InternationalWaterResourcesAssociation 44182 Water International, Volume 30, Number 4, Pages ­, December 2005 © 2005 International Water Resources Association CooperationonInternationalRivers A Continuum for Securing and Sharing Benefits Claudia W. Sadoff and David Grey, the World Bank, Washington, DC, USA Abstract: It is generally accepted that conflicting demands over international rivers will intensify. There is an active debate on whether this will lead to "water wars" or to unprecedented cooperation. Framing the debate in this manner, however, tends to cast the concept of cooperation as all-or-nothing, implying that "cooperation" is an extreme, in direct opposition to "war." This conceptual construct obscures the many practical levels of cooperation that states can undertake to their mutual advantage. It is important to recognize that it is entirely rational that states will always have a "national agenda" for a river that they share with other states, and that they will cooperate if it serves that national agenda. In practice, there can be a continuum of levels of cooperation, from simple information sharing, to joint ownership and management of infrastructure investments. Furthermore, it may not necessarily be the case that "more" cooperation reaps "more" benefits in all river basins. There are many different types of benefits that can be secured through the cooperative management of international waters, with each individual basin offering different potential cooperative benefits with different associated costs. For each international basin, the optimal mode of cooperation will depend on a mix of factors including hydrologic characteristics, the economics of cooperative investments, numbers, the relationships of riparians, and the costs of parties coming together. Keywords: international rivers, cooperation, benefit sharing, water resources, river basin management elicit extremes of cooperation or dispute or can elicit re- Introduction actions anywhere in between these extremes. Much re- cent literature exists on the imperative of cooperation Rivers wind through the histories and cultures of na- between nations sharing international rivers, but little has tions in extraordinary ways. Today international rivers form been written on the practicalities of achieving it. Achiev- an increasingly important part of the geographic, economic ing international cooperation is always a long and com- and political landscape of our world (freshwater flows plex journey, for which there is no single path and few [whether surface water or groundwater], and the lakes short cuts. Instead, there are many routes that can be and wetlands that some of these flows may pass through, followed and many steps that can be taken, with various derive from, or terminate within are described, very loosely, options to consider and choices to be made. in this text as "rivers." The term "international rivers" is This paper explores the practicalities of achieving used in the text to refer to freshwaters whose basins are cooperation on international rivers, and offers a menu of situated within the borders of more than one state). About options and choices to consider. (While this paper focuses 40 percent of the world's population lives within the basins on the challenges of international rivers, clear parallels of international rivers, and, perhaps even more significantly, may be drawn to the conflicting interests of different user over 90 percent of the world's population lives within the groups within national river basins as well). At the heart countries that share these basins. These rivers create na- of this framework is the potential to move from national tional expectations ­ both within and beyond the borders agendas that are unilateral, to national agendas that incor- of their basins ­ of the benefits they can bring. As popula- porate significant cooperation, and to converge upon a tions and economies grow, and as less contentious national shared cooperative agenda. The extent to which this will water resources become more fully exploited, an increas- occur will be determined by each party's perception of ing share of the remaining development opportunities will the benefits it can secure from cooperation. Convergence be on international rivers. Development of these rivers can towards a cooperative agenda will be facilitated by sev- 1 2 C. Sadoff and D. Grey eral important and practical steps. First, the perception of efits are perceived to be insufficient relative to the costs the range and extent of potential benefits needs to be ex- of cooperation. Benefits are broadly defined here to in- panded as much as possible, from the obvious to the less clude economic, social, environmental, and political gains. apparent. Second, the distribution of benefits and benefit- Integrated, basin-wide water resources management is sharing opportunities to redistribute the costs and benefits increasingly recognized as the ultimate goal for ensuring of cooperation need to be explored to enable the definition the sustainability and productivity of river systems and is a of a cooperative agenda that will be perceived as fair by challenge in any setting, as the priorities and concerns of all parties. Third, alternative modes of cooperation need myriad users must be reconciled. In the context of inter- to be recognized and appropriate types of cooperation iden- national rivers, efforts toward integrated management can- tified to secure the greatest net benefits. Each of these notbemadewithoutinternationalcooperation.Thecomplexity steps is examined below. and costs of international cooperation can be very great, and must be achieved in the absence of any ultimate entity with NationalAgendas: ConvergingToward the mandate and authority to impose a solution. Cooperation A useful framework for broadening the range of rec- ognized benefits of cooperation proposes the identifica- Each sovereign country will have its own national tion of four types of cooperative benefits (see Sadoff and agenda on an international river -- this is obvious, ratio- Grey, 2002). The first type of benefit derives from coop- nal, and legitimate. Thus, in a river basin shared by two eration that enables better management of ecosystems, states, there will be two separate national agendas. If these providing benefits to the river, and underpinning all other two agendas overlap in some way, there will be a third, benefits that can be derived. The second type of benefit cooperative agenda of some scale ­ from very limited to derives from the efficient, cooperative management and substantial. As the benefits of this cooperation are pro- development of shared rivers, yielding major benefits from gressively identified and secured, this third agenda may the river, in increased food and energy production, for ex- grow, with the two national agendas converging into a co- ample. The third type of benefit derives from the lessen- operative agenda for the two nations ­ each of which will ing of tensions because of cooperation, resulting in the still view the cooperative agenda to be their national agenda. reduction of costs because of the river, as tensions be- In this case, the emerging single cooperative agenda will tween co-riparian states will always be present, to a greater need to provide benefits that exceed the sum of the two or lesser extent, and those tensions will generate costs. non-cooperative national agendas, and will thus have be- And finally, as international rivers can be catalytic agents, come the rational choice of each sovereign nation (Figure cooperation that yields benefits from the river and reduces 1). Cooperation on an international river can bring many costs because of the river can yield a fourth type of ben- benefits that may allow the whole to be greater than the sum efit derived from greater cooperation between states, even of the parts, in part because treating the river basin as one economic integration among states, generating benefits systemallowsoptimizedmanagementanddevelopment(the beyond the river (Figure 2). ultimate goal of integrated water resources management). While each of these four types of benefits could po- tentially be obtained in all international river basins, the BenefitsofCooperation:LookingBeyondtheRiver scale, feasibility, and relative importance of each type will vary greatly between basins, reflecting a wide range of A first step in motivating cooperation is to recognize political, geographic, economic, and cultural circumstances. the widest possible range of potential benefits that coop- There is no hierarchy among the four types with regard to eration could bring. There will be no cooperation if ben- the magnitude of potential benefits. The relative magni- tude of different types of potential benefits within a par- 3 3 3 Type The Challenge The Opportunities Type 1 Degraded water quality, Improved water quality, riverflow Increasing Benefits watersheds, wetlands, & characteristics, soil conservation, To the River biodiversity biodiversity and overall sustainability Type 2 Increasing demands for water, Improved water resources Increasing Benefits sub-optimal water resources management for hydropower & 1 2 agricultural production, flood-drought 1 2 1 2 1 2 From the River management & development management, navigation, environmental conservation, water quality & recreation Case 1 Case 2 Case 3 Case 4 Type 3 Tense regional relations & Policy shift to cooperation & Reducing Costs political economy impacts development, away from ...all cases can be rational Because of the River dispute/conflict; from food (& energy) self-sufficiency to food (& energy) security; reduced dispute/conflict risk & military expenditure Country 1 Agenda Type 4 Regional fragmentation Integration of regional infrastructure, Country 2 Agenda Increasing Benefits Beyond markets & trade the River Cooperative Agenda 3 Figure 1. Converging agendas Figure 2. Types of benefits of cooperation on international rivers IWRA, Water International, Volume 30, Number 4, December 2005 Cooperation on International Rivers: A Continuum for Securing and Sharing Benefits 3 ticular basin will be a result of physical opportunities, costs, Benefit sharing provides riparians with the flexibility and the type of cooperation that is developed between to separate the physical distribution of river development riparian states. Nor is there a particular sequence in which (where activities are undertaken), from the economic dis- these four types should be pursued, as, wherever initial tribution of benefits (who receives the benefits of those cooperation focuses, there will be linkages with other types activities.) This allows riparians to focus firstly on gener- of cooperation. Making a start in environmental (Type 1) ating basin-wide benefits, and secondly on sharing those or direct economic cooperation (Type 2) can lead to growing benefits in a manner that is agreed as fair. Furthermore, a political (Type 3) and indirect economic cooperation (Type focus on sharing the benefits derived from the use of wa- 4) ­ or vice versa. The dynamics between types might be ter, rather than the allocation of water itself, provides far positive or negative. For example, while Type 3 coopera- greater scope for identifying mutually beneficial coopera- tion may help further advance Type 1 and Type 2 coop- tive actions. While the allocation of water, particularly in in- eration, setbacks in Type 3 relations may impede ternationalsystems,isoftencontentious,theunderlyinginterest cooperation of Types 1 and 2. of most riparians is to secure the benefits of water use. To negotiate the management and development of inter- Benefit Sharing:Achieving Fair Shares national shared rivers, riparians can focus their negotiations on the allocation of water rights or on the distribution of ben- Another very challenging step in facilitating the con- efits derived from the use of water, as shown in Table 1. vergence of national agendas is an analysis of the distri- The concept of water sharing by assigning rights char- butionofbenefitsfromcooperation.Thisanalysisisessential acterized the 20th century and remains the most widely- because a program of interventions designed to provide recognized mechanism for riparian states to engage in net gains for a basin as a whole will not necessarily pro- sharing international rivers. Today this is guided primarily vide net gains for each country. If significant benefits ac- by principles first established in the 1966 Helsinki Rules crue in one country, while significant costs are borne by on the Uses of the Waters of International Rivers and another, it is possible that a project providing net benefits then codified in the 1997 United Nations Convention on on a basin-wide scale could actually generate net losses in the Law of the Non-navigable Uses of International Wa- any one country. If benefits are secured where they are tercourses. This growing body of international water law generated under an optimal cooperative scenario (e.g., the sets out general factors upon which "reasonable and equi- most productive hydropower or irrigation sites), the distri- table" utilization of international watercourses should be bution of benefits this creates may well be perceived as based. The factors are not prioritized, except for a clause unfair by some riparians. in the UN Convention that states that "special regard" Where this initial distribution of benefits from a coop- should be given to "the requirements of vital human needs." eration management and development scenario is seen as Water use will likely be optimized within individual states unfair, benefit-sharing mechanisms can play a pivotal role and not across the basin, and thus the assigning of water in motivating cooperation. Benefit sharing can be defined rights may be inefficient. Furthermore, as populations and as any action designed to change the allocation of costs economies grow, in many river basins there will increas- and benefits associated with cooperation. This would in- ingly be insufficient water to apportion "reasonably and clude benefits of all four types (Figure 2) and all costs of equitably" between riparian states. cooperation. The costs of cooperation could be directly Direct payment for water is an alternative mecha- associated with the institutional or physical costs of river nism to (re-)assigning water rights which provides the development and management (for example, river regula- buyer the opportunity to benefit from the use of water tion and storage costs), or any other costs that the negoti- without transferring water rights. This mechanism could ating parties choose to include for consideration (for be used in a basin where the assignment of water rights example, hydropower interconnection and distribution was clear and where a cooperative scheme called for in- costs.) In most cases, benefit sharing will require some creased abstraction by one riparian. International water sort of redistribution or compensation, which will be highly markets could provide a flexible mechanism for reallocat- situation specific. ing water use among riparians within an agreed compen- Table 1. Focuses for Riparian Negotiations Water Sharing Benefit Sharing Water sharing by Direct payment for water use (e.g., municipal or irrigation supplies) (rights already assigned) assigning rights Direct payment for benefits (e.g., fisheries, watershed management) or compensation for costs (e.g., inundated land, pollution Purchase agreements (e.g., power, agriculture products) (benefit transfer through terms/price) Financing and ownership arrangements (e.g., power infrastructure) (benefit transfer through deal structure) Broadened bundle of benefits, including provision of unrelated goods and services and less tangible (e.g. reputation) benefits IWRA, Water International, Volume 30, Number 4, December 2005 4 C. Sadoff and D. Grey sation structure. Such markets would allow riparians to gainsfromtheintensifiedcooperationand,potentially,thetrust buy and sell fixed-term water use rights that would not implicitinjointownershipandmanagementofassets. necessarily affect any existing water treaty rights. The Bundling of broader benefits can also be a mecha- price and quantity of water use rights could be decided by nism for sharing benefits. It may be difficult in some cases market forces or they could be negotiated. Agreements to find a configuration of benefits that satisfies all parties. regarding the price of water, the volume of water to be In such cases, the broader the range of benefits under made available, or the eligibility of buyers would all affect discussion, the more likely riparians will be able to find a the distribution of benefits derived from that water. configuration that is mutually acceptable. If, for example, Payments for benefits (or compensation for costs) it is difficult to negotiate the benefits of a hydropower might be made in the context of a cooperative scheme. investment, the bundle could be broadened to include wa- Riparians can be compensated, for example, for land in- tershed management in the area of the reservoir, and power undation as a consequence of water impoundment by an- interconnection and trade. In addition, other benefits de- other riparian. In some instances it might be appropriate rived from unrelated projects, such as transport infrastruc- to make payments to upstream riparians for watershed ture or from areas of mutual interest, such as trade, management that bring benefits downstream (e.g. reduced immigration, communications, and environmental protec- flooding and sediment loads). Thus stewardship of head- tion, can be bundled with water use-related benefits. This waters and watersheds might entitle upstream riparians to could simplify international rivers negotiations, if the par- share some portion of the downstream benefits that their ties felt that their inclusion could produce a bundle of co- stewardship helps to facilitate, and thus share the costs of operative projects all would accept as fair. Geopolitical that stewardship. Seen the other way around, if they did and diplomatic relationships, public image and reputation, not protect the watershed it would impose costs on down- international support, and even less tangible benefits, such stream riparians. as ethical returns within a broader community of interests, Purchase agreements can be structured as flexible might also influence states engaged in discussions of co- tools for benefit sharing. Purchase agreements are gener- operative management of shared waters. The full range ally negotiated for power, but they could also be negoti- of benefit-sharing mechanisms can then be brought to bear ated for water supply, fisheries, agricultural products, etc. on this broader bundle of benefits in order to reach an The negotiated price in the purchase agreement can ef- acceptable solution. fectively re-allocate the benefits of water use among The sharing of benefits from the utilization of an inter- riparians. While each would clearly seek to be better off national watercourse is a new paradigm, for which there by the transaction if they were willing to enter into the appears to be no directly relevant reference in interna- trade, a higher agreed price would transfer proportionally tional water law beyond the principles, factors, and norms more benefits to the selling riparian, while a lower agreed for sharing water by the allocation of rights. Without such price would apportion more benefits to the buying ripar- principles and guidelines, unguided negotiations ­ and very ian. Purchase agreements can enable a range of mutu- difficult multi-party negotiations in some cases ­ may be ally-beneficial scenarios. Clearly, for example, when one the only way forward. Nevertheless, the principles of "eq- riparian has water resources or hydropower capacity but uitable and reasonable utilization" and of "no significant insufficient national demand for water and/or power, while harm" could provide a basis upon which benefit alloca- the other has meager water resources and hydropower tions can be considered. Policy makers and negotiators capacity but significant demand, both will benefit from this need to translate these principles into practice and develop trade. Purchase agreements can also provide revenue other principles and guidelines, finding practical rules and guarantees that may be required to secure financing for mechanisms for benefit allocation and mechanisms for large-scale projects. redistribution and/or compensation. Financing and ownership arrangements can be used A perception by all riparians that a cooperative basin to effect benefit sharing and transfer through the struc- management scheme which maximizes overall benefits is ture of the deal, especially when cooperative management "fair" will be essential to motivate and sustain coopera- calls for large-scale infrastructure investments. One ri- tion. While the identification and generation of coopera- parian could provide financing for another as a means of tive benefits enlarges the "pie," a larger pie will not satisfy facilitating investment, and, if the financing agreement all riparians if their particular "slice" of the pie is not larger were not concluded at strictly market terms, as a means (either literally, or even relatively to others). If benefit al- of reapportioning benefits. Joint financing of cooperative location takes place at the natural, physical location of projects, possibly including equity shares, has also been a benefit generation in an optimal cooperative scenario (e.g., successful means of facilitating cooperation and sharing hydropower or irrigation sites) and is not considered fair benefits, by embedding a redistribution of returns into the or if it provides a particular riparian with a smaller share deal structure. Jointly-owned operating or holding compa- of benefits than it would receive in the absence of coop- nies (for example, a joint power transmission company) eration, benefit-sharing mechanisms may be needed to can provide direct financial benefits, as well as broader motivate cooperation. Even when cooperation could gen- IWRA, Water International, Volume 30, Number 4, December 2005 Cooperation on International Rivers: A Continuum for Securing and Sharing Benefits 5 erate greater gains for all, perceived inequities in the dis- of benefits, such as enabling improved flow forecasting tribution of gains may make cooperation unacceptable to and greater preparedness for floods and droughts. The some. Thus concerns over the relative strength of neigh- exchange of information on development plans will help bors and trading partners can influence perceptions of fair- basin planners in different countries avoid conflicting ness and can thus also be critical factors in motivating and projects, particularly where planners assess their national negotiating cooperation. The redistribution of benefits will projects for impacts, costs, and benefits, if any, extending itself require cooperation, with some mechanisms requir- beyond their borders. (This raises an important point re- ing greater efforts than others. garding the direction of the effects of development. While upstream extraction generates externalities downstream Modes of Cooperation: Recognizing a Continuum by diminishing or deteriorating flows physically, down- stream extraction generates externalities upstream by di- A further step in promoting cooperation is to identify minishing future flows available for abstraction upstream, the various modes of cooperation that could be adopted, by virtue of perceptions of acquired rights to that water and determine the appropriate type of cooperative effort downstream.) Coordination on international rivers may to achieve a particular goal. The optimal type of coopera- enable nations to secure some Type 1 and 2 benefits (Figure tion will vary with hydrologic and investment opportunities 1) and, to some extent,Type 3 benefits, because tensions will and with the consequent potential benefit-sharing mecha- reduce as trust grows. At this level of cooperation, regional nisms in each basin. In some basins, information sharing assessments undertaken in a cooperative manner can pro- and basin-wide strategic assessments may be adequate to videaplatformofinformationsymmetrywhichmayfacilitate facilitate optimal cooperative management. In others, joint more intensive cooperation (further discussed below). actions in river regulation, water storage, and drought and Collaboration results when national plans are adapted flood mitigation would yield significant net benefits.Acon- either to secure gains or to mitigate harm in another ripar- tinuum of cooperation can be conceived from unilateral ian country. Collaboration could generate benefits of all action (independent, non-transparent national plans), to four types and could be achieved simply through ad hoc coordination (communication and information on national adaptations of ongoing plans or through agreed portfolios plans), to collaboration (adaptation of national plans for of national projects developed from a basin perspective. mutual benefits), to joint action (joint plans, management As with coordination, collaboration on international rivers or investment) (Figure 3). may enable nations to secure direct Type 1 and 2 benefits. Unilateral action in a basin would mean no coopera- Benefit-sharing mechanisms may be needed to effect a tion, not even communication or information exchange, over redistribution of benefits that is perceived to be fair; this the management and development of the shared river. Not may further bind countries together and build trust, lever- only do such arrangements forgo the opportunity to se- aging Type 3 benefits or even Type 4 benefits if the cho- cure cooperative gains, but they can lead to situations sen benefit-sharing mechanism secures or redistributes where riparian countries' development and investment broader bundles of benefits "beyond the river." schemes undermine one another. The cumulative impact Joint action occurs when riparians act as partners in of these uncoordinated developments may diminish flows the design, investment, and implementation of international or degrade water quality to the point that all activities may rivers development. This level of cooperation will need to be compromised. be formalized by treaties. Benefit-sharing arrangements Coordination is achieved through the exchange, or the such as joint ownership and management of assets repre- cooperative gathering, of information in a basin. The ex- sent the greatest cooperative effort. Situations that lend change of hydrologic information could generate a range themselves to this type of cooperation would include ba- sins in which there is strong cooperation, capacity, and ·Identify, negotiate and institutions. Joint action might include private sector en- ·Joint project implement suites of assessment and ·Communication and national investments design notification gagement among co-riparian states and state-of-the-art ·Adapt national plans to ·Joint ownership ·Information sharing mitigate regional costs ·Joint institutions management and investment scenarios that could optimize ·Regional ·Adapt national plans to assessments capture regional gains ·Joint investment direct Type 1 and 2 benefits, as well as indirect Type 3 Inform and 4 benefits. Less obviously, joint action might even be Adapt Join the right type of cooperation for basins in which there is little or uneven capacity, and high levels of mistrust, pro- Dispute Cooperation Continuum Integration viding transparency, comfort, and security to affected Unilateral Coordination Collaboration Joint Action Action riparians. Here, in addition to capturing the direct benefits of Types 1 and 2, joint action could be a powerful vehicle Type 1 benefits Type 2 benefits to deliver indirect Types 3 and 4 benefits. Type 3 benefits It is important to recognize that this cooperation con- Type 4 benefits tinuum is non-directive, dynamic, and iterative. The con- Figure 3. Types of cooperation ­ the Cooperative Continuum tinuum is non-directive as it is not intended to suggest that IWRA, Water International, Volume 30, Number 4, December 2005 6 C. Sadoff and D. Grey Indus Mekong Rhine Orange Senegal River some basins may require sizeable up-front investments in commun- cooperation before any significant benefits are secured. info sharing, convergent joint prep joint type of ication assessments national and equity cooperation Both costs and benefits in this calculation may be subtle. agendas investment ownership type of Costs will include financial, institutional, and political/rela- 1-3 1-4 1,3 2,3 1-4 benefit tional costs, and may also include the cost of unilateral purchase type of cost sharing cost sharing agreement, joint benefit opportunities (benefits) foregone. Benefits will include all financing ownership sharing four types described above, some obvious and readily as- sessed,andsomemuchlessapparent,incorporatingnon-quan- tifiablebenefitssuchasinternationalgoodwill,regionalstability, Dispute Cooperation Continuum Integration and the preservation of valuable cultural and natural assets. Unilateral Coordination Collaboration Joint Different modes of cooperative effort will create dif- Action Action ferent options for benefit sharing (Figure 5), and similarly Figure 4. Types of cooperation ­ some examples different benefit-sharing mechanisms will require differ- more cooperation is necessarily better, although it is con- ent levels of cooperation. Payments for water and pay- structed to portray increasing cooperative effort. The con- ments for benefits, for example, could be fairly tinuum is dynamic as there will be varying points on the straightforward. The establishment of water markets for continuum that are appropriate for different activities at iterative trading of water-use rights would require greater different times, and nations may choose to adapt their ac- cooperative effort, as would purchase agreements and tivities to increase or decrease the intensity of their coop- cooperative financing or ownership and management. eration in response to new opportunities or developments Benefit-sharing mechanisms themselves thus become ve- within, or to broader events outside, the cooperative process. hicles for cooperation and ties that bind riparians together. The continuum is iterative, because there will be repeated, The specific configuration of costs and range of ben- discreteopportunitiesforcooperation,andthesuccessofearlier efits in a basin will determine the mode of cooperation cooperation,particularlyintermsofrealizedbenefits,willlikely that is called for in order to secure cooperative benefits. promote increasing cooperation ­ and vice versa. The challenge then is to share these benefits fairly. Figure 4 illustrates different modes of cooperation in various international river basins around the world.Ama- Cooperative RegionalAssessments: Promoting jor challenge in each basin is to identify the right type of Cooperation cooperative effort ­ one in which the benefits of coopera- tion outweigh the costs, and the process and outcome is Cooperative Regional Assessments (CRAs) are tools politically and socially acceptable. For some basins, move- specifically designed to promote cooperation on interna- ment along the continuum may be a constructive goal over tional rivers. On the one hand CRAs are rational "regional time. In other basins, different points along the continuum assessments" of sectors (i.e., power, agriculture) or themes may be the right choice for capturing specific benefits at (i.e.,watershedmanagement,capacitybuilding).Ontheother specific points in time. In yet other basins no cooperation, hand CRAs are "cooperative," involving a process which or very limited cooperation, may be rational. Riparian states brings riparian states together (potentially including govern- will pursue cooperation only when they expect to receive ment, the private sector and civil society) to reach common greater benefits through cooperation than through unilat- understanding, change perceptions, and achieve information eral action, and when they believe that benefits can be symmetry, in order to build trust and catalyze cooperation. secured in a manner they perceive as feasible, cost effec- CRAs are essentially practical tools for identifying op- tive, and fair. portunitiesforregionalactions,forpromotingtherecognition The feasibility and cost-effectiveness of cooperation and optimization of all four types of benefits of cooperation, will vary enormously by basin, with the appropriate level for analyzing the distribution of costs and benefits associ- of cooperation needed to secure cooperative benefits de- ated with cooperative regional programs, and for identify- pending upon the combination of specific costs and poten- ingbenefit-sharingandinstitutionalmechanismsfor realizing tial benefits in any river system. Even rudimentary these benefits. CRAs identify and provide substantive ba- cooperation could be challenging and costly if, for example, sin-wide analyses of the range of development options, water rights are contested, relations are strained, or ca- and they inform the identification and selection of appro- pacity is low; however, the benefits of achieving such co- priate projects. CRAs are less detailed than, and are dis- operation could be high. On the other hand, the cost of tinct from, project-specific impact analyses which need to intensive cooperation could be modest if institutions, ca- be undertaken at the project level to thoroughly assess the pacity, and relationships already exist. Thus, some basins impacts(economic,environmental,social,etc.)ofindividual may secure a significant share of their potential coopera- projects on co-riparian states. tive gains with low levels of cooperation, and find that The design of individual CRAs will vary widely ac- further cooperative efforts are not justified.Alternatively, cording to circumstance: the complexity of the basin, theme, or sector; the number of riparians and the extent of their IWRA, Water International, Volume 30, Number 4, December 2005 Cooperation on International Rivers: A Continuum for Securing and Sharing Benefits 7 ·Negotiated/consensus bundles of projects pursuit of any shared goal. In many cases it will be strate- ·Payment for water · Joint equity ownership gic to be pragmatic and opportunist, by focusing initial ef- or water markets · Cost sharing for data gathering and · Payment for benefits · Joint financing forts on the areas where there is the greatest clarity over analysis · Joint management potential benefits and the least adaptation required from · Cost sharing for · Payment for watershed and operations management regional assessments contracts national agendas. In some cases it may instead be strate- · Purchase agreements gic to be visionary and bold. The dynamics of cooperation are such that progress in any area will help to develop Dispute Cooperation Continuum Integration cooperative processes, to establish relationships and insti- Unilateral Coordination Collaboration Joint tutions, and to gain momentum to facilitate more difficult Action Action undertakings. Initial hopes for achieving high levels of co- Figure 5. Benefit sharing along the continuum operation and large-scale comprehensive activities should cooperation; the availability of information; timeframe and not preclude efforts to secure more modest cooperative funding constraints, etc. CRAs can be as simple as desk benefits, but all cooperative activities are likely to strengthen studies that reconcile national-level assessments, or they the cooperative process and create opportunities for more can be designed holistically on a basin-wide scale. In all beneficial cooperation. Picking "low-hanging fruit" will cases they will bring riparian specialists together in a pro- secure the concrete benefits of cooperation ­ even if it is cess that builds common understanding and relationships. low-level cooperation ­ that will help motivate continued CRAs need to be strategic, highlighting the broadest pos- and likely increasing levels of cooperation. sible range of potential projects and benefits, and the op- Figure 6 illustrates some of the dynamics of coopera- tions and choices that are available to secure and share tion. At each stage of a program of development on an the benefits of cooperation. CRAs will generally include: international river, there will be opportunities for coopera- · A Transboundary Analysis of the range of potential tion, and cooperation at any stage will promote the capac- benefits of cooperation, providing a basin-wide view ity and willingness for future cooperation. Cooperation can of the best possible river management and develop- start at the analytic stage, for example using CRAs to ment opportunities (i.e. "without borders"); identify potential benefits. Riparians can together explore · A Distributive Analysis of the relative share of ben- the range of potential cooperative projects and benefits, efits and costs for each riparian nation under alterna- and the distribution of those benefits.Alternatively, coop- tive management and development scenarios (to ensure eration can begin with negotiations, when national agen- that a program designed to maximize net gains for the das are disclosed and opportunities are explored to adapt basin as a whole, will also provide acceptable gains for or bundle projects in a cooperative manner, or to redistrib- each country individually), and exploration of the vari- ute costs and benefits. Cooperation can even begin with ous options for sharing the costs and benefits of coop- coordination at the stage of project implementation, with eration more equitably (i.e. "with borders"); and information sharing and communication, which may lead · An Institutional Analysis of the possible modes of to further collaboration, if appropriate. cooperation necessary to generate the greatest net These dynamics are in part a reflection of the itera- benefits, taking account of the costs of cooperation. tive nature of cooperation on international rivers. Rivers will remain, water needs will grow, non-cooperation will CRAs are not negotiations and do not determine out- become increasingly threatening, and cooperation will be- comes. Instead they explore opportunities for cooperation come increasingly important. All interactions on interna- in an increasingly basin-wide view and they are an inte- tional rivers are therefore undertaken in an iterative context gral part of the process of building trust and confidence ­ it is assumed that in the future similar interactions will among riparian partners. The CRA can inform initial project occur and that current actions will inform and affect fu- selection and design, to enhance project efficiency, fairness, ture decisions. Thus current good faith cooperation should and feasibility from the perspective of all riparian states. The promote future good faith cooperation. CRA can also provide riparian states with the information The dynamics are also a reflection of the changes needed to reach consensus regarding the way forward in both in perceptions and realities that result from coopera- projects of mutual interest. CRAs can thus provide a com- tive experiences. The experience of successful coopera- mon point of departure for the (essentially political) negotia- tion can yield concrete benefits and build trust and tions needed to agree a cooperative regional agenda. relationships. This changes perceptions with regard to the potential benefits of cooperation, and to the feasibility of working cooperatively with co-riparians. Realities may also The Dynamics of Cooperation: Reinforcing and change as cooperation progresses. As opportunities are Iterative identified, as precedents, institutions, and relationships are Good faith cooperative efforts will be self-reinforcing established to facilitate cooperation, the real costs and ­ cooperation will promote a willingness to cooperate. Co- benefits of cooperation will change. operation can start anywhere, at any level of effort, in IWRA, Water International, Volume 30, Number 4, December 2005 8 C. Sadoff and D. Grey ·Delivers real benefits of cooperation ·Reinforces trust & builds capacity through cooperative action Identify po ·Motivates further cooperation by Cooperative delivering concrete results Assessments Regional Im Capture benefits: Analyze Implement projects Implement Assessments individually or together Negotiate ·Assures acceptability/feasibility Agree solution: ·Reinforces trust & builds capacity through good faith negotiations Negotiate a bundle of ·Motivates further cooperation by projects, benefit sharing agreeing the range of mechanisms arrangements & legal for cooperation & benefit sharing agreements (if any) Figure 6. The dynamics of cooperation ­ iterative and reinforcing Conclusions AbouttheAuthors In all international river basins, there are likely to be Claudia W. Sadoff is a Lead Economist at the World benefits to be derived from a move from unilateral na- Bank, where she leads the Water ResourcesAnchor Team tional agendas to some mode of riparian cooperation. The and is a core member of the World Bank's Nile Team. Dr. extent to which national agendas will converge toward a Sadoff specializes in natural resources economics, with shared, cooperative agenda will be a consequence of the particular emphasis on water resources and international perceptions of benefits that cooperation will bring. The rivers, and she has worked on water policy and related full breadth of cooperative benefits should therefore be issues in several countries in Africa and Asia. explored, as should the distribution of benefits and poten- tial for alternative benefit-sharing scenarios. Riparian co- David Grey is the World Bank's Senior Water Advi- operation in pursuit of these benefits can, and should, take sor, the Senior Water Advisor for the Africa Region, and many forms. The continuum presented in this paper offers leader of the multi-donor Nile Team that supports the ten a menu of cooperative options that may assist riparians in riparian states of the River Nile in their Nile Basin Initia- determining the right modes of cooperation for activities tive. Mr. Grey has worked on water issues in many coun- within their basin. This continuum is not static, but con- tries in Africa, Asia, the Middle East, Latin America, and ceived as iterative, adaptive, and dynamic. For some ba- Europe, including intensive work on water resources policy sins, movement along the continuum may be a constructive and institutions, international waters cooperation and wa- goal over time. In other basins, different points along the ter supply. continuum may be the right choice for capturing specific gains. The continuum is also not intended to be directive Discussions open until May 1, 2006. or normative in any way; it is not the case that greater cooperation is necessarily "better" or will reap greater net References gains. The uniqueness of each international basin will of- fer a different set of potential cooperative benefits, calling Sadoff, C.W. and D. Grey. 2002. "Beyond the river: the benefits for different modes of cooperation and a different set of of cooperation on international rivers." Water Policy 4, No. cooperative and benefit-sharing mechanisms. 5:389-403. Acknowledgements The authors gratefully acknowledge the input and in- sights of our colleagues Ashok Subramanian, Andrew Macoun, Christina Leb,Astrid Hillers, and IngerAndersen. IWRA, Water International, Volume 30, Number 4, December 2005