban Land Expansion and Implications for Global ­Croplands.” Proceedings of the National Academy of Sciences 114 (34): 8939–44. ­https://doi.org/10.1073­/pnas.1606036114. Dai, X. P., Y. P. Han, X. H. Zhang, J. Chen, and D. X. Li. 2017. “Development of a Water Transfer Compensation Classification: A Case Study between China, Japan, America and Australia.” Agricultural Water Management 182: 151-7. Flörke, ­ C. Schneider, and ­ M., ­ I. ­ R. ­ McDonald. 2018. “Water Competition between Cities and Agriculture Driven by Climate Growth.” Nature Sustainability 1 (1): 51–58. Change and Urban ­ de Souza Leão, R., and L. De Stefano. in press. “Making Concrete Flexible: Adapting the Operating Rules of Cantareira Water System (São Paulo, Brazil).” Water Security. Garrick, D., L. De Stefano, W. Yu, I. Jorgensen, et al. 2019. “Rural Water for Thirsty Cities: A Systematic Review of Water Reallocation from Rural to Urban Regions.” Environmental Research Letters 14 (4): 043003. doi:10.1088/1748-9326/ab0db7. Grafton, Q., D. Garrick, and J. Horne. 2017. “Water Misallocation: Governance Challenges and Responses.” Policy Report pre- pared for the World Bank. World Bank, Washington, DC. Grafton, ­ Q., ­ R. ­ J. Williams, ­ C. ­ J. Perry, ­ F. Molle, ­ al. 2018. “The Paradox of Irrigation ­ C. Ringler, et ­ Efficiency.” Science 361 (6404): 748–50. doi:10.1126­/science.aat9314. Gurung, ­Y., ­J. Zhao, ­B. Kumar KC, X. Wu, B. Suwal, and ­D. Whittington. 2017. “The Costs of Delay in Infrastructure Investments: ­ epal.” Water Resources A Comparison of 2001 and 2014 Household Water Supply Coping Costs in the Kathmandu Valley, N Research 53 (8): 7078–102. Hommes, L., R. Boelens, L. M. Harris, and G. J. Veldwisch. 2019. “Rural-Urban Water Struggles: Urbanizing Hydrosocial Territories and Evolving Connections Discourses and Identities.” Water International 44 (2): 81-94. doi:10.1080/02508060.201 9.1583311. Hooper, ­V. 2015. “The Importance of the `Urban’ in Agricultural-to-Urban Water Transfers: Insights from Comparative Research in India and China.” Doctoral dissertation, University of East Anglia. Hooper, V., and Lankford, B. 2018. “Unintended Water Allocation: Gaining Share from Direct Action and Inaction.” The Oxford Handbook of Water Politics and Policy. doi:10.1093/oxfordhb/9780199335084.013.32. Libecap, ­ D. 2008. “Chinatown Revisited: Owens Valley and Los Angeles: Bargaining Costs and Fairness Perceptions of the G. ­ Exchange.” Journal of Law, Economics, and Organization 25 (2): 311–38. First Major Water Rights ­ Marston, ­ ­ .­ L., and X ­ mplementation.” WIREs Water 3 (5): Cai. 2016. “An Overview of Water Reallocation and the Barriers to Its I 658–77. doi: 10.1002/wat2.1159. Y., N. Hatcho, and S. Shindo. 2007. “Water Transfer from Agriculture to Urban Domestic Users: A Case Study of the Matsuno, ­ Tone River Basin, Japan.” Paddy and Water Environment 5 (4): 239–46. R. ­ McDonald, ­ K. Weber, ­ I., ­ M. Flörke, ­ J. Padowski, ­ al. 2014. “Water on an Urban Planet: Urbanization and the C. Schneider, et ­ Reach of Urban Water ­ Infrastructure.” Global Environmental Change 27: 96–105. 34 Dividing the Water, Sharing the Benefits R., and ­ Meinzen, Dick, ­ Ringler. 2008. “Water Reallocation: Drivers, Challenges, Threats, and Solutions for the ­ C. ­ Poor.” Journal of Human Development 9 (1): 47–64. F. 2006. Planning and Managing Water Resources at the River-Basin Level: Emergence and Evolution of a ­ Molle, ­ Concept. IWMI Comprehensive Assessment Research Report 16. Colombo, Sri Lanka: International Water Management Institute. F., and ­ Molle, ­ Berkoff. 2009. “Cities ­ J. ­ Re-allocation.” Natural Resources Forum vs. Agriculture: A Review of Intersectoral Water ­ 33 (1): 6–18. O’Donnell, E. O., D. E. Garrick, and A. C. Horne. 2019. “Reallocation through Irrigation Modernization: The ‘Once-in-a- Hundred-Year’ Opportunity of the North-South Pipeline, Australia.” Water Security 6: 100028. ­ evelopment). 2015. Water Resources Allocation: Sharing Risks and OECD (Organisation for Economic Co-operation and D Series. Paris: ­ ­Opportunities. OECD Studies on Water ­ https://doi.org/10.1787/9789264229631­ OECD. ­ -en. Rawlins, J. 2019. “Political Economy of Water Reallocation in South Africa: Insights from the Western Cape Water Crisis.” Water Security 6: 100029. B. ­ Richter, ­ D. Abell, ­ D., ­ K. Brauman, ­ E. Bacha, ­ al. 2013. “Tapped Out: How Can Cities Secure Their Water Future?” S. Calos, et ­ Water Policy 15 (3): 335–63. Scott, ­C. ­A., and ­N. ­P. ­Pablos. 2011. “Innovating Resource Regimes: Water, Wastewater, and the Institutional Dynamics of Urban Mexico.” Geoforum 42 (4): 439–50. Hydraulic Reach in Northwest ­ Sisto, N. P., A. Ramírez, I. Aguilar Barajas, and V. Magaña Rueda. 2016. “Climate Threats, Water Supply Vulnerability and the Risk of a Water Crisis in the Monterrey Metropolitan Area (Northeastern Mexico).” Physics and Chemistry of the Earth. Parts A/B/C, 91: 2-9. https://doi.org/10.1016/j.pce.2015.08.015. L. 2015. “Informal Water Markets and Willingness to Pay for Water: A Case Study of the Urban Poor in Chennai Venkatachalam, ­ City, ­India.” International Journal of Water Resources Development 31 (1), 134–45. Wester, P., S. Vargas-Velazquez, E. Mollard, and P. Silva-Ochoa. 2008. “Negotiating Surface Water Allocations to Achieve a Soft Landing in the Closed Lerma-Chapala Basin, Mexico.” International Journal of Water Resources Development 24: 275–88. Bank. 2016. “High and Dry: Climate Change, Water, and the ­ World ­ DC. Economy.” World Bank, Washington, ­ Dividing the Water, Sharing the Benefits 35 SKU W18011