Africa's river systems have been the target of development planners since the 1960s, and many of the major rivers of the continent have been dammed for irrigation, for power generation and flood control. Indeed, river basin development planning has been widely adopted in Africa, and often enough water resource development has come to be synonymous with river basin development (Adams 1992). Integrated river basin planning was pioneered in the U.S., and the basic objective was to co-ordinate water resource development in a given basin so that individual development schemes do not work at cross-purposes. The river basin, and not the individual farmstead, served as the unit of planning, the assumption being that what was good for the basin was good for the individual farm. Such planning exercise requires a powerful interventionist state, a strong central planning authority and an over reliance on physical engineering to solve all development and conservation issues.

River basin planning was adopted in Africa, essentially in truncated form, in part because it appealed to the authoritarian interventionist states that were then in power in many countries in the continent. Moreover, African governments and their willing donor agencies, which bankrolled many of the costly river basin schemes in the continent in the 1960s and 70s, were frequently seduced by the technological promise of large-scale water projects. Planners had high hopes and the objectives frequently sought were:

Ø      to raise the level of food production;

Ø      to increase the production of export crops and hence boost foreign earnings;

Ø      to bring under cultivation what are considered to be unutilised lands;

Ø      to fight against drought and the long dry seasons, both of which exacerbated the problem of food insecurity;

Ø      to meet the energy needs of industry and urban settlements; and

Ø      to satisfy the water needs of urban and rural populations.

 The problem of food security has been keenly felt especially in the Sahel countries and Ethiopia, both of which have become increasingly drought prone. The food crises of the 1960s, 1970s and 1980s have drawn attention to the issue of environmental vulnerability and the need for its mitigation. In many of the drought prone countries, the concentration of the human population is relatively high and cannot be adequately supported by rain-fed agriculture alone. Thus, where rainfall is insufficient or unreliable, and rain-fed agriculture cannot fully support food production, water management schemes have been considered to be sound investments. Such investments, it is argued, will help stabilise agricultural production and promote food security.

But many water projects in Africa are performing poorly or have failed outright, often with damaging environmental consequences. In many instances, the benefits have gone to a small segment of the urban elite and not to the masses of needy peasants and pastoralists. Some of the reasons for this sorry record include poor planning and design on the one hand, and the lack of involvement of the primary stakeholders in policy formulation and project management on the other (Adams and Grove 1983, FAO 1986, Moris and Thom 1990). The loss of traditional farming and grazing land, population displacement and relocation, and the long term and, at times, irreparable damage to the environment are but some of the costs that communities have had to pay for the failure of water projects (Adams 1992). In Ethiopia, for example, four costly dams that were constructed in the 1980s had to be abandoned, and several irrigation schemes became unusable due to poor planning and the authoritarian approach to policy formulation and implementation that was characteristic of the government at the time (MWR 1997a).

Water is an indivisible resource, and in this sense too it is different from most other natural resources. Water users are thus interdependent, and water control and conveyance systems affect the interests of large numbers of individuals in one way or another. The interdependence of irrigation users, for example, creates an environment in which each user loses a little bit of his or her individual control over farm practices (see Bromley 1982). Some have argued that since water is a common resource and since its utilisation promotes user interdependence, its management should not be left to the responsibility of individuals. Such arguments have often been used as justification, at least in this country, for policies favouring state ownership and management of water projects both large and small.

Water Resource Development in Ethiopia

The development of water resources for agricultural purposes on the one hand and rural water supply schemes on the other are the focus of our discussion in this section. Of the two sub-sectors, the first has attracted high levels of investment, and the second was neglected until the post-Imperial period. Even today, rural water supply programs, which affect the majority of the country's population, have not been given sufficient attention.

 Modern water development schemes are a relatively new phenomenon in the country. The Imperial government took the first initiative in water resource development in the second half of the 1950s. Large-scale water projects for agricultural purposes and power generation were constructed from the end of the 1950s, and were concentrated in the Awash valley as part of the agro-industrial enterprises that were expanding in the area at the time. They subsequently spread to the Rift Valley and the Wabe Shebelli basin. Essentially, the government's interest at the time centred almost entirely on large-scale and high technology water projects: hydro-power dams, irrigation schemes, and water supply projects for Addis Ababa and a few major towns. Since then, all large-scale schemes in the country have been constructed at the initiative of the government, and managed by state or para-statal enterprises.

  Irrigation Schemes

 Until recently, the water potential of the country was not accurately known, and even today this is still a contentious area. There have been different estimates of the irrigation potential of the country, and the issue has not been satisfactorily resolved. One of the earliest estimations was made by the World Bank (1973), which suggested a figure of between 1.0 and 1.5 million hectares. Recent estimates, however, place the figure somewhat higher. According to the Ministry of Agriculture (1986), the total irrigable land in the country measures 2.3 million hectares. The International Fund for Agricultural Development  (IFAD 1987), on the other hand, gives a figure of 2.8 m, while the Office of the National Committee for Central Planning’s 1990 figure, which is based on WRDA's estimations, is 2.7 m.. The Indian engineering firm Water and Power Consulting Services’ 3.5 m ha. is the highest estimate so far and EVDSA accepted the figure and was using it in the early 1990s. Most of these figures are derived by adding up the irrigation potential of the country's eight river basins as shown in Table 1 below. Except for the Awash River and the Rift Valley lakes, all the other basins are part of the major trans-boundary river systems that drain out of the Ethiopian highlands and flow into the neighbouring countries of Sudan, Kenya and Somalia. 

 

In the 1960s and 1970s, comprehensive reconnaissance and feasibility studies were carried out on the Abbai (Blue Nile), Awash and Wabe Shebelle river basins. In 1962, a German engineering team, and in 1964, the U.S. Bureau of Reclamation undertook extensive studies of the water resource potential of the Abbai River basin, the largest basin in the country. Both reports maintained that there were high hopes for the development of irrigated agriculture in the basin. The German study (Lahmeyer 1962), which was confined to the Gilgel Abbai basin, a much smaller area, suggested that the production of oil seeds, pulses and fodder crops, using the waters of the Gilgel Abbai, would be very profitable and earn high foreign exchange. The U.S. study recommended that small-scale irrigation should be greatly encouraged but that large-scale schemes would be too costly. It argued that without a co-ordinated water development program in the basin there would be no prospects for agricultural development in north-west Ethiopia. On the other hand, the Awash River basin attracted a good deal of local and international investment, and was the subject of numerous studies and surveys in the 1960s and 1970s (Dessalegn 1986). By the beginning of the 1970s, 100,000 ha. of land was under modern irrigation in the country of which about 50 percent was located in the Awash Valley (Wetterhal 1972). An extensive survey of the Wabe Shebelle basin, which was began at the end of the 1960s, was completed in 1972. About the same time, a reconnaissance survey of the Tekezze and Mereb and Gash Rivers in the north of the country was under way. In brief, the imperial regime was keen to determine the water resource potential of the country's river basins and to invite foreign capital to invest in agro-industrial enterprises in these areas.

 

In the late 1980s, an Indian firm, WAPCOs, prepared a preliminary master plan for water development for the whole country (WAPCOS 1990).  At present, water resource master plans have been completed for the Omo-Gibe and Baro-Akobbo basins, while plans for a similar undertaking are afoot for the Abbai and Tekezze river basins. A hydropower project is under construction on the Gilgel Gibe River. For the Wabi Shebelle basin, preparation of a master plan has been initiated, and the work is scheduled to be completed by the year 2000. Plans for a surface water study of the Awash basin, a survey of the Rift Valley, and a preliminary capacity assessment of the Genale basin are also in the pipe line, with completion targets set for 2000 (MWR 1996)^[2]. On a smaller scale, pre-feasibility and reconnaissance studies of watersheds and subsidiary river valleys have been undertaken at the initiative of WRDA and EVSDA in the 1980s. These include the Birr and Koga watersheds, Gilgel Abbai, and the Borkena catchment. The main objective of all these ventures has been to determine the water potential of the country, assess the extent and nature of water utilisation, and recommend priority areas for the development of water resources.

  

Table 1. WRDA's Estimate of Irrigation Potential (1986)

 

River Basin	Irrigable Land (Ha.)
Abbai	760,000
Tekezzae & Northern	200,000
Baro-Akobbo	600,000
Gibe-Omo	250,000
Rift Valley (Lakes)	50,000
Genale-Dawa	300,000
Wabe Shebelle	355,000
Awash	185,000
Total	2,700,000

 

Source: ONCCP 1990 (based on WRDA figures).

According to recent MWR data (cited in MEDAC 1999: 484-85), some 30 large and medium-scale irrigation projects with a combined command area of over 600,000 hectares have been identified in various parts of the country for development by the state since the 1980s. Of these, about 15 percent have already been completed. The largest water project to be constructed since the 1970s is the Alwero dam in the Gambella region, which has an irrigation potential of over 10,000 hectares, but which remains unutilised two years after completion. Feasibility studies have been completed on another 25 percent of the projects and the rest have been the subject of reconnaissance studies. These projects are separate from the large river basin projects for which comprehensive master plans are now being prepared.

 

In the pre-Revolution period, the chief purpose of irrigation was to provide industrial crops to the growing agro-industries in the country, many of which were controlled by foreign interests, and to boost export earnings. The main crops grown were sugar cane, cotton, sesame, fruit and vegetables. In the Rift Valley areas, some irrigation was used to grow food crops. There was a shift of emphasis in the post-Revolution period though the earlier objectives were not abandoned. The Derg, like its predecessor, was keen to promote large-scale and complex water projects, however, other issues now came to play an important role. Initially, irrigation was seen as part of the modernisation and socialisation of the country's agricultural economy. Moreover, irrigation was considered an important investment for improving rural income through increased agricultural production, and for reducing the growing pressure on the land by bringing unused land under cultivation. Later, with the recurrence and continued threat of drought and environmental hazards, the justification for water management schemes expanded to include relieving drought and recurrent food shortages, and growing more food for the internal market to improve food security and the nutritional status of the population. In more recent years, the need to reduce the pressure on the environment and to promote sound soil and water conservation practices has become an added consideration especially in those areas which are particularly vulnerable to land degradation. Economists have often emphasised that irrigation will bring about higher income for farmers, higher employment opportunities, and increased foreign exchange earnings.

 

For much of the lifetime of the Derg, very little attention was paid to small-scale and traditional irrigation schemes constructed and managed by peasant farmers. With the nationalisation of industrial and agricultural enterprises, the government's emphasis was to promote high technology water development schemes managed by state-controlled agro-industrial and agricultural enterprises. It was only in the second half of the 1980s, as a result of the devastating famine of 1984/85, that the Derg began to show interest in small-scale water management schemes (see MoA 1986, Tahal 1988). The establishment of the Irrigation Development Department (IDD) within MoA at the end of 1984, a body entrusted with the development of small-scale irrigation projects for the benefit of peasant farmers, signalled a new approach to water development by the military government. However, progress was slow. From the mid-1980s to the fall of the Derg in 1991, IDD was able to construct some 35 small schemes (MoA 1993), of which nearly one-third were formerly traditional schemes used by peasants.

 Table 2 shows the extent of irrigation in the country by the close of the 1980s. I should note that these figures should be taken with some degree of caution. Different sources give different figures, and one is not certain which is the more accurate. For example, ONCCP's figures for large and medium-scale irrigation schemes is given as 97,000 ha.  (rounded); small-scale irrigation is said to cover 64,000 ha, but the figure probably includes traditional schemes as well. ONCCP's estimation is based on 1989 data from WRDA records. On the other hand, EVDSA's estimate in 1992 was that only 160,000 ha. was under irrigation, and this constituted only 4.5 percent of the total irrigable potential of the country (CRDA 1996).

 

Table 2. Existing Irrigation (late 1980s)

 

Scheme	Area Irrigated (hectares)
Large and Medium	89,000*
Small Scale	10,000**
Traditional	69,000***
Total	168,000

 

  Source: MoA 1986, 1992, 1993.

Notes: *Figures as of mid-1980s. **Includes micro-dams and pump schemes.  ***Based on incomplete MoA inventory in 1990.

Whichever estimates we use, it is clear that even by the low standards of African countries, Ethiopia's use of its water resources is very limited. Less than 6 percent of the country's irrigable land is now under irrigation. In contrast, according to FAO data (1987), the three countries in Sub-Saharan Africa with the largest irrigation are Sudan (2.2 million ha), Madagascar (1.00 m) and Nigeria (0.9 m). In the Sudan, 14 percent of the country's cropped land is under irrigation, while in Madagascar, the figure is 32 percent. In contrast, almost all the cropped land in Egypt is under irrigation. For comparison, irrigation in Ethiopia covers less than two percent of the country's cropped land. Assuming that all the irrigated land is utilised to produce food crops (which in actual fact is not the case as the many of the larger schemes were dedicated to industrial crops), the contribution of irrigation to the production of food cannot exceed two percent.

 

The distribution of irrigation schemes in the country is quite skewed. Almost 74 percent of the irrigated area served by large and medium schemes is located in the Awash valley. However, as shown in Table 1 above, the Awash River basin contains less than 7 percent of the irrigable area of the country. In contrast, large and medium irrigation covers less than one percent of the Abbai basin, the largest basin in the country (ONCCP).

 

Under the Derg, irrigation schemes were divided into three categories depending on their size, operation and management. These were:

 

Ø      Large-scale schemes: Over 3000 ha. Designed by NWRC and WRDA and constructed by EWWCA mainly for the benefit of State Farms.

Ø      Medium:  200-3000 ha. Head works, main and secondary canals constructed by WRDA, and tertiary and farm canals by IDD. Managed by State Farms and other para-statal enterprises.

Ø      Small-Scale: Upto 200 ha. Constructed by IDD mainly for the benefit of peasants organised in producers' co-operatives.

 

It is quite evident that irrigation development in Ethiopia did not attempt to involve the farming population. Modern irrigation by and large bypassed the peasant, and the technology involved and the operation and management of this technology was entrusted to a small technical and managerial elite working for large-scale foreign interests in the past and later for state or para-statal enterprises. On the other hand, there is a long tradition among peasant farmers of water management for small-scale agricultural use. As shown in Table 2 above, more than 40 percent of the irrigated land in the country is served by traditional schemes; the figure would be higher if a more accurate and more complete inventory was undertaken. The majority of existing traditional irrigation schemes are micro-level in size, serving a small group of households usually not more than 20 to 30 in number (see Tahal 1988). But there are some schemes that serve a large number of beneficiaries. Many of these schemes are based on stream diversion, but some may be dependent on perennial springs. The water is transported by means of furrows, sometimes for long distances involving impressive engineering works. The diversion works are frequently rudimentary, and require frequent repairs, which involve great expenditure of labour on the part of the beneficiaries. These traditional systems may be described as forms of water user co-operatives. Each beneficiary has access to water on an equal basis, and equity in water distribution is a strong factor. 

 

Traditional irrigation is a complement to rain-fed agriculture, and the crops grown are often horticultural crops and fruit trees. Peasants have a keen awareness of the benefits of irrigation and are willing to invest their labour in the construction and maintenance of the schemes. In parts of north Shoa, north Wollo, east Gojjam and the highlands of Harrage, the traditional systems still being utilised by peasants date back to the last century. Many of these schemes are managed by elected elders known as "water fathers" or "water judges" and this traditional management system has proved effective in many instances. In some cases, the irrigation schemes are managed by PAs. It is thus evident that peasants have proven ability to organise themselves and to manage small-scale irrigation systems. The labour and discipline necessary to maintain these systems over many decades is evidence of a high level of practical knowledge of water management in the rural areas. The Derg almost destroyed traditional irrigation schemes by confiscating them and handing them over to producer co-operatives.  

The experiences of the Derg period are instructive in a variety of ways, and future water development policies should benefit by the lessons of the past and should avoid the mistakes that were committed. The two most critical mistakes of Derg water policy was that it did not encourage a pluralist approach in water development on the one hand, and it did not seek to involve the beneficiaries of water schemes in any way on the other.

 Both in terms of choice of technology and scale of operations, the Derg's emphasis was on costly investments that required high management and maintenance costs. Except the grudging concession made to small-scale irrigation in response to the drought of the mid-1980s, water development policy favoured large and complex water projects. A pluralist water policy would have actively promoted the development of all types and classes of water schemes. This would have been less costly and more effective. Moreover, given the emphasis on complex and large-scale water projects at the time and their management by a small professional elite, the diffusion of new water management expertise among the farming population was out of the question and not actively sought in the first place.

 

By and large, most of the state-run water projects in this period were poorly operated and poorly managed. There was an inefficient use of water partly due to the lack of a consistent policy on water charges and to the low water rates that state and para-statal enterprises were made to pay. A number of water projects were poorly planned and designed, and as a result they were either abandoned much before the end of their life-span, or kept in operation with reduced efficiency and capacity. Other projects caused serious damage to the environment.

The Derg's programme of small-scale irrigation was in large measure a failure also. Many of the schemes that fell under the responsibility of MoA's IDD were originally owned and operated by peasants. They were taken over and upgraded by the government without the consent of the communities concerned. Quite often, such upgrading involved the transfer of the schemes to producers' co-operatives (PCs), with IDD managing the schemes for the benefit of the PCs. Peasants who earlier had access to irrigated water but who were unwilling to join the PCs were denied access to water and were relocated elsewhere.

 

Policy planning and implementation at the time was guided by a strong top-down approach. There were few occasions when stakeholders were involved in any aspect of water resource development. Neither the direct beneficiaries nor concerned institutions at the local level were consulted in the planning and implementation of water projects. Moreover, the management of the projects themselves were in the hands of party or government functionaries, and not in the hands of the beneficiaries themselves. The irrigation schemes lacked operational autonomy, and there was no sense of ownership on the part of the beneficiaries. Because of the association of irrigation with collectivisation, many peasants shunned irrigation and remained suspicious and reluctant to return to it even after the fall of the Derg.

 

Rural Water Supply

Water supply services in Ethiopia are among the lowest in Africa. Moreover, the strong urban bias on the part of successive governments since the early 1970s has kept water supply investments in the rural areas quite low. According to WRC, by the beginning of this decade, only 19 percent of the country's population, and 11.5 percent of the rural population had access to safe water. According to MWR (1996), 19 percent of the rural and 80 percent of the urban population have access to safe water, and the total coverage for the country is put at 26 percent. The reason the urban figure is high is due to the high coverage for Addis Ababa; if we exclude the capital, the picture is rather dismal. According to another MWR document (1997b), prepared for the Ministry by the international accounting firm of Ernst and Young, the average water supply coverage for urban settlements excluding Addis Ababa is 31.3 percent. If we disaggregate this by population, the coverage for urban settlements with more than 50,000 people is 58 percent while coverage for those with 5,000 or less is 14 percent. The same document states that the average coverage for rural settlements is 15.2 percent and for the whole country 17.3 percent. Regionally, its findings were 34 percent coverage for Tigrai, 14.3 percent for Oromia, 12.5 percent for the Southern Region, and 8.3 percent for Amhara Region.

 

            Thus the great majority of Ethiopians uses unsafe and polluted water and as a result is commonly exposed to a large variety of water-borne diseases. This is all the more serious in that the rural population has virtually no sanitation facilities, while in the country as a whole only eight percent of the population has access to sanitation. Moreover, in terms of accessibility to health services the country is in the worst possible situation when compared to countries in Sub-Saharan Africa (UNDP 1998). Conditions are even more critical if we bear in mind that access to safe water in our case does not mean access to adequate water. MWR defines "adequate" water supply to mean 20 litres of water per person per day and accessible within a range of 0.5 to 1.0 km from a dwelling place (MWR 1996). The WHO standard, which was once adopted by WSSA, is 45 litres per person per day. Thus, most households with access to safe water do NOT get sufficient quantities of it for a healthy life. As we shall see below, the current level of per capita water consumption is far below the adequate level set by MWR. We should also note that in the rural areas safe water does not mean water that has undergone treatment: most households have access to potable water only from wells and protected springs.

 

Table 3 shows the extent of water supply coverage for Ethiopia and selected African countries.

 

Table 3. Access to Safe Water in

Selected African Countries (% of Pop.)

 

 

Country	Urban	Rural	Total
Ethiopia	80.0	11.5	19.0
Kenya	61.0	21.0	28.4
Tanzania	75.0	40.4	52.1
Ivory Coast	100.0	75.0	82.8
Senegal	85.4	26.0	51.2
Zambia	76.2	42.8	58.9
Zimbabwe	100.0	13.5	35.5

 

Source:            World Bank 1995 (figures for 1988-93). For Ethiopia: WRC 1990 figures.

Rural water supply services began in the late 1950s under the Imperial regime. However it was not until 1971 that a body responsible for all aspects of water use and development in the country, the Water Resources Commission, was established. True, the Awash Valley Authority was set up in 1962, but its duties were to plan and promote investment activities within the Valley. The Commission was given a wide mandate and entrusted with the responsibility of planning and utilising the country's water resources including water for home consumption. Up until the late 1970s, water supply schemes relied on motor driven boreholes, but since then other technologies have been employed, including hand-dug wells (with or without hand pumps), spring protection, and occasionally artificial ponds. However, water drilling, like other development undertakings, was carried out in rural communities which were close to the main road network; communities in the interior which were inaccessible by motor transport were largely ignored. The extension of the area of coverage in the Derg period was made possible by the expansion of the transport network.

 

Rural water supply became a high priority during the Derg period, and an accelerated growth in rural water schemes was registered in the later part of the 1970s. In the early 1980s, the government pledged to implement the UN initiated International Drinking Water Supply and Sanitation Decade, which in Ethiopia ran from 1984 to 1994, coinciding with the government's Ten Year Plan, which set an ambitious target for the provision of safe water supply to the rural areas. At the beginning of the 1980s, less than 6 percent of the rural population and 19 percent of the population in the twenty major towns had access to clean drinking water. At the end of the Plan period, the coverage for the rural areas was to reach 35 percent and for the urban areas 85 percent. While the record of achievement was not as high as planners had hoped for, considerable progress was made in the 1980s (see Table 4 below).

 

An important impetus for expanding the rural water supply programme in the country was the drought that hit the country in the 1970s and the 1980s. As part of their response to the environmental crises of these years, a considerable number of NGOs and several donor agencies became actively involved in rural water supply schemes. At present, some 38 NGOs, UNICEFF and number of bilateral organisations are closely involved in rural water supply. Table 4 provides data on the type and extent of water supply in the rural areas. It should be noted here that some of the schemes are not functioning due to faulty installation or lack of maintenance. The magnitude of malfunction of water supply schemes is not known accurately though it is believed to be high. An official of MWR recently estimated that at any given time more than 60 percent of all systems are in various stages of disrepair, and thus in effect only 40 percent of the population covered by improved water systems has access to safe water (CRDA 1997: 51).

 

Table 4. Number of Rural Water Supply Schemes, 1992

 

Const. by	Borehole	Hand dug Wells	Spring	Other	Total
WSSA	873	2507	435	2	3817
NGOs	608	797	982	15	2402
Total	1418	3304	1417	17