Chapter One: Some Dimensions of the Problem

Chapter Two: The Past as Prologue

Chapter Three: Towards a Theory of Water Conflict

Chapter Four: Some Doctrines of International Water Law

Chapter Five: In The Shadow of the Future: Three Case Studies

Chapter Six: Conclusions and Policy Recommendations

Bibliography

Notes

Introduction

Freshwater is a basic, finite and fundamental resource, integral to life. But this valuable resource is extremely scarce and it is being wasted and abused globally.

The total volume of the earth's water is about 1.5 billion cubic kilometres, but about 97.3 percent of this total is salt water and only 2.7 percent is fresh. Freshwater comprises a total volume of 35 million cubic kilometres. About 77.2 percent of the freshwater is frozen in the ice caps and glaciers in Antarctica, the Arctic Circle and Greenland. Of the remaining 22.8 percent of freshwater, 22.4 percent is found in deep underground acquifers. Thus, only 0.36 percent is readily available for human use in lakes and rivers (Gleick, 1993; Housen-Couriel, 1994; Shiklomanov, 1994).

In Southern Africa, the water inventory is severely limited, unevenly distributed and subject to extreme fluctuations which are generally caused by climate changes. In most countries of Southern Africa -- with the exception of Angola and, to a limited extent, Zambia -- freshwater is scarce. In fact, physical symptoms of water stress are not only evident in water-scarce countries alone, but in water-wealthy countries as well. These symptoms include declining water tables due to chronic droughts and over pumping of underground water and dry season scarcity of water. The recurring droughts of the 1980s -- in 1981-82, 82-83, 83-84, 86-87, followed by a very severe drought in 1991-92 and another equally devastating drought in 1994-95 -- have undermined economic growth and governmental and societal efforts to improve the quality of life in the region.

One hydrologist, Falkenmark (1989), has created a measure of water scarcity. Her five water competition levels -- (1) 'adequate', (2) 'quality and dry season problems' (3) 'water stress' (4) 'absolute scarcity' and (5) 'beyond the water barrier' -- are determined in terms of the number of people who depend on a supply of 1 million cubic metres of water per year. Falkenmark suggests that countries with 10,000 cubic metres per person per year or more, have limited water problems; that general water problems occur in states with 1670 - 10,000 cubic metres per person per year; chronic water scarcity obtains in states with supplies 500 - 1000 per person per year.1

Falkenmark's levels of measurement may be controversial, but they nevertheless serve to provide a valuable water resource availability picture. In Southern Africa, that picture is one of insecurity. As this paper demonstrates, by 2025, according to projections by Falkenmark (1991) and some social scientists (Ohlsson, 1995), Lesotho and South Africa will be facing 'absolute water scarcity', Malawi will be beyond the 'water barrier', while Zimbabwe, Tanzania and Mozambique will be experiencing 'water stress'.2

Focusing on Southern Africa, this paper will argue three things: (1) that there will be increased competition for access to scarce water resources at the community, national and regional levels in an effort by states to achieve water security; (2) that colonisation and decolonisation disputes have raised water issues to a combustible level of conflict in Southern Africa; and (3) that water security and access to fresh water have become a 'high politics' issue increasingly perceived as part of the struggle for democratic empowerment in the region.

(1) Increased competition.

Conflicts can be either induced or aggravated by resource scarcity and the competition for access or possession that this triggers. Additionally, as argued by Homer-Dixon, scarcity can sharply increase '. . . . demands on key institution such as the state while it simultaneously reduces their capacity to meet those demands. These pressures increase the chance that the state will either fragment or become more authoritarian . . . .'3 The disintegration of Somalia and the savage killings that took place in Rwanda in 1993-94 are both reflective of the state's reaction to environmental scarcities (drought-induced hunger and starvation in Somalia and population-induced land shortage in Rwanda) confronting the two societies and the threats to its integrity.

(2) Colonisation and decolonisation.

The paper also argues that colonisation and decolonisation disputes provided a 'protracted conflict setting',4 thus raising water and river basin issues to a combustible level of conflict in Southern Africa. Water was heavily implicated in several historical colonial efforts: Germany's colonisation of South-West Africa (Namibia), particularly the push to have access to the Zambezi; Portugal's lusotropicalism,5 particularly the strategic imperatives behind the construction of the Cahora Bassa scheme in Mozambique and other river basins collaboration works with South Africa; and apartheid's survival and sustenance in South Africa and the imperative to maintain a cordon sanitaire around the white redoubt.

Superpower rivalry was insinuated into Southern Africa through an ingenious interpretation of the significance of Southern Africa, by both South Africa and Portugal, to the Western World. The West was made to believe that it had strategic interests in Southern Africa, that those interests were being threatened by twin forces of communism and African nationalism and that Portugal and South Africa were in the trenches in defending those interests. Their odious domestic politics should thus be ignored in the name of Western security interests in the region.

(3) 'High politics' and competitive exclusion

The paper further argues that at the local community level the struggle for access to resources (particularly water and land in both rural and urban areas) is perceived as part of the struggle for a democratic dispensation and empowerment. This is particularly so in South Africa and Zimbabwe, where resource denial and dispossession or capture are most acute. Access to resources is considered as one of the quantitative and/or qualitative indices for determining the content of democracy.

Southern Africa will therefore continue to be a veritable arena for potential conflicts over participation in obtaining access to (i.e. in political, economic and social aspects of national developmental efforts) and distribution of resources. Conflict potential over resource distribution lies in the fact that war termination arrangements in some countries -- Zimbabwe (the Lancaster House arrangements); South Africa's interim constitutional provisions; the Namibian settlement which also led to the adoption of the policies of national reconciliation -- froze legal , economic and institutional arrangements that had historically disadvantaged blacks. Thus, the policies of national reconciliation are now being misinterpreted by whites to suggest that the status quo in relation to economic arrangements and resource capture (which is sustained by institutional and legal provisions relating to property rights) should remain broadly undisturbed.

The paper examines three cases, the Zambezi, Orange and Okavango basins as units of analysis to investigate/interrogate conflict potential over water, the levels and scales of any such conflict and the conflict mitigation measures that could be instituted. Other international water basins including the Pungwe and Sabi basins involving Zimbabwe and Mozambique, the inter basin water transfer schemes involving South Africa, Swaziland and Mozambique and South Africa's domestic inter-basin transfer schemes are mentioned with the view to teasing out the water conflict potential in those areas. While conflict mitigation will not be discussed extensively, the paper will examine the role of international law and international water agreements as possible instruments to enhance co-operation and regime creation. Regime creation is itself a conflict generating exercise as some such regimes are negotiated 'in the shadow of the future' and within the context of a historically conditioned geo-political, geo-economic and geo-strategic framework.

Chapter One: Some Dimensions of the Problem

Water has become a central'high politics' issue, particularly in South Africa where over the last fifty years the development of national and shared water resources was undertaken on the basis of strategic intent. The factors that shaped the political economy of Southern Africa, particularly mineral wealth in South Africa and to a limited extent in Zimbabwe, have also profiled the broad contours of resource exploitation (including the exploitation of shared river systems) and helped to define and shape the legal and institutional framework that largely determined the issues of access to resources.

This has also created the perception of historically disadvantaged states and communities, thus raising concerns about the distribution of gains in water reforms, management and regime creation in certain basins within the region. Southern Africa is a historically conditioned geo-strategic, geo-political and geo-economic framework against which future cooperative efforts on water resource will be negotiated. There is considerable conflict potential in the process of negotiating new water sharing regimes among the states of the region, especially the issue of non-riparian rights over international rivers. Four significant factors drive this strategic intent: climate, population growth, development and historical realities.

Climate

Various factors, including latitude, altitude, topography and oceanic influences, affect the climatic make-up of Southern Africa. The eastern and northern parts of the sub-continent experience humid sub-tropical conditions while the west is dry, desert-like. Hydrological extremes are thus common with, for example, the eastern coastal area of South Africa around Durban receiving more than 1070mm of rain annually while Port Norloth, on the same latitude as Durban but on the western coast of South Africa, receives only 60mm of rain per year. Rainfall reliability, variability and evenness in distribution, follow a similar pattern. Variability in Namibia, for example, is as high as 80 percent in the west and decreases to less than 20 percent in the north east Caprivi Strip.

The ratio of run-off to total precipitation is very low in the region, at about 15 percent. That is, 85 percent of the rain evaporates again under intense heat. This ratio is even lower than the average for some individual countries, 10 percent in South Africa, for instance. In Namibia, it is estimated that on average 83 percent of the total rainfall evaporates shortly after falling. Of the remaining 17 percent, 14 percent is lost through evapotranspiration, 1 percent recharges groundwater and only 2 percent remains to be harvested behind dams.6 In the Okavango Delta, about 150km3 of water flows into the delta annually (i.e. five times the annual consumption of the whole Southern African region) and 95 percent evaporates.

In the 1980s and 1990s drought conditions have had a negative impact on communities and regional economies. These conditions also emphasise the point that water is now one of the region's most limiting natural resources. Writing about the 1994/95 agricultural season in Zimbabwe, for example, Dale noted that '. . . it was the eighth lowest on record at national level with 418.1mm of rain recorded against 335.2mm in 1991/92', the latter widely considered as the lowest season on record. He noted significantly about the 1990s, again using Zimbabwe as an example, that '. . . the five year running average of annual rainfall for Zimbabwe . . . was for the period 1990/91 - 94/95 inclusive, the lowest on record', underscoring the severity of precipitation decline.7

The Intergovernmental Panel on Climatic Change (IPCC) has predicted that drought may lower food production in Sub-Saharan Africa by as much as a 20 percent. The Southern Africa Development Community (SADC) recently observed that during the 1991/92 drought more than eighteen million people in the region required food assistance in addition to seed packs and fertiliser to prepare for the 1993 season. The 1994/95 agricultural season was equally disastrous with cereal harvest standing at 15.73 million tonnes. Direct consumption needs in the region are estimated at 23.34 million tonnes with an additional 3.73 million tonnes necessary to rebuild depleted stocks, leaving a deficit of 7.1 million tonnes of which 4.92 million tonnes is maize, the major staple for black consumers. The food deficit must be met through imports, food donations and aid, but at the costs of scarce foreign currency.

Population

In 1994, the SADC population was estimated at 136 million and, with a growth rate of 3 percent, it is expected to more than double to 298.5 million by 2025. For most of the states in the region, the population is expected to reach a growth rate of 1 percent annually after 2025 due to population policy measures, starting with South Africa in 2020. The burgeoning population is going to sharpen sectoral competition for water and also competition between states for dwindling water resources.

Per capita water availability is expected to fall sharply within the next decade. Indications are that by the year 2025 no fewer than 1.1 billion Africans, that is about 66 percent of Africa's projected population, will be living in water stressed countries. These figures must raise considerable alarm given the severe mismatch in Africa between population sizes and the level of technological development, which could empower a nation to respond adequately to a water scarcity crisis.

Development

Water demand by industries, particularly those enterprises in the region that were built on the now flawed assumption of sufficient water availability, remains high. Clarke (1991) offers the following picture to underscore the significant contribution of water as an input in industrial production: '. . . you need at least three litres of water to produce a tin of vegetables, 100 litres to produce one kilogramme of paper, 4,500 litres to produce one tonne of cement; 4.3 tonnes to manufacture one tonne of steel, 50 tonnes to manufacture one tonne of leather and no less than 2,700 tonnes to make one tonne of worsted suiting . . .'8 The industries reflected above are the mainstay of SADC economies underscoring the serious threat to industrial viability posed by growing water scarcity.

Shortages of water have major economic consequences in Southern Africa. Reductions of water flow on hydro electric power generation (HEP), agriculture and general economic performance in Southern Africa has been significant and in some respects devastating. Dale (1994) indicated that '. . . since 1980/81, the average annual energy generation capability of Kariba and Kafue have each been reduced by approximately 40 percent below their maximum practical levels and 30 percent below long term average flow levels . . .' 9 Climatic and hydrological trends suggest the urgent need to establish a new agreed hydrology which will be the basis for future plans on developing the HEP potential of the Zambezi and other rivers. No accurate figures are available to determine the impact of reduced energy supplies to both the Zimbabwe and Zambian economies. Dale (1994) speculates that energy shortage had the potential to shrink the economy of Zimbabwe by about 10 percent if it had continued. Both the governments of Zambia and Zimbabwe have jointly appealed to their publics to conserve electricity because the situation at the Kariba and Kafue stations, which are operated conjunctively, is very critical due to continuing drought conditions.

Traditional methods of water provision, with engineering achievements demonstrative of the frontier spirit of conquest and water development, are now, for a variety of reasons, unable to provide sustainable solutions to the mismatch between water demand and supply. Additionally, most readily accessible and easily exploitable dam sites, located nearer to water markets and within the territorial space of states are being, or have been developed. This implies that possible future water projects, (like the Matabeleland Zambezi Water Project in Zimbabwe) are likely to be more expensive, take longer to negotiate before work is started, and may be controversial and need extensive environmental impact assessments (EIA).

They will also require the explicit consent and co-operation of donor institutions, mostly the World Bank, and inter state agreements, if the river in question is a shared water course. In addition to social, political and institutional constraints, such water sources may not be exploited economically and may not yield as much as earlier projects. The recent phenomena of many empty dams in the region - most major dams in Zimbabwe in September 1995 were less than 15 percent full, with water availability for most major cities and towns being measured in terms of a few months - are a testimony to the diminished return on investment and the unsustainability of sinking scarce and meagre financial resources,( in most cases externally sourced funds,) in risky but essential projects. Persitant drought conditions and reduced run off, plus increased water demand contribute to this scenario.

Historical Realities

The notion of disadvantaged communities suggests that water resource shortages may not be a reflection of resource shortages per se but of a dramatisation of competitive exclusion or resource capture. In Southern Africa, one current feature of water provision is inequality with regards to access to resources and facilities. Development of water resources in Zimbabwe, Namibia and South Africa, for example, was primarily driven by the need to support these countries' wealthy economic sectors and as an instrument to encourage white colonial settlement than satisfying the needs of all communities.

As mentioned, about 55 percent of the population in South Africa does not have access to safe water.10 The major consequence of competitive exclusion is the comparatively poor access to piped water South African blacks 'enjoy'. Only 43.3 percent of black South Africans have access to piped water as compared to 95.4 percent for coloureds, 100 percent for Indians and 99.9 percent for whites. This translates to about 17.3 million blacks in South Africa alone having no access to piped water, and 150,000 coloureds and only 5000 whites.11

There is also a marked difference in the average daily per capita consumption of water between urban and rural consumers, largely due to inequality of access. The Department of Water Affairs in Namibia suggests that average water consumption per person per day in urban areas is 330 litres while the figure for rural consumers is 85 litres. Per capital daily consumption shrinks even more dramatically where water has to be carried manually over long distances. The World Resources Institute (1992) estimates the average figure at 21 litres per day. It should be noted that about 66 percent of the region's population lives in rural areas.

Simon Forster, Special Adviser to South African Water Minister, Kader Asmal, recently estimated that '. . . more than 65 percent of all water currently used in South Africa is either privately owned or used under historically obtained riparian rights'. He further observed that '. . . when viewed in conjunction with expansion in irrigation that has taken place during the last two decades, the conclusion can be made that a relatively small number of land owners now control the greater portion of the nation's utilisable water.'12 Forster admitted that government-owned water was just a tiny proportion of the total national resource. He suggested that the solution to South Africa's water problems (i.e. the question of access) also lay in a 'thorough overhaul of the country's water law to allow the state to regain control of the bulk of water resources and allocate them in a planned and equitable way.'13

Nor is South Africa the only country in Southern Africa, where competitive exclusion has resulted in the existence of historically disadvantaged communities with limited access to vital resources such as water and land. In Zimbabwe, the Water Act of 1956 enshrined water rights into the country's water legislation. Water rights are granted permanently to a user of water, thus permanently excluding new needs and interests. The majority of water rights holders in Zimbabwe are white. Additionally, major river systems where irrigation plays a significant role in agriculture (e.g. the Manyame, Mazowe, Mutirikwi) are fully righted. This legislation should be viewed against the background of growing water scarcity and water insecurity. Communal insecurity induced by resource insecurity has the enormous potential to engender political instability, national insecurity and conflicts.

Competitive exclusion is not only evident at national or sub national levels, but also obtains at inter-state levels in the unilateral measures by some upstream states to over-draw international water courses, thus enormously disadvantaging downstream states. For example, Mozambique's agriculturalists were becoming increasingly militant toward South Africa because of perceptions in Maputo that the upstream state was over-utilising shared resources. Most of Mozambique's significant rivers originate out of the country and are heavily abstracted in South Africa, Swaziland and Zimbabwe before entering Mozambique. Mozambicans argued that 'a number of South African dams on the Lesutu and Inkomati rivers in Mupumalanga province and the increased agricultural use of the Olifants and Crocodile rivers' aggravated water scarcity situation in Southern Mozambique.14 This diminished flow of rivers into the sea means that sea water may penetrate as much as 100 km upstream creating considerable ecological problems for Mozambique. Consequently, Mozambicans have been considering a variety of options, including a 'World Court lawsuit against South Africa for allegedly violating international water conventions'.15

Water Supply in Southern Africa

There is likely to be a growing trade in water within Southern Africa as water scarce countries and regions import water from resource rich states and/or basins. The Lesotho Highland Water Project, where Lesotho envisages selling huge quantities of water to South Africa, is an indication of things to come.

The project is currently the biggest and most complicated engineering endeavour of its kind in the world. When completed in about 2020, some five dams, water transfer works and 200km of tunnels will have been constructed between the two countries. A hydro-electricity generating capacity of abut 180 MW is also planned. About 2,200 million cubic metres per annum of water will be transferred from Lesotho to the South African network destined for the Gauteng Province. The cost of the project, at 1987 prices, is estimated to be R8 billion (about U.S. $1.8 billion). It is increasingly uncertain whether the project is going to continue along the lines envisaged in the 1987 treaty. Indications suggest that total water availability in the Lesotho system may have been over-estimated while water demands in Lesotho and downstream settlements in the Orange basin were either never considered or grossly under-estimated.

The largest inter basin transfer schemes are the planned transfer of water from the Zambezi to Zimbabwe's city of Bulawayo; from the Zambezi to the Botswana Capital, Gaborone; from the Zambezi to Namibia and also from the Zambezi to the Gauteng province of South Africa. The proposed transfers emphasise a dramatic swing from non-consumptive (largely hydro-electricity power generation) to consumptive use of the waters of the Zambezi. They are also challenging at national, sub-regional and regional levels because of a number of considerations, including:

• whether hydro-electricity generation is the only requirement worthy of priority;

• the need to formulate intra- and inter-basin transfer policies, particularly when one of the envisaged recipients (South Africa in the case of the Zambezi) is not a riparian state. Does a country have to be truly a riparian to have a right of extraction?

• criteria for sharing water, including who decides allocation or use of the water, on the basis of what information and for what period of time. Should quotas to countries be based on their perceived needs, their present and future populations, their frontage to the Zambezi (or any river under consideration) or contribution to total mean annual run off;

• how to reduce (not eliminate) conflict over water;

• the Zambezi's agreed hydrology as a determinant of the levels of abstractions from the river.

Transfer schemes are unlikely to be a panacea to water scarcity in Southern Africa, because there is likely to be conflict between donor basins and receiving basins as resource use rises and scarcity threatens supplies in the donor basins. Such threats of conflict and to continued water flow offers a weak basis for sustainable development. Some water transfer schemes have already significantly disadvantaged downstream states leading to demands for the revision of agreements that form the basis of such transfers or stream diversions.

The substantial return flows from these transfers, and from the Lesotho Highland Water Project when it comes on stream in 1996, are eventually into the Limpopo system.

Water Demand in Southern Africa

Water demand is driven by a number of factors including population growth and the perceived need for food self-sufficiency in individual countries. Growth in water demand levels exceed population growth rates in a number of countries in the region. Agriculture is by far the biggest consumer of water, taking over 70 percent of water consumed in the region. The demand for water for irrigation exceeds supply in Botswana, Namibia, South Africa and Zimbabwe. Consequently, the perceived need to strive for food self-sufficiency completely undermines the search for water security in individual states and at a regional scale.

Agriculture's thirsty appetite for water is largely unsustainable in a number of countries already facing looming water scarcity, suggesting the necessity to restructure the economies of these states. Economic restructuring is a long term objective that, however, needs to be embarked on quickly. Botswana's National Water Master Plan (1991) indicates that the country has already taken the lead in economic restructuring by opting for food security rather than self-sufficiency, a decision largely informed by the stark realisation that the country was severely water scarce to go all out for irrigated agriculture. South Africa, which has 10 percent of the water resources in the region, but consumes about 80 percent of regional water resources, of which 57 percent is by the agriculture sector, needs to urgently consider the merits of economic restructuring to diminish significantly the amount of water consumed by the agricultural sector. The principle of food self-sufficiency, is inappropriate for an economy as sophisticated as South Africa's and is a relic of the apartheid era when self reliance was a national security objective.

Water management problems in South Africa are also exacerbated by outdated water legislation. The South African Water Act 1956, is now widely considered as inadequate and needing revision to allow the state to regain control of the bulk of water resources and allocate them in a planned and equitable way. The Act recognises "private water", "riparian rights" and "public water', provisions that do not recognise the hydrological cycle as a continuum and water as a national asset that must be shared equitably.16 Experts in the Department of Water Affairs acknowledge that riparian owners were technically and legally capable of pumping many rivers dry, particularly during low flow periods. It is estimated that 65 percent of all water currently used in South Africa is either privately owned or used under historically obtained riparian rights. There is no guarantee that the water is being used economically and efficiently. It is also estimated that water losses associated with supply of irrigation water are around 30 percent. A one percent reduction to 29 percent would meet the basic water needs of nine million people.17 The imperative to review the entire body of South African Water legislation, should be designed to deal with issues of justice, fairness, equity and sustainable development.

Water is the most valuable input to production in most, if not all Southern African economies. However, value added to products differs significantly between agriculture and industry. Clarke (1991) cites a study carried out in California in 1980 which estimated that '... every cubic kilometre of water used to water crops, produced an added value of US $5 million, while the same cubic kilometre used by industry produced an added value of US $5 billion - more than sixty-five times as much. . .'18 The low marginal economic productivity supplied by agriculture strongly suggests that countries with stronger and broader economic bases like South Africa and, to a limited extent Zimbabwe, could start reallocating considerable quantities of water to industrial and domestic consumption. Additionally, an efficiency campaign could be coordinated within Southern Africa to ensure that farmers irrigate efficiently, adjust crops, cropping patterns and general agricultural operations to aim for lower water consumption. Farmers in the region, should be conditioned to expect permanent reductions in water supply within Southern Africa and in their sector in particular. Irrigated agriculture cannot be subsidised to the extent where the insatiable demands it makes on regional and national water resources endangers a feeble fiscus and to levels which are obviously unsustainable.

Water will be a significant issue determining the overall flavour of bilateral relations between Mozambique and South Africa. The signing of a formal treaty between the two countries establishing a joint water commission is nonetheless unlikely to bring any dramatic gains to either South Africa or Mozambique. According to Peter Van Niekerk, South Africa's Chief Engineer for planning, the joint water commission is laying the framework for 10-15 years from now which may come too late to fully address the threatening water scarcity problem.

go to top

Chapter Two: The Past as Prologue

Two truths are told as happy prologues

To the swelling act, of the imperial theme

Macbeth Act 1 Scene III

Between 1945 and 1990, the Southern African security complex was shaped by the collision of two antagonistic forces and processes: the global decolonisation process (and African nationalism in Southern Africa); and efforts to consolidate white hegemony in the region.19 The collision between these processes offered the protracted conflict setting against which the sub-text of water conflict was enacted. Three distinct but equally intermeshing strands in colonial and settler attempts to consolidate white control and hegemony in Southern Africa are identifiable

(1) Portugal's 'lusotropicalism',20 dramatised by Lisbon's 1951 redesignation of its colonies as overseas provinces, sought to integrate them into the Portuguese nation. It is significant that the timing in the changing of the imperial lexicon coincided with the start/genesis of the global decolonisation process and the rise of African nationalism.

(2) The establishment, at the instigation of the United Kingdom, of the Federation of the Rhodesias and Nyasaland (or Central African Federation) 1953 - 1963 to counter the capture of the South African State by Afrikaners in 1948, with the consolidation of Afrikaner power south of the Limpopo. The dissolution of the federation in 1963 was followed by independence of Northern Rhodesia, which became Zambia, and Nyasaland, which became Malawi. Southern Rhodesia became Rhodesia following Ian Smith's Unilateral Declaration of Independence (UDI) in 1965.

(3) The third and most significant strand was the triumph of Afrikaner nationalism as evidenced by the coming to power of the National Party in 1948. The electoral triumph was swiftly followed by the introduction of apartheid as a policy framework designed not only to institutionalise racism but also to consolidate the Afrikanerís capturing and control of the South African State. As already observed, the three strands and developments were manifest as early as the incipient stages of the decolonisation process, when African nationalism was still weak. The organisation and development of African nationalism in the 1950s to the 1980s, was in response to the challenges offered by efforts to consolidate white hegemony in the region.

South Africa emerged as the core hostility state in Southern Africa, and indeed the object of opposition by the Organisation of African Unity (OAU). The protracted anti-apartheid conflicts that emerged between 1950 and 1990 was not conducive to inter-state collaboration to develop elaborate schemes on shared river basins, no matter how urgent. Water agreements reached during this period were largely specific to the task, limited in scope and sometimes coerced with no conflict resolution potential. That is, they had no implications for the end of the larger political conflict then raging in Southern Africa.

Additionally, measures taken to consolidate colonial and settler control, the armed struggle that ensued to dislodge colonial and settler hegemony, conflict in some neighbouring states (induced by South Africa's policy for destabilisation) firmly implicated water as a terrain security issue (Naff 1993); the question of physical control of water resources was essential to the processes. The physical location of water development projects also caused conflicts over location and that conflict reflected 'the essentially contestedness' of the colonial state.

As Naff (1993) observes: '. . . water causes conflict usually as a contributing factor submerged within the context of other contentious issues among parties already embroiled. . . .' Naff also argues that '. . . resource and environmental issues and situations can be related in conflict in at least three ways: as proximate causes, as the means of conflict and as the rationalisation of conflict.'21

The Berlin conference of 1884 forced Portugal to abandon her claims along the coast line of Africa. Two distinct characteristics marked the essence of Portuguese colonialism: a very tenuous and thinly spread hold on the territory coupled with violence and exploitation .The 1951 redesignation of colonies as overseas provinces marked a new phase in Portugal's colonial presence in Africa that was likely to witness a determined economic and infrastructural development of the colonies. Henriksen (1978) identifies three goals that Portugal hoped to realise by developing her last two colonies , Angola and in particular Mozambique. These were 'to realise the golden dream of obtaining great wealth from Africa ,to justify continued retention of the colonies ,and to settle Portuguese immigrants so as to relieve unemployment in Portugal and to increase settler presence to sustain the hold on territories in the teeth of burgeoning nationalism on the African continent.'22

Between 1950 and 1960, Portuguese immigration to Mozambique grew from about 60,000 to 100,000. The white Portuguese settler population doubled from 100,000 to 200,000 between 1960 and 1970 ,a decade when armed opposition to colonial presence in the territory commenced and spread very significantly. Portugal targeted the Limpopo and the Zambezi rivers and basins for infrastructural development to entice immigrants to settle in these areas .The idea was to have settlers working in and developing the two major river basins and as the backbone for the military defence of the territory. In 1950, Portugal completed a dam on the lower Limpopo which was designed to irrigate about 100,000 hectares.

It was, however, with the construction of the giant Cahora Bassa dam and hydro-electric power project on the Zambezi, 1968 to 1973 , that Portugal intended to capture the essence and symbolism of a determined and continued imperial presence in Africa. The dam also focused contest (both political and military) between the colonial /imperial presence and the liberation of Mozambique (FRELIMO) in military opposition to the project. Portugal saw the Cahora Bassa in strategic and economic terms. The dam was constructed on the Zambezi river in the Tete province of Mozambique . The province is rich in coal ,iron ore and other mineral deposits. The dam had the capacity to generate 3,600,000 kilowatts of electricity to be used by the mining sector in the Tete province with the excess to be sold to South Africa . In fact the power generated by the project was envisaged to satisfy 8 percent of South Africa's total power requirements. Portugal was reportedly envisaging settling one million Portuguese immigrants in the Zambezi valley after the completion of the dam. The proposal for the dam also envisaged water borne transport on the Zambezi for 260 miles from the coast to the dam wall and for another 190 miles upstream to the border with Zambia and Zimbabwe.23

In military strategic terms, Portugal saw the filling of the dam as a way to help hamper guerrilla infiltration and operations south of the Zambezi. FRELIMO fighters were equally determined 'to bust the dam'; they saw the project as symbolic of Portugal's determination to stay in Mozambique. FRELIMO rapidly opened up a military front in the Tete province to oppose the Cahora Bassa project. The dam project also led to FRELIMO reassessing its military tactics. Hitherto, guerrillas had desisted from attacking any economic infrastructure which they hoped to inherit after Portugal's departure, but from 1968 onwards any infrastructure perceived to be enhancing Portuguese imperial presence in the territory became a military target. This included water projects seen as conceived by Portugal as instruments of settler occupation. Equally, guerrillas had desisted from attacking Portuguese settlers in Mozambique arguing that they were at war with Portuguese Colonialism and not the Portuguese people. Portugal's determined push to settle immigrants who then would be considered as being the backbone of colonial defence led to such settlers being considered as legitimate targets.

After independence the dam infrastructure also became targets of South Africa's military destabilisation campaign in Mozambique, particularly between 1983 and 1990. Electricity generation had been out since 1985 as a result of military action by the Mozambique Resistance Movement (RENAMO) and by 1988 about one third of the pylons along the 3000 km power lines to South Africa had been sabotaged. There was another wave of sabotage of the power lines two months following the signing of an agreement among Portugal, South Africa and Mozambique to co-operate in repairing the lower lines. The cost of repairs is now estimated at more than US$ 50 million.

Did the riparian dispute in Mozambique cause the war? Certainly not; the combatants were already embroiled. The riparian dispute in the Zambezi valley, however, did serve to escalate the conflict as Portugal demonstrated her determination to use the project as a means of conflict. The change in FRELIMO strategy after 1968 implicates water in its rationalisation of the conflict. Equally, black states within the region interpreted the dam in terms of the rationalisation of the conflict. Portugal's allies in the region, Rhodesia and South Africa, saw the dam as symbolic of Portugal's determination to stay in Southern Africa. Mozambique under Portugal could offer Rhodesia the necessary defence shield, and South Africa a cordon sanitaire against African nationalism and communism.

It is within the context of water as a terrain security issue where Portugal's water infrastructural development and FRELIMO's armed struggle strategies after 1968, plus RENAMO's sabotaging of development infrastructure including water projects, become significant and pertinent to this paper. Portugal saw water infrastructural development in the Limpopo and the Zambezi in strategic terms, that is in terms of terrain control, thus militarising both public discourse and developmental efforts. FRELIMO's extension of its military campaign into the Zambezi Valley with the dam structure as the principal military target is equally an issue of terrain control at the local level (Zambezi valley) and national level, the latter aspect to be accomplished with the dislodgement of Portugal from Mozambique and the attainment of independence under FRELIMO. Water as a terrain security issue is equally implicated in RENAMO's bloody and systematic destruction of infrastructure in rural Mozambique, including the power lines from Cahora Bassa to South Africa. This destruction was seen as demonstrative of the capacity of the RENAMO war machine to shrink the ruling FRELIMO's writ in Mozambique.

The second observable trend towards the consolidation of colonial and settler control in Southern Africa was the British government's establishment of the federation of the Rhodesias and Nyasaland in 1953. In 1938, a Royal Commission had recommended the creation of a body to co-ordinate common services among the three adjoining territories. The resurgence of South Africa in the region after World War II added impetus to British intentions to establish the federation as a counter weight to Afrikaners capturing the state in South Africa. Southern Rhodesia was the linchpin of the federal system. Individual governments of the three territories and the federal government in Salisbury, encouraged and offered incentives for new immigrants to settle in the Federation. Southern Rhodesia in particular attracted more immigrants. There were significant developments in the industrial, mining and agricultural sectors of the federal economy leading to increased energy demand in Southern Rhodesia and Northern Rhodesia's Copperbelt.

Two possible sites on which to locate hydro-electricity generating plant were debated in the two territories of Southern and Northern Rhodesia. One site was on the Kafue River wholly on Northern Rhodesia territory, while the other possible site (Kariba) was on the Zambezi, a shared watercourse. Additionally, growing economic activities in the federation introduced the notion (into the debate) that hydro electricity sites were going to be developed not for the purposes of meeting relatively small deficiencies of power at the end of the 1950s, but to change the whole character of power supply within the federation by making hydro rather than thermal the base power for the era.

Debate in the 1950s clearly favoured the Kariba site on the Zambezi with its much larger proven capacity. In 1955, the World Bank concluded that 'on technical and economic grounds, the choice of Kariba was fully justified'. But that debate generated at least five distinct themes:

(1) The considerable opposition by politicians from Northern Rhodesia to the development of Kariba ahead of the Kafue site clearly implicated water as a terrain security issue. Where was the project to be located? Who would have control? Hydrological data on the Kafue river was still considerably less accurate than hydrological data on the Zambezi from which to make a thoroughly informed choice.

(2) Another theme is the question of the balanced achievement of gains when two riparian states co-operate to develop a shared water course. This study will return to the issue for a detailed analysis when we examine the Zambezi case.

(3) The collapse of the federation with Zambia and Malawi attaining independence and the Rhodesia Front under Smith unilaterally declaring independence (UDI) for Rhodesia, served to develop a protracted conflict setting against which the issues to manage the assets of the Central African Power Corporation (CAPCO)24 were handled. The United Nations imposed mandatory economic sanctions against Rhodesia which Zambia observed scrupulously. There was therefore in political terms a situation of non-recognition and acute hostility as the overall framework between Zambia and Rhodesia. There was also very limited contact between Rhodesia and Zambia on development and management of the power utility on the Zambezi. This tends to support observations by Lowi (1993) that it is sometimes difficult to delink issues when a riparian dispute coexists with a protracted political dispute. Extended contact between the two countries would have suggested either recognition or acceptance of the radical otherness UDI wanted to consolidate in Rhodesia.

(4) The federation which had Rhodesia as its linchpin created the perception of historically disadvantaged states, as both Northern Rhodesia and Nyasaland argued that the federal system hugely favoured Southern Rhodesia. The theme recurs incessantly in discussions between Zambia and Zimbabwe over sharing/division of unitary held assets, particularly the assets of the now defunct Central African Power Corporation. The federal government largely benefited whites more than it benefited blacks, thus sustaining the perception of historically disadvantaged communities.

(5) The large white immigrant population that settled in Rhodesia were the major beneficiaries of resource development and allocation embarked on by the government to satisfy the existing settler community and to entice new immigrants. The theme of competitive exclusion from access to resources including water is implicated in these political developments. The CAPCO regime which governed the management of the Kariba system and the Cahora Bassa regime in Mozambique have developed and remained separate notwithstanding the fact that they are drawing from the same basin and same watercourse.

When the National Party (NP) won the 1948 elections in South Africa, it swiftly introduced the policy of apartheid. The policy of racial supremacy encapsulated in the apartheid ideology was at odds with trends toward decolonialism and with the mood borne out of the experience of combating Hitler's Germany. As apartheid was bitterly contested both at home and abroad, the security of both the regime and the white redoubt was a major preoccupation of government. Thus, from 1948, according to Price (1984) '. . . what ultimately lay behind all Pretoria foreign policy efforts, was the desire to achieve, if not acceptance of, at least international acquiescence in white minority rule within the Republic. . .'25 Threats to white supremacy and apartheid (with apartheid just one form of that supremacy) were seen to come from three external sources. These were global decolonisation, Communism, and the liberation movement, i.e. African nationalism. The liberation movement was seen as an external threat particularly after 1961 when the African National Congress and the Pan Africanist Congress were banned and driven underground and into exile. After 1960, particularly following the Shapesville massacre and South Africa's voluntary withdrawal from the Commonwealth, economic sanctions became another serious external threat as South Africa increasingly depended on the importation of foreign capital.

South Africa conceived its defence perimeter as extending to the Zambezi and the whole Southern African region as its sphere of influence. It sought to be allowed to absorb/incorporate the High Commission States of Basutoland, Bechuanaland and Swaziland, which Britain rejected. South Africa also extended support to other white ruled states in the region, that is Rhodesia and Portugalís presence in Mozambique and Angola. In the 1970's South Africa also embarked on the policy of d'Ètente which was essentially an attempt to win friendship and construct peaceful coexistence between black ruled Africa in general and black ruled Southern Africa in particular. Pretoria hoped to use its economic power to co-opt newly independent states in the region to accept South Africa's legitimacy.

Superpowers were increrasingly drawn into the fray. In the United States, National State Security Memorandum 39 dated 1969, had concluded that whites were in Southern Africa to stay and that economic efforts should be used to moderate the consequences of white rule. Following the collapse of lusotropicalism (1974) and Smith's UDI in 1980, Soviet and Cuban military power was projected into the region. South Africa's defence perimeter had shrunk dramatically. Pretoria launched a vicious destabilisation strategy as an instrument to fight the perceived threat to White supremacy in South Africa and to cause black states in the region into non -aggression pacts or into accepting Pretoria ís legitimacy. Water was intimately implicated in the high politics encapsulated in the total strategy designed to respond to conceived and perceived external threats.

Firstly, Pretoria entered into specific agreements with Portugal as the colonial power controlling Mozambique and Angola for the water development project on the Cunene river between Angola and Namibia in 1969, and on the Zambezi. While the Cunene projects (Hydro electric generation and irrigation ) were also designed to respond to the question of water scarcity in Namibia , they equally were part of the high politics designed to strengthen lusotropicalism .The Cahora Bassa scheme was equally designed to supply energy to South Africa while strengthening lusotropicalism in Mozambique. The objectives of lusotropicalism and apartheid were thus congruent as they were designed to sustain control over White ruled territory in the region.

Secondly, South Africa, like Rhodesia and Portugal's lusotropicalism, suffered the experience of non-recognition and acute hostility. South Africa thus found black ruled states in the region reluctant to enter into elaborate agreements over water issues except under coercion. The 1987 treaty between Lesotho and South Africa on the development of the scheme to export large quantities of water from Lesotho to South Africa, was essentially a function of coercive military action by the hegemonic power in the basin .The agreement followed a South African engineered military coup in Lesotho which overthrew Chief Leabua Jonathan. Jonathan's successor Major General Justin Lekhanya signed an agreement /treaty with South Africa ending over 30 years of negotiation over the issue. The military action by the hegemonic power in the basin was engendered by the issue of threatening water scarcity facing the Johannesburg-Pretoria (Gauteng province) area of South Africa suggesting that water was directly implicated in the causes of the conflict between the two countries. But as all the water that flows out of Lesotho eventually passes through South Africa, what was the point of forcing a coup as a route map to an agreement? Perhaps the answer lies in the fact that South Africa's superior power resources do not coincide with a superior riparian position in the basin. This is specifically in relation to the expected location of the project in Lesotho which would have some impact on yield of the cost and efficiency of the project. South Africa is downstream to Lesotho in the Orange basin and locating the project in South Africa would have meant increased cost of construction and operation.

Thirdly , South Africa sought to use regional water sharing arrangements (and also bilateral arrangements )as a means to acquiring bilateral security treaties with some states in the region. The protracted conflict setting , however , effectively undermined those hopes with those states that had riparian agreements between themselves and South Africa not allowing them to function. That is, there were hardly any meetings between South Africa, Angola and Mozambique on agreements on the Cunene and Cahora Bassa. South Africa could thus allow RENAMO to sabotage the power lines from Tete to South Africa to bring the FRELIMO government on its knees towards coerced co-operation. Even when such agreements are signed without coercion, like the Limpopo basin technical committee involving Botswana, South Africa, Mozambique, and Zimbabwe, they were highly delimited, highly specific and highly technical arrangements and had no conflict resolution implication26. The Limpopo committee has met only once since 1985. This suggests, as the Rhodesian case revealed, that it is difficult to delink issues when a riparian dispute coexists with a protracted political dispute. Co-operation over a riparian development project is not viewed as a straight forward and unambiguous issue.

But is it prudent to talk of the potential for inter-state conflict over water in Southern Africa today, when the Southern African impulse suggests inclination towards regional co-operation? For example, the signing in 1995 of a SADC protocol on shared watercourse systems in the region could be signalled as an eloquent statement demonstrative of the co-operative impulse in Southern Africa. But is a protracted conflict setting the only requirement to either induce or aggravate conflict over renewable resources , including water? Do conflicts over renewable resource evolve over time? What actually evolves to determine the dimensions and flavour of resources conflict in a basin? The next section will attempt to interrogate these issues with the view to the future dynamics of hydropolitics in Southern Africa. But first a theoretical exposition on security and water conflict is pertinent.

go to top

0: Towards a Theory of Water Conflict

Academic discourse and interest in water as a potential source of conflict have concentrated largely on the Middle East and North Africa. This region is marked by combustible hydropoltics over such shared water resources as the Jordan and Nile Rivers. (See Rogers and Lydon, 1993, Gleick 1993, Beschorner 1992, Starr 1991, Waterbury 1989, Lowi 1993, etc.). For example, King Hussein of Jordan, one of the Arab leaders most willing to make peace with Israel, has stated that water is the only issue which could cause Jordan to go to war again. Egypt's foreign minister has also warned Ethiopia and Sudan about potential dams along the Upper Nile. Significantly that ominous statement suggests not only Israel as a possible opponent, but also other Arab and African states as potential adversaries over water.

Predictions of future water wars (Starr 1991) are based on the growing competition for an increasingly scarce resource. They are also premised on projections that all states in the Maghreb and the Middle East are already facing absolute water scarcity and high population growth rates, urbanisation and the absence of fresh sources to augment supplies only underscore a bleak future.

In terms of such global hydropolitics, Southern Africa has essentially remained at the edges of international concern. Attempts by some hydrologists (Falkenmark, 1989) to warn the world about Africa's threatening water scarcity do not appear to have provoked widespread interest, which is somewhat curious given the fact that environmental scarcity is heavily implicated in collapse of the state system and the humanitarian nightmare that followed in Rwanda and Somalia.

In a multi-ethnic state, competition for access to limited resources is also likely to be viewed in terms of ethnic security and survival. Even where ethnicity may not be overtly or immediately implicated, competition between consuming sectors for a resource which is shrinking could lead to disadvantaged stakeholders and other interested parties perceiving and undertaking access in terms of group security. For example, the Zimbabwe Farmers Union (ZFU) representing small-scale, communal and resettlement farmers (wholly black), threatened in 1995 to break the law pertaining to water rights that denied them access to water -- a denial that threatened their survival. The ZFU clearly sees access as a matter of group security for small-holder agriculturists.27

As discussed earlier, Southern Africa is already experiencing critical water scarcity. Falkenmark (1993) suggests five 'competition intervals' for water with categories defined in terms of the number of people who depend on a supply of 1 million cubic metres of water per year. She argues that countries that have 10,000 cubic metres per person per year or more have limited water problems; general problems over water occur in states with 1670 - 10,000 cubic metres per person per year; 'chronic water scarcity' obtains in states with supplies of between 500 - 1000 cubic metres per person per year and 'beyond the water barrier' status occurs in states with less than 500 cubic metres per person per year. Other hydrologist (Shuval, 1992) proposes different figures to describe water scarcity levels. Shuval suggests that states with over 2000 cubic metres per person per year should be considered as water abundant while those with less than 500 cubic metres per person per year are water stressed. These figures though contested offer concrete basis for quantitative measurements of scarcity and help to indicate the water resource vulnerability of states.

According to the World Resource Institute (1992), a critical stage is reached when a country uses more than one third of its total renewable supply. Falkenmark suggests that many countries in Southern Africa will be in this category by 2025. Using Falkenmarkís competition intervals to determine scarcity, the following picture emerges for Southern African countries by 2025. Malawi will be beyond the water barrier; Lesotho and South Africa will be facing absolute water scarcity while Mozambique, Tanzania and Zimbabwe will be experiencing water stress. Angola, Namibia, Botswana, Swaziland and Zambia will be experiencing only quality and dry season water problems. But national averages sometimes mask chronic water scarcity that is usually experienced at local levels, this suggests that even resource rich states could be facing water resource vulnerability.

Gleick argues that water resource vulnerability '. . . is a function of many things including economic and political conditions, water availability and the extent to which a source is shared. . . .28 Water sharing reflects the degree of vulnerability to competing interests. For example, about 94 percent of Botswana's water resources are shared by other neighbouring states. The Zambezi, the largest water course in the region, is shared by eight states with South Africa, although a non-riparian, also planning to abstract water from the Zambezi by the first quarter of the next century. How and by what political agreements and technologies remains to be seen. As discussed, other quantitative indicators of water resource vulnerability suggested by Gleick are the population growth rate (average for the region is 3 percent), and high dependence on hydro-electric power as a percentage of total electricity supply. About 99.6 percent of the electricity consumed in Zambia, 97.6 percent in Malawi and 69.8 percent in Tanzania is from hydro-electricity stations. This source is particularly vulnerable to effects of climatic change on run off and increase in population leading to increases in consumption of water, thus reducing water availability for power generation.

Gleick also observes that because of water scarcity '. . . some countries could eventually reach an absolute limit on the type and extent of development' adding that this has the potential to 'cause tensions between water poor and water rich states.'29 Naff (1993) argues that water is intrinsically a security issue because nations cannot survive without enough water. Water scarcity, therefore,

' . . . especially [as a] mismanaged scarcity, combined with maladministration, contributes significantly to the creation of an environment of uncertainty and instability in the basic political, economic and social institutions of society, most destructively in situations where the reciprocal factors of ecological marginality and rising poverty obtain.'30 Scarcity, poverty and environmental decay induce rural to rural migration or rural to urban migration. Such migration could induce identity conflicts especially in a multi-ethnic state. Homer-Dixon asserts that '. . . there is substantial evidence to support the hypothesis that environmental scarcity causes large population movement, which in turn causes group identity conflicts. . .'31

Water problems and their relationship with security may need to be viewed with other variables like population, culture, values, total nature of relations between states, national and regional value systems. Naff suggests that 'no single theory or model has as yet been developed that can deal with layered political, socio-economic, legal and strategic entanglements of freshwater that underlie hydro politics. . .'32 Therefore, any attempt to categorise water as a conflict issue 'must employ a multi-dimensional typology or a combination of typologies'. But there are also certain distinguishing characteristics water displays as a security issue in international relations which are pertinent to a theory of water conflict. These characteristics include the fact that water as a commodity and water as an issue are very complex problems; there is no substitute for water. Secondly, water as a terrain security issue compels all concerned to seek control over the product/commodity or terrain on or through which it passes. Who has rights to water? We see in Southern Africa that South Africa, as a large and powerful economic actor, believes that it has such rights and is reaching across international borders and other states to obtain water and meet its water needs. While water may not implicated as a direct or overt cause of conflict, it is, however, a contributory factor '. . . submerging within the context of other contentious issues among parties already embroiled. . .'33 Thus, a protracted conflict setting in a shared watercourse or basin, a political system that has lost legitimacy, a social dislocation induced by environmental scarcity could offer the necessary backcloth to induce or aggravate water conflict.

Naff also suggests a power matrix model which could be used to determine the potential for water conflict. The causal agents for a power matrix model are ' interest or need including the perceptions and motivations of the actors, the riparian position of the actors and projectable/defensive power.'34 States may opt to co-operate or enter into conflict if their needs are seen as either being helped or frustrated by other riparian states. A riparian position gives advantages to an upstream state over downstream competitors. Upstream states can determine, through a variety of actions, the quantity and quality of water that can be available to downstream states. Potential conflict over water is heightened when the downstream riparian is more powerful but faces resource scarcity and perceives its needs as either frustrated or likely to be frustrated by the actions ( or non-actions) of upstream states (as in the case of Egypt with Sudan and Ethiopia, upstream states on the Nile). '. . . [T]he key determinants of water-driven conflicts/strife are scarcity and maladministration, followed closely by perceived need and the relative power status of the international basinís riparians. . .'35 The truth of this assertion will be tested in the discussion of the three cases to be considered below.

When a particular riparian issue concerns two (or more) states, how can we explain that one state may consider it as a source of conflict, while others may not? The next section will interrogate this question and the concerns encapsulated in it, in our detailed discussion of cases. Classical conflict theory views conflict as a form of social interaction, a struggle over claims to limited resources (including power) in which opponents strive to eliminate rivals or injure them. Conflict in this case is different from competition. Zartman (1994) argues that conflict has many layers and at one level it can be seen as a '. . . contest of parties trying to impose a unilateral solution to a problem. . .'36 At another level conflict could be reflective of 'shifts in elements of old order or regimes. . .' The collapse of an old regime and the construction of a new regime or negotiations to establish such a regime could be accompanied by conflict. Equally there are different scales and levels of conflict and the analysis will attest to the elasticity of the concept of conflict by recognising the various but relevant types of hydro -conflicts and the different levels and scales of such conflict.

go to top

Chapter Four: Some Doctrines of International Water Law

What do we mean by rights to water? This section will broadly review five doctrines of international water law that have been formulated by publicists who have studied the problems of the utilisation by states of shared water resources. These doctrines attempt to delineate the rights of riparian or basin states to utilise waters from a international water course.

The doctrines are :

(a) the doctrine of absolute territorial sovereignty;

(b) the doctrine /theory of absolute territorial integrity;

(c) the doctrine /theory of limited territorial sovereignty;

(d) the theory /doctrine of limited territorial integrity;

(e) the doctrine /theory of community of interest in the waters of a international water course.

Some of the problems the doctrines have attempted to address are :

• how to distribute water resources between basin states especially in regions of scarcity;

• different prioritisation of water uses by basin states;

• prior appropriation /existing use 'natural,' 'historic' vested or 'ancient rights' claimed by hydraulic civilisations, particularly, versus new water's needs and interests;

• compensation to some basin states for damages to their interests.

Various factors influence the pattern of claims by basin states to the use of shared water resources. Dellapena asserts that the '. . . patterns of international claims and counter claims, diverge sharply according to the riparian status of the state making the claims. . .'37 Upstream states usually make claims based on the doctrine of absolute territorial sovereignty, while downstream states usually (invariably support the doctrine of absolute integrity of the watercourse. For example, Egypt has traditionally supported the doctrine of absolute integrity of the Nile while the upstream states, particularly Ethiopia, support the doctrine that gives them absolute territorial sovereignty over the waters of the Nile. The implications of these divergent views is hydropolitics that are conflictual. Different levels of economic development in basin states and the differences in bargaining power have also contributed in part to competing positions taken by basin states on resource sharing. As eloquently argued by Spulber and Sabbaghi '. . . indeed the demand for water, power and related resources are not identical at all points of a basin. Due to differences in rate of regional development, demand may be stationary in one part of the basin while it is growing in another part. The result is an intensified struggle for the apportionment of the waters of the axial river and often political pressures for water transfer from proximate and even far off

basins. . .'38

Again the example of Southern Africa is illustrative. The statement is true of the economic development levels and water demand patterns in the Zambezi basin. The countries to the South of the basin are relatively more developed economically than those on the Northern basin. Also the Southern basin states face more water scarcity problems than those on the Northern basin. These factors are likely to make the process of regime negotiations more acrimonious and long. Basin or parts of some basins may be characterised by exceptional political strife thus providing a protracted conflict setting against which water issues are going to be discussed and negotiated.

International Lawyers and hydrologists are agreed on the international drainage basin as the critical unit of analysis and management for international water resources. The Helsinki Rules (1966) Article 11, define the drainage basin as '. . . a geographical area extending over two or more states determined by the watershed limits of the system of waters including surface and under -ground waters flowing into common terminus. . .' The Helsinki Rules likewise define a basin state as '. . . a state the territory of which includes a portion of the international drainage basin . . . .' However, the legal provision may not cover all aspects of the defining characteristic of an international drainage basin (or even a micro basin ) particularly as it would relate to joint planning processes since the international drainage basin does not always conform geographically to other planning basins or resource basins.39

The Doctrine of Absolute Territorial Sovereignty

This is sometimes known as the Harmon doctrine. The doctrine regards a watercourse as being subject to the exclusive sovereignty of a riparian state through which it flows, as it relates specifically to that portion that flows through its territory. Some jurists regard the doctrine as anarchic and not reflective of a position of international law. They assert that the US Attorney General Harmon may have been expressing an opinion reflecting the absence of international law on the subject then (1895). Some countries, Ethiopia for one, actually support this doctrine.

The Doctrine of Absolute Territorial Integrity

The doctrine essentially asserts that states should not interfere with the natural flow of a water passing through their territory in a manner that will impact on the behaviour of the flow downstream. The principle is viewed with disfavour by modern juridical opinion.

The Doctrine of Limited Territorial Sovereignty

The gist of this doctrine is that international rivers cannot be the subject of the exclusive appropriation by one state. Thus it restricts the principles of absolute integrity and absolute sovereignty to the extent that ensures all riparian states reasonable utilisation of the waters of a shared course some jurists believe that the doctrine forms the basis of the equitable apportionment principles now widely acknowledged (but interpreted differently depending on the part and interest of affected states in a particular basin) as the basis of international water law. The differences over this doctrine arise from the fact that the principle of equitable apportionmentís, precise formulation remains contentious. Additionally, conflicts arise from such issues as determining the hierarchy of uses of the waters of a shared river; prioritising between prior and new needs or uses and even agreeing on the criteria to determine who gets water and how much water.

The Doctrine of Community of Interest

This doctrine attempts to advance and improve on that of limited territorial sovereignty by suggesting that all basin states have a common interest in developing the basin. The doctrine therefore underpins the expectations of an integrated approach to the development and management of a shared international basin. The doctrine suggests equitable division and sharing of benefits.

The multiplicity of doctrines on the utilisation of shared international waters suggests that '. . . international law as an instrument of regulations on the trans-boundary freshwater issues is at present inconclusive and weak . . .'40 Naff attributes this problem to the absence of prior formal political agreements to govern 'the general specific terms of shared waters, together with essential international or inter -riparian institutions that ensure compliance among the users.'41 It is at this intersection between politics and law or absence of convergence, where the organic connection with the security and conflict setting (particularly as it relates to international watercourses) can be located.

go to top

Chapter Five: In The Shadow of the Future: Three Case Studies

The demise of apartheid, which had made South Africa the focus of international hostility, shifted water to an issue of high politics in regional relations. water became a resource issue around which domestic political agitation and inter-state relations are organised. By 1997, environmental issues and particularly water scarcity were confronting Southern Africa in a dramatic manner; scarcity was being driven by population growth, climatic change, resource degradation, distribution and accessibility problems. Each of the Southern African states is directly implicated in the dilemmas of environmental scarcity because '. . . scarcity can sharply increase demands on key institutions as such the state while it simultaneously reduces their capacity to meet those demands. . .' 42

Although Southern Africa demonstrates the impulse to co-operate, regime formation and the effectiveness of the regimes formed to deal with interstate co-operation will remain stunted by the overwhelming concern for sovereignty. National interest will inform foreign policy behaviour, thus suggesting that historically formed values and beliefs are bound to affect outcomes on regime negotiations. The 'shadow of the future' will affect negotiations towards the establishment of regimes in a number of ways including the fact that some states may 'remain fearful that an agreement for sharing resources might leave them much weaker at some later date.'43 This suggests that there is likely to be overwhelming concern, exacerbated by scarcities and the perception by historically disadvantaged states and communities, over the issue of the distribution of gains. Zambia and Angola's reluctance to sign the regional water protocol could be seen in this light.

Needs for water vary in any basin. The possibility is there for the least needy state (with nothing to lose) to refuse to engage seriously in negotiations to establish regimes, thus indirectly exercising a veto power in the basin. While the least needy state(s) would not view its/their behaviour as antagonistic, the needy would read the behaviour of the other state(s) as designed to frustrate their needs and therefore view relations as conflictual. As regimes are going to be negotiated within a historically conditioned framework, the possibility is there for some states within a basin to '. . . consider the river system primarily as a geopolitically strategic resource' that becomes an instrument of foreign policy.44

The issue of inter basin transfers, particularly in relation to the Zambezi, is likely to be problematic as many basin states have not even estimated their present and future demands. The issue for the Zambezi is further complicated by the fact that South Africa, a non-riparian state, is also interested in abstracting water from that system. A southern transfer that includes South Africa could be read by some basin states as further harnessing the political economy of the region into the South African orbit without encouraging symmetrical economic development of the region.

We next consider conflict at the sub-national, national, sub regional and regional levels using the case studies of the Zambezi, Okavango and Orange Rivers.

The Zambezi Basin

The Zambezi is the largest river in Africa that flows into the Indian Ocean. The river is about 3000 km long and drains a basin of nearly 1,300,000 square kilometres. The major tributaries of the Zambezi are the Kafue and Luangwa in Zambia, the Chobe in Botswana, the Sanyati and Manyame in Zimbabwe and Shire in Malawi. The major lakes formed behind dam infrastructures are Kariba and Cahora Bassa with surface areas of 5250 square kilometres and 2739 square kilometres, respectively. There are also some large dams on some of the major tributaries of the Zambezi like the Kafue and Manyame that reduce the run off from such tributaries to the Zambezi. Riparian states are currently engaged in unilateral exploitation of the resources of the basin thus eventually reducing the capabilities of other riparians.

The Zambezi basin lies roughly between latitudes 10 and 20 degrees south of the equator, an area dominated by the inter-tropical convergence zone as the major cause of precipitation. The wet season lasts from mid-November to end of March. Thus, the runoff of the Zambezi reflects this seasonality of the wet season.

The Zambezi drains eight of the twelve member states of the SADC. The countries, parts of whose territories lie in the Zambezi basin, are Angola, Botswana, Namibia, Malawi, Tanzania, Mozambique, Zambia and Zimbabwe. About 26 million people live in the basin. Three capital cities, Harare, Lusaka and Lilongwe, lie in the basin.

According to the IUCN (1992) Zambia and Angola contribute about half of the total annual run off of the Zambezi, Zimbabwe between 10 and 15%, Tanzania and Malawi contribute through the Shire river, Namibia/Botswana contribute to the run off only when the Okavango waters flow over the Selinda Spillway. Mozambique has the highest river flows as indicated by the readings at Cahora Bassa but has very little catchment.

Total run off of the river at the coast is estimated at 108 billion cubic metres annually. Runoff at Cahora Bassa in Mozambique is estimated at 88 billion cubic metres and at Victoria Falls, 42,20 billion cubic metres annually. The lowest seasonal flow average in November is estimated at 8,84 billion cubic metres per year. The Zambezi around Victoria Falls is the most likely area of possible water transfers to Zimbabwe, Botswana and South Africa.

Table 1 below shows the distribution of the Zambezi River catchment area between countries.

Country	Zambezi Basin Area (sq km)	% of Total Basin	% of Country in Basin
Angola	150,000	12	12
Malawi	120,000	9	98
Mozambique	140,000	11	18
Zambia	580,000	45	77
Zimbabwe	220,000	17	56
Botswana	80,000	6	1
Tanzania	N/A
Namibia	N/A
Total	1290,000	100

Source: Sir Mott McDonald and Partners: Hydrology of Southern Africa, Regional Report, 1990

The Zambezi has been predominantly utilised for hydro-electric power (HEP) generation at Kariba, Cahora Bassa and Victoria Falls (north bank). The Kafue is the only major tributary of the Zambezi with considerable HEP potential; part of the total 900 MW at Kafue gorge has already been exploited. Lower Kafue has potential of 450 MW. Only 36 percent of the Zambezi river's HEP potential has been developed. Potential yet to be developed and exploited includes the Batoka gorge (1600 MW) Devil's gorge 1240 MW which both lie between the Victoria Falls and Lake Kariba. There is an additional 300/600 MW potential at Kariba, 1000 MW at Mupata gorge (all on the Zambia - Zimbabwe border). In Mozambique, the exploitable potential includes 1200 MW at Cahora Bassa, 1600 MW at Mapanda Uncua, 444 at Boroma and 654 at Lupata. The central question confronting riparian states to the Zambezi is whether this hydro-electricity potential can be fully exploited and developed while also meeting other demands on the waters of that river.

It is noteworthy that even without any dramatic increase in the consumptive use of the waters of the Zambezi, power generation at Kariba and Kafue has been badly affected by the downward flow trends on the Zambezi evident since 1981/82 when Kariba last spilled. The IUCN (1992) and Dale (1994) have noted that average inflow of water into Lake Kariba that is available for power generation is only 62 percent of the equivalent of 30 years figures from 1961-1991, while net inflow since 1981/82 is only 52 percent of inflow during the preceding 20 years. Dale has estimated the annual energy generation at Kariba and Kafue to have 'each been reduced by approximately 40 percent below their maximum practical levels and 30 percent below long term average flow

levels. . . .'45 The implications of these developments include the requirement that the energy potential profiled above needs to be seriously reassessed in the light of new flow trends that have been evident since 1981/82.

It is not even certain as to when these new trends will settle down, suggesting that power planners will have to adopt the precautionary principle in their plans. In fact Magadza's (1991) mathematical models on possible climatic change in the region in general and the Zambezi in particular projects those changes at anything between 4 and 5 degrees Celsius over the next half century. Studies by Gleick have shown that a 2 degrees Celsius warming could lower runoff in California's Sacramento River basin by 22 percent, while a 4 degrees Celsius increase in temperature would lower runoff by as much as 62 percent. If climate change in Southern Africa would lower precipitation by say 20 percent as projected by mathematical modelling, the impact on resource availability for power generation in the Zambezi and other projected uses will be catastrophic. Kariba and Cahora Bassa power schemes were 'built to designs that expected river flows to contribute as they had done before impoundment'. Runoff was quite certain with cycles that were also predictable.

Another issue raised by the question of climate change is the uncertainty in planning and negotiations towards an equitable beneficial use of the Zambezi. Competition for the increasingly limited resource whose supply continues to diminish and is uncertain creates an incentive for states to pursue their national interests at the expense of other competitors. Potential for conflict increases in such circumstances.

Additionally, some states and other consumers are questioning whether hydro-electricity generation is the only requirement worthy of priority. These questions arise from the fact that under international water law, existing uses occupy preferential position over projected or anticipated uses. The existing use benefits primarily Zimbabwe and Zambia in the middle Zambezi, and Mozambique in the lower Zambezi. Mozambique has plans to export power to Malawi and Zimbabwe in addition to the existing grid that feeds South Africa. But an increase upstream in the demand for consumptive utilisation of the waters of the Zambezi and its tributaries will eat into the potential for expanded development of hydro-electricity generation in the lower Zambezi. A parallel use of Kariba, Kafue ( and possibly Batoka if the project went ahead) would still be unable to simultaneously satisfy the HEP demand and consumptive utilisation of the Zambezi.

As already indicated, more states in the region have plans to take water from the Zambezi early next century. In fact, Botswana, Zimbabwe, Zambia, Namibia and South Africa 'have operated on the assumption that the ultimate source of water for their requirements and developments is the Zambezi'.46 These plans suggest both inter-basin transfers and a possible extra riparian transfer.

The envisaged transfers (which are dominated by the transfer of the Zambezi waters to states south of the basin, and hence 'the southern transfer') would entail, inter alia, the following:

(1) Namibia plans to expand the LONRHO operated sugar irrigation project in eastern Caprivi from 40 to 140 hectares. The IUCN (1992) also suggests that Namibia has plans to channel Zambezi water from Katima Mlilo to Lake Liambezi to irrigate several thousand hectares of cane. Part of the extractions from the Zambezi may eventually find their way to the country's eastern carrier.

(2) Botswana envisages drawing from the from the Zambezi after 2020 to meet the expected demand for water in eastern Botswana and the Greater Gaborone region. Botswana has already indicated interest to join either Zimbabwe or South Africa in taking water from the Zambezi.47

(3) Zimbabwe proposes to draw water from the Zambezi to Bulawayo and from Kariba to Harare ( the latter after 2016 when alternative projects close to Harare will have been fully exploited). The Zambezi - Bulawayo extraction envisages abstracting the equivalent of 2 percent of the Zambezi flows during lowest flow or 0.25 percent of the mean annual flow which translates to 1.2 cumecs.48

(4) South Africa plans to draw from the Zambezi between 2.5 billion to 4 billion cubic metres annually after 2020 when the Lesotho Highland Water Project is fully developed. The possible extraction by South Africa of up to 4 billion cubic meters (the equivalent of 130 cubic metres/sec) would exceed the lowest recorded flow of the Zambezi at Victoria Falls ( of 120 cubic metres/sec in 1924) by 10 cubic metres/sec. This raises the question of whether or not there is enough water in the Zambezi after all for the simultaneous satisfaction of non-consumptive use ( hydro-electric generation ) and expected consumptive uses.

(5) Existing extractions for Zambia include 10,000 hectares under irrigation at Nkambala. Planned projects include the Kafue - Kariba irrigation and Mambova irrigation project which both plan to abstract about 9.28 cumecs49.

The inter-basin transfers profiled above, plus the in -basin abstractions from the Zambezi, do not include plans by basin states to build dams on the tributaries of the Zambezi within their own territories. Such structures will also diminish the amount of water flowing into the main river from tributaries, thus reducing the Zambezi which was profiled above. The construction of dams on the tributaries of the Zambezi plus the plans to abstract water from the Zambezi itself, are being executed unilaterally. This is largely because of the absence of an overall basin-wide regime to plan, co-ordinate and execute developments within the region. While a Southern African impulse of co-operation is evident, it is still seriously conditioned by a variety of factors which include the history of the region, the different levels of economic development and natural plus human resources endowments enjoyed by member states. Angola and Zambia, the least needy states as far as water is concerned, did not sign the regional protocol on shared water course systems produced by the SADC. Their decision to postpone the signing could perhaps be explained by the fact that both countries wanted further studies and consultations to determine the implications of the protocol to their undefined water needs. Angola has not enjoyed peace long enough to allow a determination to be made of its water needs in its part of the Zambezi basin. Zambia could equally worry that signing a regional protocol on the shared water systems significantly takes away freedom to manoeuvre in relation to the Zambezi. If it had viewed the Zambezi as a geopolitically strategic resource vis-a-vis other states in the basin, then signing the protocol significantly alters that perception because sovereignty was an important consideration in the draft protocol. Stanley consultants (1995) have observed that 'when the state signs and ratifies the draft protocol, the authority to make certain decisions concerning a river basin located within its territory will be decided by a river basin institution. The decision to develop a river basin area can no longer be a unilateral decision decided by the national government. This results in a relinquishing of national sovereignty to SADC and the joint decisions made by the organisations. . .' 50

It remains uncertain whether signatories to the protocol were aware of the implications to their sovereignty, as interpreted by Stanley Consultants, because the water protocol would then suggest a radical departure from the way the SADC defines co-operation. Previous SADC approaches to co-operation, including the SADC treaty, deliberately avoided construction of supra-national authorities to which member states ceded authority or sovereignty.

By far the most contentious issue on the Zambezi is likely to be how to share the beneficial use of the waters of the basin. Regime negotiation, which may take a long time, could confront the following challenges:

(a) An agreed hydrology for the Zambezi. This is likely to be contentious for a variety of reasons including the issue of compensation to those basin states likely to be disadvantaged by the inter basin transfer of water from the Zambezi. Water rights (for states) will also be influenced by the agreed hydrology which will have to be determined at various significant points. Abstraction points from the Zambezi would need to be collectively agreed upon so as to properly synchronise the various uses of the Zambezi. For example if South Africa would be allowed to withdraw between 2.5 billion and 4 billion cubic metres per annum from above Victoria Falls, this will be between 6 percent and 11 percent of mean annual flow at that point in the river. This will also translate to between 24 percent and 33 percent of mean annual low flow.51 The figure is higher if all planned withdrawals are upstream of Victoria Falls. Such abstractions are likely to impact negatively on the Victoria Falls itself, which is considered an international heritage. Additionally, the abstraction by South Africa alone is likely to have a negative impact on Zambia Electricity Supply Corporation's (ZESCO) Victoria Falls hydro-power station. The station requires 117 cumecs to keep it in production which is compatible with the 80 cumecs extracted during average flows. The IUCN has observed that the 'minimum recorded flow at the Victoria Falls of 142 cumecs' would keep 'either the South African project or the ZESCO power station supplied [but] not both' projects simultaneously. A sustainable yield at various points on the Zambezi needs to be worked out quickly to enable long range planning by those states interested in abstracting water from the river.

(b) Inter- and intra-basin transfer policies. The agreed hydrology will help in the formulation of inter- and intra-basin transfer policies, particularly when one of the recipients (South Africa in the case of the Zambezi ) is not a riparian state. In fact, the potential for non-riparian abstractions from the Zambezi offers the opportunity for general negotiations on non-riparian rights to water systems for all countries within the subcontinent. This will obviously be a very contentious issue which will need lengthy negotiations. Countries with access to the Zambezi will interpret any such negotiations as designed to offer South Africa, which possesses both the technical skills and the monetary resources (plus being the most needy state) access to the only river that is perceived to have the potential for such abstractions.

(c) The water-exporting basin considers itself as having priority over the commodity as opposed to the receiving or importing basins.

(d) Water rights and quotas. Criteria for sharing the water, including who decides allocation or use of the water, on the basis of what information and for what period of time. Should quotas or water rights to countries be on the basis of their perceived needs, their present and future populations, their frontage to the Zambezi (or any river system under consideration ) or contributions to total mean annual run off?

(e) Use reduction. How to reduce conflict over water.

Central to the critical issues above is the question concerning distribution of gains that should accrue from a co-operative regime on the Zambezi. Absence of such a regime suggests that gains for water rich states will remain absolute because they can unilaterally develop their section of the Zambezi without any regard for other basin states. They can equally exercise veto power over developments on the Zambezi by making it difficult if not impossible for other basin states to develop infrastructure on the Zambezi without their consent. Concern for gains does not only implicate the future of the basin, but also engages the historical experience of states in the basin --that is the history of co-operation between states. The perception by some states of having been historically disadvantaged in any co-operative relationship is thus likely to cause the regime negotiation process to be very complicated and even acrimonious. There are potential disagreements over such issues as water allocations and their destinations, conflicting views of rights, needs and international legal precedents on water rights.

Additionally, there are likely to be considerable technical problems arising from having a comprehensive basin wide regime for the Zambezi. The only unifying thread on the 3000 km long Zambezi is the hydrological potential of the river. Everything else, including land use patterns, population patterns, water demands etc., varies considerably on the macro-basin and its subsystems. Thus, having centrally determined, basin-wide measures is likely to implicate many issues, interests, planning basins and zones beyond the limits of the basin that will not have been consulted for the purposes.52 Equally, attitudes, values and interests beyond the basin's borders will impinge on the basin's plans and strategies creating perpetual tension.

Sub - Regional Conflict Potential : The Zambezi River Authority.

A conflict is evident in the Zambezi basin. The Zambezi River Authority (ZRA) is one of the limited, exclusive agreements that govern a section of the Zambezi. It involves only two of the eight states that lie in the basin. The lower Zambezi in Mozambique is governed by a different agreement that oversees the generation of power at Cahora Bassa. The ZRA is thus a point dramatising both a sub-regional impulse towards co-operation while also being the fundamental focus of considerable tension between the contracting parties -- Zambia and Zimbabwe. In fact, the tension sometimes implicates the core values of the two states to the extent where there is reluctance to co-operate on many bilateral issues beyond the technical matters of the ZRA, thus reflecting its 'high politics conflict' potential.

The ZRA was established by simultaneous Acts of Parliament in Zambia and Zimbabwe in 1987. The objective was to use Zambezi water on the section between two countries for power generation. The authority's functions included the maintenance of the Kariba complex and recommendations for new dams, recruitment of manpower and submission of development plans and programmes. These are essentially task-based arrangements. The ZRA can sue and be sued. The authority's agreement has conflict resolutions mechanisms. The ZRA was established at the dissolution of the Central African Power Corporation (CAPCO) itself preceded by the Federal Power Board. This historical profile suggests a conditioned framework.

As mentioned, the Kariba dam site was chosen despite opposition from Northern Rhodesian politicians who favoured the Kafue site and from blacks who saw the whole dam project as a part of the British strategic designs to strengthen imperial presence and white control in central Africa. Implicated in the arguments over location of the new dam project were the issues of water as a terrain security problem and the question of the 'balanced achievement of gains or equitable distribution of benefits'.53 The latter issue exercised debate over the whole federal system. That is, Which state and racial group benefited most from the federation? This debate created the perception of historically disadvantaged states and communities, issues which still inform debate about the ZRA. Against this backdrop, Zimbabwean and Zambian attempts to co-operate over the Zambezi are clearly not straight forward and unambiguous. Both countries view the Zambezi as conflictual and an issue of high politics, thus sometimes making the process of delinking issues relating to their bilateral relations a very difficult exercise. That is to say, hydro-political issues and differences are intimately woven into the fabric of the two states' bilateral relations.

General Zambian perception of the nature of relations between the two states as projected in their co-operation within the ZRA was reflected in an exhaustive coverage of the subject by the Weekly Post (October 1993) which depended heavily on views of Zambians employed by the ZRA. Those views could be interpreted to be opinions of potentially disillusioned employees with an axe to grind. They also, however, raise issues quite pertinent to this section of the study -- issues at the centre of troubled Zambian-Zimbabwean relations.

The Weekly Post's account, titled 'An Unequal Relationship' reviewed such issues as the division of CAPCO assets, the employment structure within ZRA, infrastructure development by the ZRA at Kariba and how the ZRA is carrying out its mandate to identify and develop power generating projects on the Zambezi section between Zimbabwe and Zambia. The latter focuses on the Batoka gorge project feasibility study. The Weekly Post asserted forcefully that Zambia was being given a raw deal in the ZRA partnership. On the distribution of jobs within the ZRA, the Post observed that Zimbabweans hold most jobs: 'Of the total establishment of 252 employees, only 27 were Zambians', thus, underscoring inequitable distribution of employment within ZRA.54

On infrastructural development, The Post asserted that there was great imbalance between infrastructure owned by the two countries, a perpetuation of what was inherited from colonial days. The paper observed that during the construction of Kariba power station and dam most Federal Power Board assets were located in Zimbabwe, particularly at Kariba. It alleged that ZRA 'continues to perpetuate the imbalances' in infrastructural development between Zimbabwe and Zambia concentrating construction of company houses and work to stabilise South bank at Kariba. Although the ZRA headquarters is in Lusaka, according to the Post, this is essentially a technicality as most ZRA '. . . functions like engineering, purchasing, tendering and stores' are effected from Harare.

On the division of CAPCO assets, the paper asserted that the arrangement favoured Zimbabwe which was given 50 percent of the assets with the other 50 percent going to the ZRA. The ZRA assets were then shared 25 percent each to Zambia and Zimbabwe. The paper, however, noted that there were anomalies in the repayment of the Commonwealth Development Corporation loans inherited from the CAPCO with Zambia having ìoverpaid by £1.6 million since 1987î.

The paper, again quoting Zambians within ZRA, made some observations on the feasibility study on the proposed joint project to exploit the power potential on the Batoka gorge. It observed that the ZRA, which co-ordinated the study, again perpetuated the imbalance obtaining at the construction of Kariba. The camp site at Batoka was located on the Zimbabwe side of the border, and so was the access road to the camp. Only Zimbabwean companies, Cementation Company and Blasting Specialists were hired during the exploratory work. Co-ordination of the project which cost US$3 million was effected from Harare where most of those funds were therefore spent. The Weekly Post then observed that Zambians working for the ZRA were adamant that the Batoka project should not be allowed to proceed '. . . until all the facts which disadvantage Zambia. . . are removed'.55

The paper made projections on electricity/energy requirements/needs for both Zambia and Zimbabwe and acknowledged that joint available potential at Kariba, Kafue and the thermal power stations in Zimbabwe was exceeded in 1993. Zambia will, however, continue to be self sufficient until 2003 with some projections putting self-sufficiency to 2010. Zimbabwe, on the other hand, has been experiencing energy deficits since 1990, thus making it urgent that Batoka be brought on stream as soon as possible. The Weekly Post is not alone in dissuading Zambia from going ahead with the Batoka project. A University of Zambia assessment of the Executive Summary of the Batoka feasibility study warns Zambia against embarking on the project when the country is currently self-sufficient in energy to about 2010. The study suggests that Zambia has not done a proper study of its power requirements and would most likely be roped into the project on the strength of Zimbabwe's more rigorous assessment of its needs. It also suggests that Zambia should in fact give priority to exploiting power potential wholly inside Zambia on the Kafue and the Luapula rivers with the Zambezi potential given secondary consideration. It remains largely uncertain as to whether and when the Batoka scheme will go ahead and the circumstances and dimensions the project would assume if it went ahead. It is, however, noteworthy that the Zambia power utility, ZESCO, strongly supports the Batoka scheme and wants the project to go ahead as soon as necessary preparations and studies are completed.

The hydropolitical differences between Zambia and Zimbabwe as dramatised by the ZRA case reflect the troubled relations between the two countries with misunderstandings largely driven by differences over the sharing of unitary held assets inherited from the federation days. This has had significant impact on the fabric of Zambian - Zimbabwean relations with differences over a range of issues gnawing at the foundation of sub-regional co-operation. For example, on July 3, 1995, the Zambian Ministry of Transport barred Zimbabwean heavy vehicles (weighing 2.5 tonnes or more) from entering Zambia. Zimbabwe quickly retaliated by imposing similar restrictions on Zambian commercial vehicles. In April and May 1995 a row between Air Zimbabwe and Aero Zambia, which is privately owned, over whether the new Zambian airline should be allowed to inherit the slots held by the defunct Zambian Airways, resulted in Air Zimbabwe being refused permission to land in Lusaka. These minor altercations are reflective of strained relations.

The ZRA case provokes pertinent questions about conflict potential in the Zambezi basin. The case also offers empirical insights into the effectiveness of regimes formed against the backcloth of suspicion and misunderstanding; that is, a protracted conflict setting. How stable are negotiated solutions, particularly against the backdrop of perceptions about historically disadvantaged states and communities? This addresses the question of regimes. Does the change of government(s) that will have negotiated a settlement implicate the stability of such settlements? The ZRA regime is clearly an unstable settlement for a variety of reasons including the fact that it has yet to realise its mandate on new power production projects like the Batoka gorge project. This is essentially a reflection of differences between the contracting parties to the agreement rather than lack of resolve within the ZRA. The perception of imbalances in the distribution of benefits and gains, past and future, also contributes to the instability of the agreement that brought about the ZRA. The instability is further exacerbated by the fact that Zambia and Zimbabwe are still bitterly divided over the ownership of the north bank of the Kariba scheme. They are seeking independent arbitration over the subject. It is equally possible that the new Zambian government which did not negotiate the 1987 ZRA settlement may not be wholly supportive of the agreement and could be deliberately defecting from the broader commitments suggested by the agreement; that is, on new power production projects.

So, do regimes then matter for peace and stability of agreements and are they a 'civilising' instrument that helps to bring about sanity between embroiled states? The ZRA case suggests that regimes themselves could be tools of combat implicated in the larger conflict between states already embroiled. If some individuals within an institution could recommend to a contracting party to take a particular course of action that may undermine the effectiveness of a regime, then the institution's neutrality is compromised. The ZRA has failed to be an effective regime to foster sub-regional co-operation by binding states into multi-level game strategy. The fact that the main body of Zambian intellectual opinion and policy analysts favour that country's exploitation and development of power potential away from the Zambezi (i.e. on the Kafue and Luapula rivers) suggests that the ZRA could be perceived as a single task regime with little scope for spill over into broader co-operation. Defection from the requirements of broader co-operation is readily executed suggesting the inability of the regime to civilise behaviour patterns. One of the major considerations by Zambia to exploit resources wholly within the national territory is the question of control over the project suggesting elements of water as a terrain security issue. Although Zimbabwe is equally refurbishing its thermal power stations, and expanding power importation from South Africa and Mozambique, the Batoka gorge still remains the immediate project the country wishes to develop to address urgent domestic energy demand.

It can be argued that by seeking to develop Kafue and Luapula power potential and deferring decision on Batoka, Zambia could be exercising veto through denial of the project to Zimbabwe. This is veto power of the less needy state in the basin. As a shared resource, Zimbabwe would need the consent of Zambia before it goes ahead with the project more so if Zimbabwe would need to mobilise international capital for the project. Negotiations over the project have been going on for sometime and if similar projects in the region (the Lesotho Highland Water Project took over 30 years to negotiate) offer any lessons, it is still a while before work starts on the Batoka. The player with the least to lose, in this case Zambia, can refuse to play any particular game with obvious implications on regime negotiations and the start to projects.

Is the ZRA the most suitable institutional home for the envisaged comprehensive regime for the Zambezi? Not unless it is significantly recast to ensure that issues currently troubling the authority, that is differences between member states, are made not to hold the larger regime to ransom. Perhaps the involvement of more states in a regime that replaces the ZRA will ensure that the regime ceases to be a platform on which nations play out their differences.

Intra-State Conflict Potential: The Matebeleland Zambezi Water Project.

The Bulawayo-Zambezi water project which envisages the abstraction of approximately 2 percent of the Zambezi flows during the lowest flow (which translates to about 0.25 percent of the mean annual flow) to Zimbabwe's second largest city, offers an example of intra-state conflict over water. The conflict potential as it currently manifests itself is driven, induced and aggravated by environmental scarcity that has sharply increased societal demands on the state while it simultaneously reduces its capacity to meet those demands. The current inability of the post-colonial state to satisfy the competing demands of society/communities within the state creates a crisis of legitimacy for the political system. This crisis of legitimacy which also confronted the colonial state, however, has different implications for the post-colonial state, as will be demonstrated below.

The water problem facing Bulawayo is not new. Proposals to abstract water from the Zambezi date back to 1912 and successive governments of the day have either rejected or postponed taking action. From 1912 to 1980 challenges to the colonial state to address the issue, culminating with armed struggle between 1966 and 1980, were conducted within the broader context of political control of the state. The answer to the question was integrative as competition for power between various groups centred on capturing the state. The colonial state never had any legitimacy. The same problem confronting a post-colonial state assumes different nuances and dimensions. The security and political challenges to the post-colonial state from Matabeleland, between 1982 and 1987 had ethnic dimensions and threatened the integrity of the state. Although water then did not assume the high profile it has now since the severe drought in 1991/92, the state's inability to address the issue was seen within the context of the conflict then raging and probably interpreted as a deliberate strategy to neglect a rebellious part of the country. The absence of any significant economic investment in Bulawayo is sometimes cited as instance of this alleged negligent strategy.

Since the 1991/92 drought, the water problem has become the single most significant organising issue politically in Bulawayo. The Bulawayans took the initiative and established the Matebeleland Zambezi Water Project and raised substantial funds to finance a comprehensive study of issues relating to the abstractions of water from the Zambezi. The government responded by establishing a second trust the Matebeleland Zambezi Water Trust (MZWT) and charged the Ministry of Local Government (not Water Affairs) with responsibility over the project. Although the government has promised to provide funding for the project as a long-term solution for the water problem and for interim dam projects (Gwayi/Umgusa and Gwayi/Shangani), no financial commitments were made in the 1993/93, 1994/95 and 1995/96 budgets. In fact, there was nothing committed for new water projects in the 1995/96 across the country. And this is against the backcloth of severe water scarcity confronting the country. This clearly demonstrates the inability of a distressed economy to respond to current environmental scarcities, particularly water shortages. The entrepreneurship role by Bulawayans to establish a trust to attempt a long term solution to the city's water problem outside the state's efforts underscored some significant points including the spirit of self-help and self-empowerment of citizens. It equally underlined the point that environmental problems (plus the forced retreat of the state), may be contributing '. . . towards a crisis of the nation state by eroding the normative appeal of the state and the idea of the nation state as the primary, if not exclusive focus of human loyalties. . .'56 The Zimbabwean governmentss swift response in establishing a new trust led by senior cabinet ministers from Bulawayo suggests a recognition that the stateís primacy as provider of infrastructure and focus of loyalty suffered the possibility of being irreparably undermined with serious stability implications for the state and the political system. The events of 1982 to 1987 were still fresh in the minds of ZANU PF political entrepreneurs.

It is uncertain at this stage, whether the Matebeleland Zambezi Water Trust, essentially a non-governmental organisation, has the capacity to supplant the state or could be said to be only supplementing the state. The question is pertinent when both the cost of the project, estimated at Z$2.5 billion, and the requirements of international negotiation with other riparian states are considered. The question of standing is very important in international negotiations. The Trust has no standing in negotiations with other riparian states that also want to abstract water from the Zambezi, and even those not interested in drawing water from the Zambezi. Equally, the Trust would have no standing in negotiating a huge international financing package for the project.The Trust, however, suggests that it has been promised significant amounts for the project (as outright gifts and grants) and hopes to form a private company where ordinary citizens can buy shares as a means of raising funds for the project.

The question of negotiating with other riparian states is pertinent when it is considered that the project may actually end up being larger than currently projected thus necessitating drawing larger quantities of water from the Zambezi. The Zimbabwean Farmers Union has suggested that it wanted to raise Z$100 million to help extend the Zambezi-Bulawayo project to Masvingo and parts of Midlands provinces. Extending the project to the Midlands and Masvingo entails an extra basin transfer and like all inter-basin transfers profiled in this study, makes certain assumptions that are largely questionable. The assumptions are that the Zambezi source is available, that the resources are plentiful and that the envisaged recipients have a right to the water. The validity of these assumptions will be tested in negotiations on the sharing of the Zambezi waters.

What will the Zambezi water be used for? Domestic, industrial and commercial demands are expected to consume the bulk of the water. It is also believed that large tracts of land in Matebeleland north province, with irrigable land estimated at 57,000 hectares, will be irrigated. Additionally, if the project is extended to Masvingo and Midlands provinces, more land will be irrigated by small scale land holders. It is considered wildly optimistic that small scale farmers will afford the cost of the water unless it is heavily subsidised by the government. Experts estimate the true cost of the water from the Zambezi at between Z$12.00 and Z$16.00 per cubic meter, which will be very steep for even domestic urban consumers. To irrigate at anything near the above cost would mean that farmers would need to go for exotic export crops like flowers, which will have little bearing on Zimbabwe or, indeed, the irrigator's own food security. It will not be surprising to have a scenario where water will be readily available but inaccessible to the less affluent because of cost. It would not be the first such occurrence. Manyuchi Dam in Mwenezi district, a multi-million dollar project has failed to benefit small scale farmers in the district. None of the 3,476 hectares which were to have been irrigated has been developed. In 1995, at the height of a severe drought in the district and nationally, the dam was over 80 percent full and benefiting only a few large scale commercial farmers. Mazvikadei, a large government owned dam built in the midst of commercial farms, remained inaccessible to small scale farmers, even fishing co-operatives, which wanted to use the dam, as commercial farmers controlled physical access to it. These two are but a few examples of scarcities induced by either legal instruments (water rights) or financial disabilities on communal farmers which are quite common in Zimbabwe and which can befall large scale billion dollar projects like the envisaged Zambezi-Bulawayo project. Conflict potential inheres in these situations and projects.

In sum, the Zambezi-Bulawayo water project draws attention to two problems that have confronted the post colonial state in Zimbabwe: the perceived inequitable manner in which the national question has been addressed, and the inability of the state to respond to environmental scarcity challenges confronting it. Both problems confront the state with a legitimacy crisis as the state increasingly becomes dependent on external forces for sustenance to enable it to address these difficulties. The project also emphasises the issue of the rise and strengthening of civic society. The post colonial Zimbabwe state perhaps needs both a strong civic society and a strong state so that the two counter balance and mutually reinforce each other for a stable environment. The danger lies in one trying to supplant the other rather than each supplementing the other. Water is implicated as both a terrain security issue and as an environmental scarcity problem that have stability implications for the state and the political system. Conflict levels include political conflict, economic, legal, diplomatic and potentially military conflict if there is a perceived serious threat to the integrity of the state.

The Lesotho Highlands Water Project

This project is currently reputed to be the most complicated and biggest engineering endeavour of its kind in the world. As noted previously, it took more than 30 years to negotiate ans those negotiations preceded Lesotho's independence. They commenced at a time when South Africa was negotiating with the Untied Kingdom so that the latter could be allowed to incorporate the High Commission territories of Basutoland, Swaziland and Bechuanaland. It is equally noteworthy that Lesotho has previously indicated territorial claims over large parts of South Africa's Orange Free State province. Implied in the two countriesí designs on each other territories is the issue of non-recognition.

The question of non-recognition was exacerbated in Lesotho's case by the apartheid issue with South Africa. Even when the two countries established diplomatic ties, relations remained clouded by the territorial issues observed above. In South Africa's case, Lesotho under Premier Chief Leabua Jonathan was seen as threatening Pretoria's security, particularly when Maseru established diplomatic relations with the Soviet bloc and Communist China and allowed the ANC to have a presence in Maseru. The protracted conflict setting profiled above shaped perceptions on water negotiations and explained, in part, the lengthy processes leading up to an agreement in 1987. In fact, the agreement only followed a military coup which Pretoria engineered.

When completed in about 2020, the Lesotho Highlands Water Project (LHWP) will consist of five dams and 200 km of water transfer tunnels. A hydro-electricity generating capacity of 180 kW is also envisaged. It is hoped that the electricity generated will make Lesotho self-sufficient in energy. The project is expected to transfer about 2,200 million cubic metres of water annually from Lesotho to the South African network for consumption in Gauteng province alone. The cost of the project, at 1987 prices, is estimated at R8 billion.

It remains uncertain whether the project will continue along the lines planned in the 1987 treaty. Firstly, there was no agreed hydrology between the two countries from 1987 to 1995. The hydrology will be used to determine the amount of water transferred and also the price of exported water. Secondly, there appears to have been an underestimation of water demand in Lesotho and in downstream settlements in the Orange basin to the sea. Falkenmark (1989) has shown that Lesotho will be facing absolute 'water scarcity' by 2025, five years after the completion of the LHWP. This suggests that the country may be exporting itself into water scarcity. Provision should have been made, in the treaty, giving priority of access to a resource exporting basin, with water export reduced proportionately to growing demand in the donor basin. There is conflict potential in this over sight.

Lesotho has, in fact, been privately demanding a re-negotiation of the 1987 treaty to address some of the concerns raised above. The country has been stalling on the start to the second phase of the project perhaps as a strategy to cause South Africa to address some of the concerns and perceived inadequacies of the 1987 treaty. In fact the expectations of the 1994 Reconstruction and Development Programme in South Africa have dramatically changed the water demand equation in both Lesotho and South Africa.

The Orange basin, of about 1 million square kilometres, embraces four countries: Lesotho, South Africa, Botswana and Namibia. The LHWP involves only two states. Namibia's interests were said to have been represented by South Africa, then effectively the colonial power in Namibia. Consultation with Namibia was, however, not carried out as extensively as stipulated by international law. Water demand patterns in Namibia were, therefore, not extensively researched. An expected increase in off-take from the Orange basin upstream of the Namibia -- South Africa border will negatively impact on water availability for Namibia and the estuary. This suggests that the needs of the environment and Namibia must be factored into any re-negotiation and re-planning of the project.

At the intra-state level, the LHWP though very popular in Lesotho has failed to generate the employment levels widely envisioned. Social benefits to the displaced population are also not flowing as readily as hoped. Less timely environmental impact assessments, looming water scarcity in Lesotho and the lingering perception that the 1987 treaty was negotiated by a pliable military regime, contribute to some demands for re-negotiation of some aspects of the project. While Lesotho is not expected to defect from the commitments of the 1987 treaty, any future water supply uncertainties, driven by factors such as climate change or population increase, are likely to cause considerable strain within the exporting basin and between an exporting and importing basin.

In South Africa, conflict potential is fuelled by competitive exclusion. About six million people within the Orange basin do not have access to clean drinking water. Less than half the rural population in the basin has access to a safe water supply. Conflict potential will be sharpened if the Reconstruction and Development Programme fails to deliver on the promise on water supply and sanitation. Also, inadequate access to water supply for blacks because of limited reform of water legislation, or resistance by those who historically have had riparian rights, will fuel conflict potential at national level, particularly in the politically volatile Gauteng province.

At sub-regional level, massive water development and diversion schemes to meet new expanded water needs in the Orange and Vaal basins will impact on the downstream state, Namibia. Water taken from the Orange basin to Gauteng is unlikely to be returned to the basin, but will be fed into the Crocodile and Limpopo basins. In the absence of a basin-wide treaty on water development, the downstream consumers will depend on the goodwill of upstream states. That goodwill will be difficult to guarantee in an atmosphere of scarcity. A deprived downstream consumer is likely to view the river resource as a source of conflict. It is as yet uncertain whether the proposed project for the Orange river will lead to Lesotho demanding a reduction to its commitment to deliver over 2 billion cubic metres of water annually to South Africa by 2020. Such a defection remains a possibility as long as Lesotho continues to stall o