Patterns in International Water Resource Treaties:
The Transboundary Freshwater Dispute Database

by Jesse H. Hamner and Aaron T. Wolf
Published in:
Colorado Journal of International
Environmental Law and Policy. 1997 Yearbook, 1998.

Footnotes (not yet incorporated in text)

Introduction: The Transboundary Freshwater Dispute Database

Approximately 261 international watersheds, and untold number of transboundary aquifers, cover about one-half of the globe's land surface. This vital resource -- a scarce resource which has no substitute, which has poorly-developed international law, and the need for which is overwhelming, constant, and immediate -- has driven its share of political tensions. Water has exacerbated tensions around the globe, most famously in the arid and hostile Middle East, but also throughout Africa and Asia.

The fortunate corollary of water as an inducement to conflict is that water, by its very nature, tends to induce even hostile co-riparians to cooperate, and even as disputes rage over other issues. In fact, the weight of historic evidence tends to favor water as a catalyst for cooperation: organized political bodies have signed 3,600 water-related treaties since AD 805, versus only seven minor international water-related skirmishes (each of which included other non-water issues).  The only water-related war between states on record occurred about 4,500 years ago.  Given this disproportionate evidence in favor of "hydro-cooperation," the processes of conflict resolution and amelioration warrant more study, although at present many scholars focus on the potential for violent conflict sparked by water disagreements.

The UN Food and Agriculture Organization has identified more than 3,600 treaties relating to international water resources dating between AD 805 and 1984, the majority of which deal with some aspect of navigation.  Since 1814, states have negotiated a smaller body of treaties which deal with non-navigational issues of water management, flood control, hydropower projects, or allocations for consumptive or non-consumptive uses in international basins. Including only those dating from 1870 and later which deal with water per se, and excluding those which deal only with boundaries, navigation, or fishing rights, the authors have collected full and partial texts of 145 treaties in a Transboundary Freshwater Dispute Database at the University of Alabama. The collection and translation efforts continue in an ongoing project of the Department of Geography and the Center for Freshwater Studies, in conjunction with projects funded by the World Bank and the US Institute of Peace. Table 1 lists the treaties in the Database chronologically.

Negotiating notes and published descriptions of treaty negotiations are also being collected in the Database, and at present the Database includes fourteen detailed case studies. These cases include nine watersheds (the Danube, Euphrates, Jordan, Ganges, Indus, Mekong, Nile, La Plata, and Salween); two sets of aquifer systems (US-Mexico shared systems and the West Bank Aquifers); two lake systems (the Aral Sea and the Great Lakes); and one engineering works (the Lesotho Highlands Project). Table 2 lists these case studies with some defining characteristics.

At present, few authors have undertaken systematic work on the body of international water treaties as a whole, although some use treaty examples to make a point about specific conflicts, areas of cooperation, or larger issues of water law.

Treaties can tell about regional hegemony, about how and which water needs are met, about the relative importance of water in the political climate, about development issues, and whether earlier treaties have successfully guided or guaranteed state behavior. This article summarizes the general findings from a comparative assessment of the treaties in the Database. To organize and analyze these treaties, a systematic computer compilation has been developed which catalogs the treaties by basin, countries involved, date signed, treaty topic, allocations measure, conflict resolution mechanisms, and non-water linkages.

Major Findings

The contents of the Database are qualitatively and quantitatively assessed for their provisions regarding the following criteria: basin involved; principal focus; number of signatories; non-water linkages (such as money, land, or concessions in exchange for water supply or access to water); provisions for monitoring, enforcement, and conflict resolution; method and amount of water division, if any; and the date signed (two sample summary sheets are included as Figure 1). Preliminary descriptions of our findings follow, and Table 3 provides a statistical summary.

Treaty Signatories:

One hundred twenty-four of the 145 treaties (86%) are bilateral. Twenty-one (14%) are multilateral; two of the multilateral treaties are unsigned agreements or drafts.

It is unclear whether so many of the treaties are bilateral because only two states share a majority of international watersheds or because, according to negotiation theory, the difficulty of negotiations increases as the number of parties increases.  In basins with more than two riparians, this preference for bilateral agreements can preclude the comprehensive regional management long advocated by water resource managers. One who ignores the watershed as the fundamental planning unit, where surface- and groundwater, quality and quantity are all inter-related, also ignores hydrologic reality. The Jordan basin, for example, has been characterized by bilateral arrangements; the only regional talks on the basin, the Johnston negotiations of 1953-1955, went unratified. As unilateral development in the basin proceeded in the absence of agreement, each state's goals and plans abutted against those of the other co-riparians, leading to inefficient development and even to exchanges of fire in the early 1950s and mid-1960s.  Similarly, India has a long-standing policy of adhering to bilateral negotiations,  presumably because it can best address its own needs vis-á-vis each of its neighbors separately. Partly as a consequence, neither the Ganges-Brahmaputra nor the Indus River systems have ever been managed to their potential efficiency.

Of the twenty-one multilateral treaties/agreements, developing nations are parties to thirteen. Only one multilateral treaty exists between industrialized nations for access to a water source, namely the treaty regarding water withdrawals from Lake Constance signed by Germany, Austria, and Switzerland in 1966.  None of the preindustrial-nation multilateral agreements specified any water allocations-all involved hydropower or industrial uses or both.

The states surrounding the Aral Sea have an agreement, dated 1993, that deals with several joint issues, but the text lacks allocations and provides too little detail for planning water use.  Like the Aral Sea, Lake Chad also suffers from intense, poorly managed use and current deficit water withdrawals.  The Chad Basin treaty (1964), among Cameroon, Niger, Nigeria, and Chad, deals with economic development inside the basin, the lake's tributaries, and industrial uses of the lake, but lacks allocations.  The agreement does create a commission, which, among other things, arbitrates disputes concerning implementation of the treaty. The commission prepares general regulations, coordinates the research activities of the four states, examines their development schemes, makes recommendations, and maintains contact among the four states.

Principal Focus:

Most treaties focus on hydropower and water supplies: fifty-seven (39%) treaties discuss hydroelectric generation, and fifty-three (37%) distribute water for consumption. Nine (6%) mention industrial uses, six (4%) navigation, and six (4%) primarily discuss pollution. Thirteen of the 145 (9%) focus on flood control. The Database includes one treaty which primarily discusses fishing (less than 1%) (included in the Database for other elements).
Monitoring (where mentioned; some treaty sources are from condensed works):

Seventy-eight (54%) treaties have provisions for monitoring, while sixty-seven (46%) do not. When monitoring is mentioned, it is addressed in detail, often including provisions for data-sharing, surveying, and schedules for collecting data.

Information-sharing generally engenders good will and can provide confidence-building between co-riparians. Unfortunately, some states classify river flows as secrets,  and others use the lack of mutually acceptable data as a stalling technique in their negotiations.  Most monitoring clauses contain only the most rudimentary elements, perhaps due to the time and labor costs of gathering data.

Data collected by signatories of the treaty can provide a solid base for later discussions. India and Bangladesh previously could not agree on the accuracy of each other's hydrologic records, but eventually agreed on Ganges flow data and based a workable agreement on that data in 1977.  The cooperation between engineers or among council members of different nations can result in the formation of an epistemic community, another positive outcome of data gathering and sharing.  Treaties do not yet include provisions to monitor compliance, but such additions may bolster trust and increase the strength of these epistemic bonds.

Method for Water Division:

Few treaties allocate water: clearly defined allocations account for fifty-four (37%) of the agreements. Of that number, fifteen (28%) specify equal portions, and thirty-nine (72%) provide a specific means of allocations. All but three multilateral agreements lack definite allotments, although a few establish advisory and governing bodies among states.

There are four general trends in those treaties which specify allocations: (1) A shift in position often occurs during negotiations from "rights-based" criteria (whether hydrographical or chronological) in favor of "needs-based" values (based on irrigable land or population, for example); (2) In the inherent disputes between upstream and downstream riparians over existing and future uses, the needs of the downstream riparian are more often delineated (agreements mention upstream needs only in boundary waters accords in humid regions), and existing uses, when mentioned, are always protected; (3) Economic benefits are not explicitly used in allocating water, although economic principles have helped guide definitions of "beneficial" uses and have suggested "baskets" of benefits, including both water and non-water resources, for positive-sum solutions; and (4) The uniqueness of each basin is repeatedly suggested, both implicitly and explicitly, in the treaty texts.

This last point is exemplified in the unique treaty elements devised by negotiators. The 1959 Nile Waters Treaty divides the average flow based on existing uses, then evenly divides any future supplies (projected from the Aswan High Dam and the Jonglei Canal Project).  The Johnston negotiations led to allocations between Jordan riparians based on the irrigable land within the watershed; each party could then do what it wished with its allocations, including divert it out-of-basin.  The Boundary Waters Agreement, negotiated with a hydropower focus between Canada and the United States, which allows a greater minimum flow of the Niagara River over the famous falls during summer daylight hours, when tourism is at its peak.


Fifty-seven of the treaties (39%) focus on hydropower. Power-generating facilities bring development, and hydropower provides a cheap source of electricity to spur developing economies. Some, however suggest that the age of building dams will soon end, because of lack of funding for large dams, a general lack of suitable new dam sites, and environmental concerns.

Not surprisingly, mountainous nations at the headwaters of the world's rivers are signatories to the bulk of the hydropower agreements. Nepal alone, with an estimated two percent of the world's hydropower potential,  has four treaties with India (the Kosi River agreements, 1954, 1966, 1978, and the Gandak Power Project, 1959) to exploit the huge power potential in the region.


Only three agreements deal with groundwater supply: the 1910 convention between Great Britain and the Sultan of Abdali,  and the 1994 Jordan-Israeli  and 1995 Palestinian-Israeli  agreements. Treaties that focus on pollution usually mention groundwater, but do not quantitatively address the issue.

The complexities of groundwater law have been described by more than a few authors.  Overpumping can destroy cropland through salinity problems, either by seawater intrusion or evaporation-deposition, and therefore the allocation of too much water (or one party's overpumping) can decimate future freshwater supplies.

The Bellagio Draft Treaty, developed in 1989, attempts to provide a legal framework for groundwater negotiations.  The Draft  requires joint management of shared aquifers and describes principles based on mutual respect, good neighborliness, and reciprocity.  While the Draft recognizes that obtaining groundwater data can prove difficult and expensive,  and mutually acceptable information relies on cooperative and reciprocal negotiations, it does provide a useful framework for future groundwater diplomacy.

Non-Water Linkages:

Negotiators may facilitate success of a treaty by enlarging the scope of water disputes to include non-water issues. If pollution causes trouble in a downstream country, an upstream neighbor may opt to pay for a treatment plant in lieu of reduced inputs or reduced withdrawals. In such a case, lesser amounts of high-quality water may improve relations more than a greater quantity of polluted or marginal-quality water. Such tactics "enlarge the pie" of available water and other resources in a basin. Non-water linkages include capital (44, or 30%); land (6, or 4%); and political concessions (2, or 1%). Other linkages account for ten of the 145 treaties (7%), and there are no linkages for eighty-three treaties (57%).

Examples of these linkages can be found in the 1929 Nile agreement, in which the British agreed to give technical support to both Sudan and Egypt.  In another example, the Soviet Union agreed, in lieu of payments, to compensate lost power generation to Finland in perpetuity (the 1972 Vuoksa agreement).  Britain even established ferry service across newly-widened parts of the Hathmatee in India, in compensation for the inaccessibility problems created by a dam project in the late 1800s.

Compensation for land flooded by dam projects is common. For example, British colonies usually agreed to pay for water delivery and reservoir upkeep, and the British government agreed to pay for flood damage to houses.  However, capital can provide compensation for a greater array of treaty externalities and requirements, such as the construction of new water facilities; the India-Nepal Kosi River Project Agreements, signed in 1954 and 1966 provide two examples.

Treaties which allocate water also include payments for water; forty-four treaties (30%) include monetary transfers or future payments. As early as 1925, Britain moved toward equitable use of the rivers in its colonies: Sudan agreed to pay a portion of the income generated by new irrigation projects to Eritrea, since the Gash river flowed through that state as well.  Treaties also recognize the need to compensate for hydropower losses and irrigation losses due to reservoir storage: the 1951 Finland-Norway treaty  and the 1952 Egypt-Uganda treaty  both include such compensation. Again, these agreements emphasize the monetary aspect of water; they do not describe water as a right.


Treaties may handle disputes with technical commissions, basin commissions, or via government officials. Fifty-two (36%) of the treaties provide for an advisory council or conflict-addressing body within the parties' governments. Fourteen (10%) refer disputes to a third party or the United Nations. Thirty-two (22%) make no provisions for dispute resolution, and forty-seven (32%) of the texts are either incomplete or uncertain as to the creation of dispute resolution mechanisms. Can a technical advisory body address disputes? Perhaps, but the treaties do not expressly provide for such activity.

Historically, force or the threat of force has ensured that a water treaty will be followed (e.g. British colonial treaties and the 1947 Allied peace treaty with Italy ), but power is less desirable and more expensive as a guarantor of compliance than mutual agreement. Britain could oversee its colonial water treaties because it had one of the more powerful administrative and military organizations in the world. Similarly, agreements on the Nile generally favor Egypt, while those on the Jordan River favor Israel because of their respective power.

While the conflict resolution mechanisms in these treaties do not generally show tremendous sophistication, new enforcement possibilities exist with new monitoring technology. It is now possible to manage a watershed in real time, using a combination of remote sensing and radio-operated control systems. In fact, the next major step in treaty development may well be mutually enforceable provisions, based in part on this technology of objective and highly detailed remote images, better chemical testing, and more accurate flow computations than previously available.


The 145 treaties which govern the world's international watersheds, and the international law on which they are based, are in their respective infancies. More than half of these treaties include no monitoring provisions whatsoever and, perhaps as a consequence, two-thirds do not delineate specific allocations and four-fifths have no enforcement mechanism. Moreover, those treaties which do allocate specific quantities, allocate a fixed amount to all riparian states but one-that one state must then accept the balance of the river flow, regardless of fluctuations.

One problem hampering the development of sophisticated water treaties may have been the difficulty in acquiring information on similar settings. Thus far, each set of negotiators has had to, in effect, independently invent solutions. However, with the compilation of treaties in a single, searchable collection, along with negotiation notes and case studies, the Transboundary Freshwater Dispute Database hopes to provide researchers and diplomats a useful tool to assess negotiating trends and workable treaty solutions in the future.



  1. Acknowledgments: The authors are indebted to those agencies which have helped fund different aspects of the Database, including the US Institute of Peace, the World Bank, the US Agency for International Development, Pacific Northwest National Labs, the Alabama Water Resources Institute, and the University of Alabama.
  2. Address for correspondence: Dept. of Geosciences, Oregon State University, Wilkinson Hall 104, Corvallis, OR 97331-5506, USA. Tel: +1-541-737-2722.
  3. A. Wolf, J. Kinsler, J. Natharius, and J. Danielson, Transboundary Rivers of the World: An Updated Register. Unpublished Manuscript on file with the author.
  4. Thomas Naff, Conflict and Water Use in the Middle East, in WATER IN THE ARAB WORLD 273-74, 253-84 (Peter Rogers and Peter Lydon, eds., 1994).
  5. Aaron T. Wolf, "Water Wars" and Water Reality: Conflict And Cooperation along International Waterways (1997) (presented at the NATO Advanced Research Workshop on Environmental Change, Adaptation and Human Security, Budapest, Hungary).
  6. JErrold S. Cooper, Reconstructing History from Ancient Inscriptions: The Lagash-Umma Border Conflict (1983).
  8. See, e.g., Evan Vlachos, Prologue, WATER INTERNATIONAL, 15, 185-88 (1990); Ludvik A. Teclaff, Fiat or Custom: The Checkered Development of International Water Law, 31 NAT. RESOURCES J. 45 (1991); David and Joseph Eaton, Joint Management of Aquifers between the Jordan River Basin and the Mediterranean Sea by Israelis and Palestinians: An International Perspective, in PROCEEDINGS OF JOINT MANAGEMENT OF SHARED AQUIFERS: FIRST WORKSHOP, 131-52, (Eran Feitelson and Marwan Haddad, eds., 1994); DEBORAH HOUSEN-COURIEL, SOME EXAMPLES OF COOPERATION IN THE MANAGEMENT AND USE OF INTERNATIONAL WATER RESOURCES (1994); Joseph Dellapenna, Building International Water Management Institutions: The Role of Treaties and Other Legal Arrangements. in WATER IN THE MIDDLE EAST: LEGAL, POLITICAL, AND COMMERCIAL IMPLICATIONS, 55-89 (J. A. Allan and Chibli Mallat, eds., 1995); NURIT KLIOT, DEBORAH SHMUELI, AND URI SHAMIR, INSTITUTIONAL FRAMEWORKS FOR THE MANAGEMENT OF TRANSBOUNDARY WATER RESOURCES (1997).  In two exceptions, Dellapenna describes the evolution of treaty practice dating back to the mid-1800's, and Wescoat assesses historic trends of water treaties dating from 1648-1948 in a global perspective. In addition, McCaffrey offers theories about trends in treaty-making, specifically the move towards integrated management from unilateral development, the move away from navigation as the primary use, and the trend towards "equitable utilization." Hayton has argued that international law should better address hydrologic processes. See, respectively, Joseph Dellapenna, Treaties as Instruments for Managing Internationally-Shared Water Resources: Restricted Sovereignty vs. Community of Property, 26 CASE W. RES. J. INT'L L. ___ (1994); James L. Westcoat, Jr., Main Currents in Early Multilateral Water Treaties: A Historical-Geographic Perspective, 1648-1948, ___ COLO. J. INT'L ENVTL. L. & POL'Y ___ (1995); Stephen C. McCaffrey, The Evolution of the Law of International Watercourses, 45 AUS. J. PUB. & INT'L L. 87 (1993); ROBERT D. HAYTON, U. N. DEP'T OF TECHNICAL CO-OPERATION AND DEVELOPMENT, NATURAL RESOURCES WATER SERIES, NO. 20: RIVER AND LAKE BASIN DEVELOPMENT, 362 (1998) and Robert D. Hayton, Reflections on the Estuarian Zone, 31 NAT. RES. J. 123 (1991).
  9. The compilation was created by Jesse Hamner. We expect that both the full text of each treaty and the compilation of summaries will be up-loaded to the World Wide Web by the fall of 1998. In the meantime, an electronic version including one-page summaries of each treaty, is available on disk from either author on request.
  10. I. William Zartman, The Structure of Negotiation, in INTERNATIONAL NEGOTIATION, 65-77 (Victor Kremenyuk, ed., 1991).
  13. See BINGHAM, WOLF, AND WOHLGENANT supra note 16, at 111, 121.
  14. Agreement Between the Federal Republic of Germany, the Republic of Austria and the Swiss Confederation Regulating the Withdrawal of Water from lake Constance, April 30, 1966, Austria, Federal Republic of Germany, Switzerland, 620 UNTS 191.
  15. Unpublished manuscript , heads of states of Central Asia, 1993.
  16. Robert Rangeley, Bocar M. Thiam, Randolph A. Andersen, and Colin A. Lyle, INTERNATIONAL RIVER BASIN ORGANIZATIONS IN SUB-SAHARAN AFRICA, 10 (1994).
  17. Convention and Statutes Relating to the Development of the Chad Basin, May 22, 1964, Cameroon, Chad, Niger, Nigeria, TREATIES CONCERNING THE UTILIZATION OF INTERNATIONAL WATER COURSES FOR OTHER PURPOSES THAN NAVIGATION, 1984, at 8, U.N. Doc ST/ESA/141.
  18. Id. at 11.
  19. Sumit Ganguly, Personal Comm., 1997.
  20. Lincoln Kaye, The Wasted Waters., FAR EASTERN ECONOMIC REVIEW, Feb. 2 1989, at 17.
  21. Government of India, Ministry of Agriculture and Irrigation Central Water Commission, Agreement Between the Government of the Republic of India and the Government of the People's Republic of Bangladesh on Sharing of the Ganga Waters at Farakka and on Augmenting Its Flows, Nov. 5, 1977, Agreements on Development of Inter-State and International Rivers (1979) at 381.
  23. Aaron Wolf, Criteria for Water Allocations: The Heart of International Water Conflict, on file with the Journal.
  24. Agreement Between the Republic of the Sudan and the United Arab Republic for the Full Utilization of the Nile Waters, Nov. 8 1959, 453 UNTS 51.
  26. Treaty Between the United States of America and Canada Relating to the Uses of the Waters of the Niagara River, Feb. 27, 1950, 132 UNTS 224.
  27. See e.g. SANDRA POSTEL, THE LAST OASIS 40-43 (1992).
  28. Manisha Aryal, Dams: The Vocabulary of Protest, HIMALAYAN MAGAZINE, July-August 1995, at 16.
  29. Terms of a Convention Regarding the Water Supply of Aden Between Great Britain and the Sultan of Abdali, 148 Consolidated Treaty Series 70.
  30. Treaty of Peace Between the State of Israel and the Hashemite Kingdom of Jordan, Done at Arava/Araba Crossing Point, Oct. 26, 1994, Ministry of Foreign Affairs of Israel.
  31. Israeli-Palestinian Interim Agreement Annex III, Art. 40, Sep. 28, 1995, Ministry of Foreign Affairs of Israel.
  32. See for example Robert Hayton, The Law of International Aquifers, 22 NAT. RESOURCES J. 71 (1982); Albert Utton, The Development of International Groundwater Law, 22 NAT. RESOURCES J. 95 (1982).
  34. Hayton supra note 36 and Utton supra note 36.
  35. Robert Hayton and Albert Utton. Transboundary Groundwaters: the Bellagio Draft Treaty, 29 NAT. RESOURCES J. (1989).
  36. Although costly, gathering accurate groundwater data to help determine safe yield for wells is probably not as expensive in human terms as ruined cropland and saline water supplies.
  37. Exchange of notes between His Majesty's Government in the United Kingdom and the Egyptian Government in Regard to the Use of the River Nile for Irrigation Purposes, May 7 1929, United Kingdom Foreign Policy Documents No. 4, 1978, at 24.
  38. Agreement between the Government of the Republic of Finland and the Government of the Union of Soviet Socialist Republics Concerning the Production of Electric Power in the Part of the Vuoksi River Bounded by the Imatra and Svetogorsk Hydroelectric Stations, July 12, 1972, 884 UNTS 57.
  39. Articles of Agreement Between the Edur Durbar and the British Government, July 20 1874, Consolidated Treaty Series 148 at 70.
  40. See supra note 43.
  41. See Government of India, Ministry of Agriculture and Irrigation Central Water Commission, Agreement Between the Government of India and the Government of Nepal on the Kosi Project, Apr. 25, 1954, Agreements on Development of Inter-State and International Rivers (1979) at 355, 369.
  42. Notes Exchanged between the United Kingdom and Italy Respecting the Regulation of the Utilisation of the Waters of the River Gash, June 15 1925, United Kingdom Foreign Policy Documents No. 4, 1978, at 14.
  43. Agreement Between the Governments of Finland and Norway on the Transfer from the Course of the Näätämo (Neiden) River to the Garsjöen, Kjerringvatn, and Förstevannene Lakes, Apr 25, 1951, Legislative Texts and Treaty Provisions Concerning the Utilization of International Rivers for Other Purposes than Navigation,  United Nations Document ST/LEG/SER.B/12 (1963), at 609.
  44. Exchange of Notes Constituting an Agreement Between the Government of the United Kingdom of Great Britain and Northern Ireland and the Government of Egypt regarding the construction of the Owen Falls Dam in Uganda, July 16 1952, 207 UNTS 278.
  45. Treaty of Peace with Italy, Feb 10 1947, 49 UNTS 1.
  46. Broad accessibility to the Database is encouraged.  The Database is not copyrighted (although due credit is appreciated), and, as mentioned above, should be available in its entirety on the World Wide Web by the end of 1998. Any comments or suggestions for future work are welcome. We would also appreciate knowing about any omissions or errors.

  47. This summary assessment is the first result from what we hope will be continued systematic study of the treaties we have collected in our Database. In other work, Wolf describes generally the process of water disputes and dispute resolution, and specifically how the Database treaties allocate water. See, respectively, Aaron T. Wolf, Interntational Water Conflict Resolution: Lessons From Comprartive Analysis, 13 INT'L J. WATER RES. DEV. In future study, we hope to assess the relationship between the substance of the treaties and other geographic variables, particularly climate, power relationships, types of government, and changes over time. We would also like to assess mechanisms for conflict resolution for their relative effectiveness.
Collaborating Partners