The Water Crisis in Yemen

The central place of agriculture in Yemen will be clear from the discussion in Part I of the role of agriculture in its history. In an arid land, the possibilities of agriculture are determined by the availability of water and by how that water is developed and managed.

The present chapter starts, then, from two premises. As agriculture uses more than 90 per cent of Yemen’s water, to understand agriculture is to understand a large part of the water question. And, by corollary, to understand agriculture and its potential, we need to understand what water is available to farmers and how best they may use it.

The chapter begins with a snapshot of Yemen’s agriculture and its potential and challenges today (section 6.1), before examining ‘water for agriculture’ and the contemporary development of irrigation (section 6.2). A third section (6.3) looks at the crisis today in agriculture and water, and at responses to stagnant productivity and poor sustainability. The following section (6.4) evaluates results to date of reforms and investments and looks at options for the future to avoid the exhaustion of groundwater reserves and the immiseration of the rural economy. The discussion is completed (section 6.5) by an examination of the political economy of agricultural water: can Yemen make the right choices in agriculture and water and put them into practice?¹

This section provides an overview of Yemen’s agricultural economy, its current structure and problems. Agriculture represents about 93 per cent of total water use, and intensive use of water has always been the mainstay of the rural economy. Since the development of commercial agriculture based on groundwater, agriculture has flourished as never before.

Yemen divides into three broad agro-ecological zones: coastal, highlands, and the eastern plateau

The hot, humid coastal plain receives little rain, and cropping is largely dependent on spate flows and on groundwater. The highlands receive rainfall between 300 and 1,200 mm. Here, cropping is practised on terraces and, when groundwater, springs or run-off are available, on the plains that lie between the mountains. In the dry eastern plateau, cropping is dependent on spate flows and groundwater, or is practised in the scattered oases. Table 6.1 summarizes the principal crops.

The transition from traditional farming towards a market-based agriculture sector

Traditional agriculture was largely for subsistence, but since the 1970s a rapid transformation towards a largely market-orientated agriculture has taken place

The population explosion together with rising incomes have led to a vast expansion of demand. This, combined with a change in the nature of demand towards higher value products, particularly fresh fruit and vegetables, meat and qat, have driven modernization and the rise of commercial farming. Improved roads, transport and market facilities have helped commercial agriculture to grow.

Rapid growth in the production of high-value irrigated crops was made possible by the tubewell and by availability of capital. Market demand for high-value crops soared with urbanization and rising incomes. Social aversion to cash crop production (see Chapter 2) disappeared with the generalization of the market economy and individualistic profit-driven patterns of thinking (Figure 6.1).

Figure 6.1 Yemenis have adapted rapidly from subsistence farming to commercial agriculture. Photograph courtesy of Dorte Verner.

Even after the rapid modernization of recent years, Yemen remains a predominantly rural country, where agriculture is still an important sector, accounting for 33 per cent of employment and 11 per cent of GDP.³

Agriculture today is characterized by market orientation, by diversification, and by the importance of irrigation, particularly from groundwater. Production of cash crops has increased enormously, and Yemen is largely self-sufficient in fruit and vegetables. Qat cultivation (see below) has spread to cover one-eighth of prime land. By contrast, the importance of cereals has declined. Few rural households are now self-sufficient in cereals, and imports account for three-quarters of consumption. Linkages to industry are weak, and agriculture has an inward-looking orientation – exports are very small.

The expansion of irrigation has contributed to rapid growth of commercial agriculture

The total irrigated area has doubled since 1970: now, more than two-fifths of farmed land is irrigated. The area under cash crops has shot up from 3 per cent of the total in 1970 to 14 per cent today, and production of high-value fruit and vegetables has increased 20 times, from 40,000 tons annually in 1970 to 800,000 tons today.

Levels of husbandry and productivity are low, with no improvement for most crops in recent years. Yields remain well below technical potential (Figure 6.2). Irrigation efficiencies nationwide average only 40 per cent. Producer services such as research, extension and credit have been in crisis for years.

Groundwater irrigation now accounts for two-thirds of value added in agriculture, but sustainability is threatened by over-pumping, with rapidly declining water tables, and by watershed and range degradation upstream that provoke erosion and reduce groundwater recharge.

The rural–urban terms of trade have deteriorated, partly as a result of the government’s structural adjustment programme, as diesel prices move to border parity, and fruit and vegetable prices decrease as import restrictions are lifted. Increases in the price of imported cereals in recent years may improve incentives for a minority of farmers but will increase costs for the majority of rural people who buy cereals. Rural areas are under the double pressure of shrinking water availability and rising population.

Poverty in Yemen is largely a rural phenomenon: 75 per cent of the population, but 84 per cent of the poor, live in rural areas. Rural poverty rates remained at around 40 per cent for the period 1998–2005, while urban poverty rates dropped from 32 per cent to 21 per cent. Today, the combination of population growth and the combined impacts of the global food and financial crises are worsening the situation. Currently, rural poverty is estimated at more than 50 per cent. Inequality is also on the increase, with the gap between the rural rich and the rural poor widening, as the better-off can exploit their ownership of land and water.⁵

Irrigation has kept poverty down in areas where water is available – but for how long?

About two-fifths (41 per cent) of rural households live in districts where the poverty rate is relatively low (30 per cent). In these districts, there is better access to water and land – about a third of households practise irrigation (32 per cent), and about half (47 per cent) of their incomes come from crop and livestock production. By contrast, in poor districts, where almost half the rural poor live and where the incidence of poverty exceeds 60 per cent, irrigation is much less accessible (21 per cent of households). In these areas, agriculture is more focussed on extensive livestock production and on rainfed cropping, and labouring and migration for work are dominant coping strategies. However, declining groundwater tables threaten even the comparative prosperity of districts where irrigation is currently important.⁶

A particular set of challenges surrounds the drug qat, which is both a major water consumer and the source of much rural prosperity

Qat is a tree which Yemenis have long grown in order to harvest and chew the leaves, which act as a mild, amphetamine-like stimulant.⁷ Qat production and consumption have exploded in recent years. Today, the drug occupies 12 per cent of the cultivated area, consumes 30 per cent of Yemen’s water, produces about 4 per cent of gross national product, and employs one-seventh of the workforce (Figure 6.3).⁸ Largely a smallholder crop and grown only at altitude, qat is in many ways an ideal crop for the water-scarce Yemeni highlands, as it is relatively hardy and easy to grow, and it returns by far the highest net revenue per unit of water consumed of any crop. Qat has well-developed year-round markets accessible to even the smallest producer, bringing in huge daily cash flow to rural areas, keeping production zones relatively prosperous and providing second-round stimulus to neighbouring food producing areas. Qat’s profitability also sustains parts of Yemen’s precious heritage of terraces. On the negative side are the high water use and the crowding out of other cash crops; but seen from the farmer’s perspective, qat is the crop of choice wherever it can be grown.⁹

Figure 6.3 Harvesting qat in the vast qat monoculture of Al Azraqayn in northern Hamdan. Photograph courtesy of Peer Gatter.

On the consumption side, qat chewing is now universal in Yemen, and it is the national preoccupation. Qat sessions promote sociability and social mobility. However, many families spend more than they can afford, menfolk pass long hours away from their families, and there are mild health impacts which may be damaging in the longer term. Attempts at regulation and taxation have had little effect.¹⁰

A particular set of risks and challenges also attaches to possible climate change

Recent scenario analysis¹¹ suggests that Yemen will have to adjust to warmer temperatures and to manage risks from more unpredictable rainfall patterns and heavier rains. The risks are not, however, all ‘downside’ – if farmers can adapt, rising temperatures could increase output where water is available. In general, it is groundwater depletion rather than climate change that will have the largest effect on agriculture, although depletion effects could be exacerbated by climate change impacts, particularly in lowland areas if the weather turns hotter and drier. The risks of drought and flooding will also affect farming. Overall, climate change effects are likely to be felt only at the margin of the central challenge of groundwater depletion; and, given Yemen’s diverse topography and climates, would be different for different areas.

Given the demographic explosion in a still largely rural country, the rural economy will have a key role in absorbing labour and providing incomes for new entrants for the foreseeable future – at least if an uncontrolled and impoverishing rural exodus is to be avoided. The rural population is expected to grow at almost 2 per cent annually, and because of past growth, the labour force will expand even faster – at 4 per cent nationwide, adding hundreds of thousands of new entrants to the rural workforce each year. Agriculture remains the largest source of rural income and must play an important part in future rural livelihoods. Given high rural poverty, this creates a massive challenge.

Yemen has a large population – 25 million people – rising at up to 3 per cent a year, so it will double within 30 years. Domestic demand will thus continue to rise, and will move ‘up market’ to higher-value foods as urbanization continues and incomes rise. Demand from industry could increase for cotton and fruit. Export niches also exist: original Mocha coffee, frankincense, myrrh, saffron, etc.¹²

On the supply side, there are potential sources of future growth in agriculture

Many crops have economic and technical potential for further growth, including cotton, grapes, papaya, coffee and market garden crops. These crops have low domestic resource costs (see Tables 6.2 and 6.3 below) – i.e., they are in economic terms efficient users of Yemen’s scarce natural resources. They also have potential for productivity improvements by improving the efficiency of water use and crop husbandry. Potential genetic improvements exist, and post-harvest, marketing and agro-processing activities can increase the value added. Investment in these crops would be in line with Yemen’s comparative advantage – that is, Yemen should, if possible, specialize in these crops.

Table 6.2 Domestic resource costs: crops with domestic resource costs of less than 0.5¹³ (i.e. highly efficient users of domestic resources)

Table 6.3 Crops with domestic resource costs of 0.5–1.0 (i.e. relatively efficient users of domestic resources)

However, growth in agriculture will face massive problems – of water shortage, of productivity, and of equity

Agriculture currently faces daunting challenges: dwindling water availability, and lacklustre productivity growth. If the rural economy is to survive and grow, attention is needed on both these fronts – conserving water and boosting productivity. This will require improved services to the sector: research, extension, credit, as well as a favourable enabling environment. Beyond the challenges of restoring growth lies the distribution and poverty dilemma: those who lack access to adequate water and land will have difficulty in sharing in a growth path based on increased value added in crop production.

6.2 The development of irrigation infrastructure and agricultural water management

This section describes the development of agricultural water management in Yemen and gives a snapshot of current status and issues.

Yemen’s water resources and the range of techniques traditionally used to develop and use water for agriculture are described in Chapters 4 and 5. In summary, rainwater is used directly in rainfed agriculture. Due to the low rainfall, the erratic distribution and high evaporation rates, most rainfed agriculture uses some form of water harvesting, such as ‘run-off/run-on’, small dams or terraces, or combines rain with supplementary well irrigation. Spate flows and baseflows in wadis are diverted for agriculture by weirs. Springs are harnessed by simple diversion structures and conveyed by small canals. Water from the alluvial aquifers is tapped, largely for agriculture, by shallow wells. The innovation in recent years has been the spread of tubewell technology to tap groundwater from deeper rock aquifers, sometimes several hundred metres down. Deep groundwater is used largely for agriculture but with growing household use. The growth of groundwater use has led to the depletion of Yemen’s aquifers and poses challenges for the future – not only of farming but for the country as whole. The issues of resource depletion and sustainability are discussed throughout this book, and particularly in Chapter 9.¹⁴

6.2.2 The modern development of irrigation infrastructure and agricultural water management

Traditionally, Yemenis had many ingenious techniques to husband their scant water, and these had equally well-developed institutional systems to ensure their working. As discussed in Chapter 5, in the highlands, water harvesting systems were commonly employed, both run-off/run-on systems and terrace systems. To date, water harvesting is the predominant system in the northern highlands in terms of farmed area (although not of value added). A 1986 survey in Amran (average rainfall 300 mm) showed that 80 per cent of land was rainfed/water harvested: 2 per cent of land was rainfed/water harvested with supplementary irrigation by pump; and 8 per cent of land was primarily pump irrigated.

In the ‘run-off/run-on’ systems (Chapter 5), hillsides are cleaned and prepared to deliver run-off water to levelled fields below, or to storage tanks. This allows cultivation in low rainfall areas (as low as 50 mm). Efficiencies can be high: up to 50 per cent of precipitation from a run-off area can arrive on the run-on field. The key factor is the existence of a reasonably impervious and well-prepared run-off area: gently sloping rock, an already saturated soil profile.

Terrace systems (Chapter 5) are still found throughout the highlands (Figure 6.4). Levelled and embanked fields descend the hillsides, each retaining precipitation and catching run off, benefitting and passing water on to the terraces below. Terraces provide level fields even on steep slopes. A third water harvesting technology is hill dams (Chapter 5), to check the flood flows either for diversion or for the recharge of groundwater.

Some systems are no longer economic and have fallen into disuse, particularly where they compete with higher yielding pump irrigation. In some areas land previously used as a run-off area, watering individually owned run-on land, has now been converted to pumped irrigation. Lichtenthaler graphically describes the process (see Box 6.1). Other constraints include: declining profitability, due to the high labour and animal requirements; environmental problems, due to the deterioration of the upper catchments; and land tenure problems, particularly land fragmentation. Nonetheless, water harvesting systems are sustainable, and planners have long tried to devise a mechanism to help restore terrace systems to their old vitality, if only as part of larger watershed management schemes (see section 6.3.5).

Box 6.1: Groundwater replaces water harvesting in the Sa’ada basin – but is now itself ailing.¹⁵

The dry Sa’ada basin in Northern Yemen has seen a population explosion. In 1975 there were 40,000 people in the basin; by 1997 there were 180,000. This was due to natural growth, returning migrant workers and internal in-migration in response to economic opportunity. Much of this opportunity has been in agriculture. Returnees from the Gulf brought capital in search of investment opportunities, and investment in the land suited the traditional values of this tribal region. Government improved agricultural profitability with a fruit import ban, so that orange and pomegranate farming looked an attractive investment. More recently, soaring demand for qat has made that a preferred crop.

Growth of agriculture was based on the rapid development of groundwater. In Sa’ada until the 1970s, most land was communally owned grazing land. The run-off rights from this land belonged, however, to individual proprietors in the bottom lands. Agriculture was not allowed on the grazing lands as this would impair the run-off. Therefore, tubewell irrigation could not develop on the slopes. Despite this, a deal was negotiated that allowed the owners of the pasture rights to convert half the slope land to agriculture on condition that the owner of the run-off rights was compensated with the same right on the other half. In 1976, a local cleric promulgated this practice in a fatwa and the rule change has been followed ever since.

Many tribal communities privatized their common land and distributed it to each household. Private tubewell development took off. A new elite of commercial farmers emerged; land and water resources are freely bought and sold; qat (previously scorned as ‘the tree of the devil’) is widely cultivated. Now the water table is plummeting and springs have dried up. The new groundwater boom may prove to have been short lived but it will be hard to return to the old water harvesting systems that it briefly replaced.

Figure 6.4 Ancient terraces capture run-off from the hillsides. Photograph courtesy of Gerhard Lichtenthaler.

Spring irrigation was traditionally important (see section 5.3), with each spring channelled and divided by canal systems as complex as the rights and rules that governed the sharing of the water. Organization and husbandry were well developed. However, many of these springs have dried up with the motorized depletion of the aquifer. Mundy, studying Wadi Dahr just outside Sana’a, gives an excellent description of this process and of the related social and economic tensions (Box 6.2). Varisco, who documented a historic spring system in Wadi Ahjur, makes a laconic citation of the named springs in the Sana’a basin that have disappeared from history in recent years. As a result of the depletion of the aquifer and the advantages of the tubewell, the economic importance of spring irrigation has dwindled, and springs now irrigate only 2 per cent of the farmed area.¹⁶

Box 6.2: Wadi Dahr – adjusting from spring to tubewell.¹⁷

Up to the end of the 1960s, irrigation water in Wadi Dahr, just outside Sana’a, was provided from springs, and villagers had rights distributed according to the land they owned. Historic upstream/downstream disputes since the seventh century had led to a brokered agreement that allocated 15 days to upstream users (‘the upper wards’), and 15 days to downstream users (‘the lower wards’). Supervisors were appointed by the community to distribute the water.

Rights to spring water were documented, with a register recording land and water rights. For every 100 libnah of land, a farmer had one ta’sah of water. These rights could not be permanently alienated from the land. However, from the early 1960s, the system was adapted to allow farmers to trade or sell their water entitlement. This greater flexibility was introduced due to qat, so that farmers could apply extra irrigation when the market was high.

Then, in 1970, the tubewell irrupted into the valley. Very quickly farmers in the upper wards drilled 21 wells and installed diesel pumps. The springs began to dry up. The lower wards went to court, but the sheikh who controlled the court was from the upper wards – and an owner of one of the new wells.

‘Justice’ was not forthcoming. One farmer from the lower wards said: ‘People are fools to be taken in by the sheikh. They are like sheep but with one difference: when they are sold, sheep go with the buyer, but here, after being sold, people return to the pen to be sold again.’

The lower wards transferred their allegiance to another sheikh, but then the wealthier farmers in the lower wards also purchased pumps. Mundy comments: ‘The spread of the technology began to render absurd the political moves adopted to contain it… .’

The pumps allowed qat to be planted on cereal land. By 1980, a new equilibrium had been established that depended entirely on private pump irrigation. There was more water than before, and money flowed in the wadi.

A dynamic economic and social situation emerged. Those who had money drilled wells and appropriated water. Those who had traditional rights but no money were losers. But then the qat market of Wadi Dahr – which is home to one of the most prized varieties and very close to the rich market of Sana’a – carried everyone up on a wave of prosperity. Even the landless became qat traders.

At this point, the water table began to decline, and this promised to bring new dislocation.

Traditional spate irrigation schemes survive throughout Yemen, although their economic importance has waned

Spate irrigation may be the oldest form of irrigation (see section 5.3). Box 6.3 describes how traditional spate schemes are managed to this day in Wadi Hadramawt. Traditional communal schemes are now less important in economic terms than modern schemes, although simple technologies like gabion reinforcement have been used in many schemes to improve diversion and protect wadi banks (Figure 6.5).

Box 6.3: Traditional spate water management in Wadi Hadramawt.¹⁸

Spate water rights are typically upstream or highest first. At Hureidah in Wadi Hadramawt, each field has its own name and its order of priority. However, there are limits to the amount of water a farmer may take before passing the water to his downstream neighbour: cereals in Hureidah, for example, may ‘drink up to the height of their stalk’ – about 18 inches – while date palms may be irrigated up to the top of the bund.

The nature of storm events in Yemen has meant that floods have always been a problem for spate irrigation systems. Nonetheless, spate systems have continually regenerated themselves. After many destructive floods throughout history, there is little evidence of cultivated fields being abandoned. However, siltation is a big problem. Fields at Shibam in Hadramawt have shrunk because their banks are up to 10 m high. Eventually, fields may be abandoned from this cause.

Cooperation is essential. At Hureidah, Serjeant describes how the whole village turned out when the sayl (flood) came down. If there was a breach in a field bund, the farmers called ‘Maqsar!’ (Breach!) and everyone came to help. When the main diversion structure in the wadi bed was breached, a drum would be beaten at night to announce what the farmers should bring next day to work on it.

Organization and management have always been an imperative. At Hureidah, they selected a ra’is al ma. At Hainan, there was a khaiyil who would divide the water.

Figure 6.5 Gabion baskets filled with stones are used to protect the wadi banks and improve spate irrigation (Wadi Hadramawt). Photograph courtesy of Matthias Grueninger.

Early development of modern spate schemes in the 1950s adapted traditional spate technology to a larger scale

Modern spate development began under the British in the Western Protectorate in Abyan in the 1950s. The Colonial Welfare Development Fund made a grant to finance a scheme to grow cotton. The scheme was so successful that the Abyan Board quickly repaid the grant.

Subsequently, with donor support, the two Yemens developed large spate schemes with fixed structures

The southern republic, PDRY, made heavy investment, with support from the Soviet Union and the World Bank, in spate schemes in Lahej (Wadi Tuban), Abyan (Wadi Bana, Wadi Hassan), and Wadi Ahwar. The northern Republic of Yemen invested very heavily in developing spate irrigation in the five Tehama wadis, largely with World Bank support. These investments created a large area under improved spate irrigation totalling about 117,000 ha countrywide, including Lahej 20,000 ha, Abyan 31,000 ha, and Tehama 67,000 ha. In addition to financing all the investment costs, the two governments also assumed responsibility for the operation and maintenance of the schemes.

By spreading water over wide areas, spate schemes increase aquifer recharge and also provide a leaching dose of freshwater to the fields, dissolving and diluting salts accumulated in the soil horizon and draining them down deeper. As the schemes are largely in the coastal plains adjacent to the sea, this recharge also helps resist the intrusion of seawater into the freshwater aquifer. In addition, spate structures reduce the possibility of flood damage from torrential spate flows.

First, although the introduction of permanent diversion structures increased the reliability of spate irrigation, it also changed water distribution patterns. For example, in the Tehama wadis, the upstream farmers had always benefited from first rights to spate water. With the introduction of more efficient diversion, these benefits increased at the expense of downstream farmers.

Second, the economic returns are less assured than in schemes fed by permanent river flows. Spate is dependent on the chance of flood flows. The canals have to be enormous to capture the vast flows, so capital costs are high, yet the canals may be in use for only a few days a year. The nature of spate means also that risks are high, as the chances of a flood arriving are erratic. Overall, water supply may be less than design predictions, and the spate may arrive at the wrong time in the cropping year. Returns are further depressed by the way in which poor, risk-averse farmers internalize this risk. Unwilling to invest in high-value cash crops that demand expensive seeds and fertilizers, farmers favour production of low-value cereal crops for household consumption. Cropping patterns rarely conform to the ‘high-value mix’ on which improvement projects are predicated.

Third, the capacity of the public sector to manage these schemes had been eroded by the crisis in public budget and services. In 1997, the Tehama Development Authority (TDA), which presides over the sizeable spate schemes in the Tehama wadis, had a large staff for operation and maintenance (32 engineers, 69 technicians and 582 support staff) but a non-staff budget of only Rls 8.5 million ($65,000). The situation was the same in Lahej and Abyan. Most of the spate irrigation infrastructure was deteriorating due to poor maintenance.

Alarmed at the deterioration of these modern schemes, government launched an improvement programme in 2001. The results and lessons are discussed in section 6.3, and more particularly in section 6.3.4 below.

Groundwater use for irrigation developed rapidly and now accounts for two-thirds of the value of crop production

As discussed above, until the 1960s the use of groundwater for irrigation was confined to supplemental irrigation from shallow wells. During the last third of the twentieth century, the tubewell and motor pumps revolutionized Yemeni irrigation. More than 100,000 farm wells are today irrigating 400,000 ha, 40 per cent of the cropped area. Full or supplemental groundwater irrigation now accounts for two-thirds of the value of crop production, and despite ever-increasing pumping depths, groundwater use remains currently financially profitable in many areas and for many crops. At current market prices, tubewell irrigation is viable for vegetables at depths of over 250 m, and for qat at over 300 m depth. Although conveyance is sometimes by pipe over long distances, technology is generally simple, with little localized irrigation, low levels of know-how, and virtually no pressurized irrigation (drip, bubbler) or protected agriculture (plastic houses and tunnels).

The availability of tubewell technology and government’s promotion of groundwater development through the incentive framework encouraged rapid development of tubewell irrigation. There was active government support for groundwater development through projects in the 1970s and 1980s. A favourable incentive structure included cheap diesel, cheap credit, absence of import duties on equipment, an import ban on competing cash crops, and the absence of any constraining regulatory framework. Private capital from remittances also helped many farmers to finance the investment. Profitability was boosted by the rapid growth of markets for irrigated crop production, and in particular by the explosion of demand for qat.

Figure 6.6 This big groundwater irrigation scheme was developed by the government in Wadi Hadramawt to provide farms for poor people. Photograph courtesy of Matthias Grueninger.

The past role of dams in Yemen is discussed in Chapter 5. In recent times, only a single large dam has been constructed: the Ma’rib dam, completed in 1986 with finance from the ruler of Abu Dhabi. Unlike the ancient Ma’rib dam, which was in fact a spate diversion weir,¹⁹ the modern dam is designed to impound water in a reservoir. However, the water available from the dam has never approached the design level of 200 MCM. Up to now, only 1,700 ha out of a command area of 7,000 ha has been developed for irrigation. The dam and irrigation network are managed by the Ministry of Agriculture and Irrigation’s local development authority, ERADA.²⁰

Figure 6.7 Many new hill dams such as this one have been constructed but they often do not increase overall water availability or improve sustainability of the resource. Photograph courtesy of Gerhard Lichtenthaler.

Small dams, largely small check dams and retention structures, have existed in Yemen since the days of the ancient South Arabian kingdoms (see Chapter 5) as a means of improving water control, breaking the force of a spate flow, or enhancing infiltration of groundwater. Recently, government has been promoting small dams in answer to the growing water resource constraint. The programme, which was strongly supported by the then president, has been widely criticized as contributing little to the overall water balance and altering water rights between upstream and downstream. Technical conception and execution, and transparency and accountability in procurement and implementation, have also been faulted.

Three-quarters of Yemen’s farmed land gets less than 600 mm of rain, and inadequate soil moisture constrains production and yield. In such dry conditions, supplemental irrigation²¹ can increase yields significantly if applied during dry spells. Supplemental irrigation is relatively low cost and uses little water, and is already commonly practised. A survey found that supplemental irrigation is quite common in some areas. For example, in Hoban (Ta’iz), where rainfall is 400–600 mm, two supplemental irrigations were applied on 7,000 ha out of a total cropped area of 40,000 ha, doubling net returns for maize.²²

Box 6.4: Qat and supplemental irrigation as a coping strategy.²³

Mawiyya, outside Ta’iz, is in a very low rainfall area – about 300 mm a year on average, at the very margins of agricultural feasibility. Only qat can give farmers enough income to stay in business. With declining water tables, farmers are turning to supplemental irrigation from tankers.

We travelled through Sueda, a rickety market town that has sprung up from nothing in the past five years, thriving on the qat business. Just outside the town, we met Ahmed, a qat farmer. He says his water is dwindling fast: now he can pump for only two hours every 20 days. He grows qat as the only crop that will give him enough return to stay in farming. ‘Without qat,’ he says, ‘we would be dead.’ But to keep his qat going he has to buy water from a tanker regularly. While we are discussing matters, a tanker carrying water for qat arrives and parks on the road next to his farm.

Farmers with stands of largely rainfed qat will watch market prices – and the skies – and will provide a supplemental dose of irrigation water about ten days before they plan to harvest, in order to bring on a flush of new shoots. In some cases, even costly tanker water is used (see Box 6.4). With the growing water shortage in the country, supplemental irrigation techniques are becoming more important.

This section first looks at the current crisis in agricultural use of groundwater (6.3.1). This is followed by an analysis of experience in improving spate irrigation (6.3.2) and attempts to improve agricultural water management in rainfed agriculture (6.3.3). There is a glance at the role of improved agricultural productivity in reducing pressure on water resources (6.3.4). Section 6.4 then examines the potential impacts of these solutions on water use and on farmers.

In recent years, groundwater reserves have been mined on a vast scale, and many springs have dried up

Deep aquifers, tapped over four decades by 100,000 tubewells, have been the source of the groundwater boom that has driven the flourishing of the Yemeni rural economy. In recent years, the boom turned to bust, with pumping depths plummeting several hundred metres, and some reserves completely drained. It has been reported that in the Sana’a basin, if the current rate of overdraft continues, groundwater will be pumped dry in the near future. A recent study concluded that, by 2025–30, depletion of groundwater reserves could reduce agricultural production by more than 40 per cent. Chapter 9 examines this critical challenge in detail, together with related policy options. As groundwater is part of an interconnected hydrological system, groundwater overdraft has also affected other systems, leading to the drying up of most of Yemen’s springs and the waning of the age-old agricultural systems that depended on them.²⁴

The causes lie in the incentive structure, in stagnating agricultural productivity, and in institutions

As discussed above (section 6.2.2), the incentive structure provided a powerful stimulus to groundwater development but gave scant incentives for efficient use. At the same time, there are no communal or governmental institutions that can regulate water use: no defined property rights, no regulatory system. Each individual irrigator competes to drain out as much of this common resource as possible as quickly as possible – ‘the race to the bottom’.

As early as the 1990s, policymakers and planners were aware that the groundwater boom could not last, and that groundwater mining not only threatened the future of the agricultural economy but was also yielding incomes and financial capital accumulation well below the true value of the mined resource. Adjustments to the incentive structure were made: in a first round, subsidized credit for groundwater development was ended. Then, in line with good practice to separate the water resource ‘manager’ from the water ‘user’, responsibility for water resources planning and allocation was transferred from the Ministry of Agriculture and Irrigation (the user) to the NWRA (the manager).²⁵ Programmes were launched to improve the efficiency of water use. By the mid-1990s, a vision of the need for a three-part solution – incentives, technical solutions, institutional change – was being articulated in policy analysis.²⁶ These three solutions were incorporated into the 2004 national water strategy, NWSSIP (Box 6.5). At the same time, a series of field programmes was launched to test out some or all of these solutions on the ground. The following paragraphs describe some of these experiences.

The 1993 Land & Water Conservation Project (LWCP) subsidized pipes for groundwater farmers to reduce water losses in conveyance and distribution. The project covered almost 11,000 ha (3 per cent of the groundwater-irrigated area). It was reported that water losses in the area dropped by 20 per cent and that most of this ‘saved’ water stayed in the aquifer and was not used for other production, as farmers wanted to save money on pumping and also were afraid they would have to repeatedly deepen their wells if they did not reduce pumping. However, some farmers certainly did use the ‘saved’ water to expand their irrigated area, or even sold it to the tanker trade. Although the payback period on the investment was reported to be only two years, there was almost no spontaneous take-up of the technology by farmers outside the project. Little interest was aroused by a small component to test pressurized irrigation (drip, sprinkler, bubbler, see Figure 6.8).

Box 6.5: NWSSIP’s three-part solution.

The 2004 national water strategy (NWSSIP) retained three sets of solutions to the crisis in the agricultural use of groundwater. First, economic solutions would reduce incentives to water mining and inefficient water use, and encourage farmers to strive for ‘more crop per drop’. Then, technical solutions would develop and disseminate the technology needed to maximize water use efficiency and improve the value added. Finally, institutional solutions would develop an institutional structure to allow farmers to understand the challenge of groundwater depletion and to take joint action to control it.

Figure 6.8 Unusually among Yemenis, these farmers have adopted efficient drip irrigation – the drip lines are visible between the trees – which gives them more income and leisure for qat. Photograph courtesy of Gerhard Lichtenthaler.

The follow-up Groundwater & Soil Conservation Project (GSCP) 2003–12 expanded the piped conveyance packages to a further 51,000 ha (about 12 per cent of the groundwater-irrigated area). Three key innovations were added to the LWCP ‘model’: (i) an irrigation advisory service was set up to advise farmers on water management and irrigated agriculture; (ii) participants were required to form a water user association (WUA) as a counterpart to the project; and (iii) a pro-poor emphasis was introduced through slightly larger levels of subsidy for smaller farms.

The GSCP programme proved popular, with demand being ‘five times’ availability (Box 6.11). The 2008 poverty and social impact analysis (PSIA)²⁷ found water savings of up to 40–50 per cent being reported, but little improvement in on-farm water use efficiency (Box 6.6), and with the better-off cornering the lion’s share (Box 6.17). A pressurized irrigation component again found few takers. Surveys at project completion in 2012 found ‘annual savings of 83 MCM’ of groundwater (about 3 per cent of total agricultural use nationwide) and increases in farmers’ incomes of up to one-quarter due to lower pumping costs (46 per cent reduction). However, there was no observed impact on aquifer depletion, and no improvement in water use efficiency – water consumption per hectare actually increased. The project did not succeed in setting up institutions for sustainable aquifer management.²⁸

A more elaborate approach was adopted for the Ta’iz Water Supply Pilot Project (TWSPP) 1996–2000. Here Ta’iz city wanted to transfer water from Habir, a nearby rural area, and offered a package of ‘compensation’, including support for the formation of WUAs, investment support for water conservation and efficiency in both groundwater and spate, and help to improve farming techniques and cropping patterns. The World Bank financed part of the physical investments and all the compensation and management support. Numerous problems arose during implementation: the water resource proved less than expected; local people proved uncooperative and often hostile, as they said it was sheikhs and not they who had made the deal; and the World Bank, after an intensive effort during 1996–9 to help WUAs and water conservation efforts to work, changed tack mid-term and insisted that government complete the engineering investments as quickly as possible with scant further attention to the social, economic and institutional developments in the Habir source area.

Box 6.6: In Wadi ‘Arafa, the Groundwater & Soil Conservation Project (GSCP) contributes to water saving – but lack of knowledge keeps productivity below potential.²⁹

Wadi ‘Arafa is a dry area of granite outcrops. The GSCP farm is just beside the road. It is a demonstration farm, and the farmer has received pipes and drip irrigation on half his farm. The well dates from 1990, and has been deepened from 80 m to 120 m. The farmer is keen on GSCP: he reckons that he has reduced his water pumping by 40 per cent, and the irrigation time that used to be 12 hours is now six hours. He has saved on labour costs, too: he employed ten labourers before, now he employs only five. Production has gone up as well: in the season, he is harvesting 100 baskets of tomatoes every three days compared to 70 before. However, he has changed nothing in his cropping pattern, nor in his production practices, and he has learned how to handle drip irrigation by trial and error.

A more integrated but very small pilot project – the Community Water Management Project (CWMP) – was implemented in 2005–9. The objective was to test a replicable model for sustainable self-management of groundwater by WUAs that represent all water users in a discrete hydrological area. The project helped set up nine WUAs in three areas and gave them institutional support and capacity building so that they acquired skills in WUA management. The heart of the project was then for the WUAs and the project to work with NWRA and GSCP to define the water balance and to prepare and execute a water management plan targeting a specific ‘water management goal’ – effectively reducing pumping to more sustainable levels while sustaining or improving incomes. Investment support from GSCP was made available so that water-saving investments could be implemented under the water management plan. The project is complete but was too short to create the required institutional capacity, and in fact most time was spent by the project learning ‘how to do it’. Continuation, capitalization and scaling up were needed if anything specific was to be achieved.

Yemen’s most ambitious attempt to solve the groundwater problem in agriculture took place as part of a water basin plan for the national capital area, the Sana’a Basin Water Management Project (SBWMP). In 2003, Yemen committed itself to a long-term process of integrated water resources management in the basin, where critical water problems were pressing. A 15-year time horizon was selected, and divided into three five-year phases. The World Bank agreed to finance the full 15-year programme. The first phase, designed to test technologies, demand and supply management approaches, and institutional arrangements, was implemented under the SBWMP. The governance and management model adopted was essentially a fully integrated water resources management (IWRM) model.³⁰

Implementation started slowly but the first phase was completed in 2010. The governance structures were set up, water resources were assessed and monitored, a master plan was prepared, public awareness campaigns were conducted, and regulatory measures were implemented. Recharge structures were built to enhance supply. On the demand side, an intensive effort was made to work with WUAs on sustainable water management and on water use efficiency in agriculture. Investments were made in piped conveyance (10,000 ha) and in localized on-farm systems, typically drip irrigation (8,700 ha). Support was provided on water use efficiency and agricultural productivity. Project surveys showed that farmers could – and did – pump 40 per cent less water and still increase their incomes by 10 per cent.

In summary, the first phase piloted IWRM approaches in the Sana’a basin, and demonstrated that many of them can work. Sadly, there has been little initiative to continue the experience subsequent to the closing of the first phase implemented under the SBWMP project.³¹

There is no guarantee that water ‘saved’ would not be reused by the farmers

Although all the projects discussed above report ‘water savings’, there is also anecdotal evidence which suggests that Jevons’ Paradox may often apply: that any improvement in efficiency of use of a natural resource will result in more use (as it is more profitable).

In only a few cases have project-initiated WUAs cooperated on water management or had any corporate life beyond the transaction with the programme bringing subsidies.

Groundwater is a common pool resource and so all users must cooperate in water saving

Programmes so far have essentially addressed the agricultural water behaviour of individual farmers. However, groundwater can hardly be ‘saved’ if only one farmer reduces his pumping, as groundwater ‘flows’ and savings by one farmer will be available for a neighbour to pump out. Real savings are only possible where all water users in some discrete hydrological unit agree together on reduced levels of pumping.

Some observers have questioned how much water is actually saved anyway (see Box 6.15). Programmes often replace earth channel conveyance from the wellhead by piped conveyance. However, most of the transmission losses from earth channels merely seep back into the water table and so are not lost to the hydrological cycle at all. On this interpretation, the main ‘saving’ of the groundwater conservation programmes is the reduction in pumping cost.³²

Bad money drives out good. It may be that the public subsidy is ‘crowding out’ private initiative. In any case, the rationale for a public subsidy of essentially private benefits is not really clear (see below on this).

Despite government and donor commitment to NWSSIP’s three-part solution and some apparent success, there has been a considerable amount of stop-start

Projects have been abandoned before they have had a chance to succeed. Clearly, tackling the groundwater problem is a long-term business, and approaches have to be tested, scaled up and applied consistently over many years.

Arguments for public subsidy of groundwater irrigation improvement have proved questionable

Three arguments are used to justify subsidy: (i) Yemeni farmers would only invest in water saving if there were a subsidy incentive; (ii) there is a public interest in saving groundwater; and (iii) groundwater programmes can be pro-poor. Each of these arguments has proved open to question.

The investment appears profitable, so farmers may not need the subsidy incentive – a credit programme might be more efficient

If, as project figures show, farmers can recover their investment in water saving through reduced pumping in just two years (see Box 6.9), then farmers should be encouraged and enabled to invest themselves. If the problem is not lack of incentives but lack of capital, then a credit programme would be more efficient because it could be market based and would not create the kinds of moral hazard that subsidies do.

Box 6.7: A useful rule of thumb for evaluating the financial benefits of water saving.³³

Under LWCP, government and farmers invested $250/ha to achieve water savings of approximately 2,300 m³/ha/year. The investment costs were thus about 11 cents for every m³ of annual water saving. The annual savings in pumping costs averaged 6 cents/m³. Thus the investment cost could be recouped in two years on the basis of saved diesel cost alone, without taking account of the opportunity cost of the water saved in the aquifer.

Unless the saved water can be reclaimed for the public good, it is hard to justify subsidy on the argument of public interest

There might be a public interest, given that groundwater in the aquifer has some of the character of a common resource: it can be accessed by many people and may be appropriated by the state. Certainly, the ‘groundwater crisis’ is discussed in Yemen as though it were a public good crisis. However, in Yemen the absence of regulation means that access to groundwater is an open free-for-all. Open access to a finite resource is essentially a form of private property with a floating title that can be asserted by the early bird. He who pumps deepest and fastest is the owner. Unless regulation can be imposed, and the groundwater resource reclaimed for the public good, subsidies essentially enhance the value of a private good.

Alternatively, the anti-poor impacts of groundwater mining and of conservation programmes could make a case for pro-poor subsidies

If subsidies really help the poor, they could be justified. Groundwater overdrafting certainly has an impoverishing effect, and this effect is greater for the poor. In addition, some of the instruments used to combat groundwater overdraft have a preponderant negative impact on the poor. For example, reducing the implicit diesel subsidy may reduce incentives to groundwater overdraft but has a negative impact on incomes through consumer price rises. Even irrigation improvement tends to reduce agricultural employment on which the poor depend.³⁴

A pro-poor bias was in theory built into the government groundwater conservation programmes

In order to impose a pro-poor bias under LWCP, a ceiling was placed on the area that the project would co-finance. Under GSCP, this was modified so that the subsidy increased for farmers with smaller farms. These measures were designed to ensure that smaller, supposedly poorer, farmers got the larger share of subsidy, but that all farmers had an incentive to join the programme.

However, the poorest have little or no access to irrigation and, in practice, skewed access to groundwater and the tendency of subsidies to go to the more powerful undermines the pro-poor justification for subsidies

Designing pro-poor subsidies for water conservation is difficult: as the poor have less water so they would receive less subsidy. In fact, as only a fifth of people in poor areas have any irrigation at all (see section 6.1 above), most poor would get nothing. In practice, it is hard to find evidence that the poor benefit from water conservation subsidies – and there are plenty of examples of the better-off collaring the largest share (see Box 6.17).

In any case, experience in Yemen in general is that, whatever the safeguards, any subsidized programme will tend to mainly benefit the better-off

‘Subsidy,’ the Minister of Water said in 2006³⁵ – speaking ‘as a citizen and not as a minister’ – ‘subsidy in Yemen is a form of corruption that does not reach the deserving people. There is a budget of $600 million for subsidies. None of it benefits the poor.’ Clearly, the rationale for any subsidized programme needs to be carefully thought out, the targeting mechanism carefully designed, and the programme transparently implemented – difficult conditions to satisfy in Yemen’s political economy.

Water use efficiency is likely to become a profitable investment as scarcity grows

Investment in irrigation efficiency should be profitable, and Yemeni farmers do invest in piped conveyance without subsidy. There is certainly scope to promote – or simply allow – the development of private markets.

One aspect of this is to level the playing field for private markets to develop

There is good experience internationally with development of commercial markets for irrigation efficiency equipment, including local manufacturing of pipes and filters. In India, for example, the private sector has shown considerable initiative, bringing costs down and generating considerable business (Box 6.8). Equipment is already manufactured in Yemen at prices reported to be cheaper than imports. Relatively small reductions in cost could make a substantial impact on the viability of improved irrigation and the speed with which it is adopted. When established, the manufacturers of improved irrigation equipment could become a major force for promoting uptake. The lessons are that subsidies should not ‘crowd out’ private sector initiative – but they might be used to promote it by encouraging competition and innovation.

Box 6.8: Irrigation efficiency subsidies slow adoption of drip technology.³⁶

Drip and sprinkler technologies were aggressively promoted in India from the mid-1980s; yet, by 2002, the area using them was only 60,000 ha. One problem was subsidies, channelled through the big irrigation equipment companies. Their equipment typically cost $1,750/ha, which put it out of the reach of most farmers – except the few who could access the subsidy programmes. These subsidies directed at branded, quality-assured systems prevented viable, market-based solutions from emerging.

Fortunately a grey market of unbranded products began to offer drip systems at $350/ha. Then, one innovative manufacturer introduced a new product labelled ‘Pepsi’ – basically a disposable drip irrigation system consisting of a lateral with holes. At $90/ha, Pepsi cost a fraction of the price of all other systems.

6.3.2 Improving the performance of spate: the Irrigation Improvement Programme

Alarmed at the deterioration of modern spate schemes, government launched an improvement programme in 2001 that incorporated best practice on participatory irrigation management (PIM)

The Irrigation Improvement Programme (IIP) was conceived as a 15-year programme (2001–16) to rehabilitate and modernize all of Yemen’s major spate schemes, and to develop sustainable management. Responsibility for running the schemes would be progressively taken over by farmers. The objectives of a first phase project were to: (i) improve water service through increased reliability and efficiency of the physical structures; (ii) improve productivity, helping farmers to increase income per drop through improvements to water use efficiency and adoption of better water management, higher value cropping pattern and improved crop husbandry; (iii) improve the efficiency of operation and maintenance of the schemes by promoting decentralized management by WUAs, which would gradually take over operation and maintenance of the tertiary and secondary canal systems; and (iv) improve the financial viability of spate schemes by getting WUAs to share in the rehabilitation and operation and management costs of the schemes. The last pair of objectives built on principles developed through the worldwide movement for participatory irrigation management (PIM).³⁷

IIP was implemented 2001–9 in two of the eight main spate wadis: Wadi Zabid in the Tehama, and Wadi Tuban in Lahej. The legal framework for participatory management was established, and WUAs were set up at the level of each command area, and Irrigation Councils at the level of the scheme.³⁸ By 2009, the two schemes had been rehabilitated, an agricultural demonstration programme under farmer management on 5,000 ha had been set up, and WUAs had been established.³⁹

There were positive results from the participation and productivity measures under the programme

All farmers joined WUAs, which had clear responsibilities. Irrigation Councils were set up as the governance bodies. Before the project ended, WUAs were progressively taking over responsibility for operation and management of the secondary and tertiary systems. In Wadi Zabid, the associations were even able to reconcile the interests of small and large farmers, working with larger farmers to remove the control works that had been placed in the canals to increase the large farmers’ irrigated area. The programme produced good production results, too. Actual production increases exceeded expectations (Table 6.4).

Table 6.4 Planned versus actual yield increases under the Irrigation Improvement Programme.⁴⁰

Careful design and implementation were at the heart of the success of the institutional component

Information about the project was well transmitted, with farmer ownership emphasized. The institutional design was carefully constructed, initiatives to participate were built in, and capacity-building was continuous. The project unit prepared the institutional development carefully, with well thought out manuals and training programmes, and the project team worked hard to make the popular community contracting initiative a reality.

Overall, IIP showed the potential for decentralizing spate irrigation management on a participatory basis

The project showed that participation and cost-sharing create ownership. WUAs showed potential, and the project showed that incentives to participate in them include: (1) public awareness; (2) empowerment through participation in design, supervision, etc.; and (3) monitoring and evaluation (M&E) that is shared and can show results. Clearly, both WUAs and Irrigation Councils can work but they require lots of support, and it is a slow process. Finally, the project did begin to reduce the burden on the exchequer through the higher level of cost sharing.

However, only time will tell whether these early positive results can be consolidated

Whether the participatory model can resolve the classic upstream/downstream, large farmer/small farmer inequities of spate is not clear. Overall, a recent evaluation concluded that ‘it is too early to draw conclusions on the quality of irrigation services provided by the Irrigation Councils and WUAs, compared to those previously provided by government entities.’⁴²

More alarming than this modest diffidence is lack of follow-up, despite the good results of the project

The implementation unit was disbanded at project completion, and the World Bank did not finance a follow-up project, despite its original commitment to finance three successive projects.⁴³ Notwithstanding a commitment to a long-term strategic approach, government and donors have not yet worked out how to learn lessons and scale them up in a long-term investment programme for spate any more than they have for groundwater.

With the dwindling of groundwater, terraces could be a recourse for farmers, although they would probably need technical and financial support

As discussed in section 6.2.2, rainfed agriculture has struggled to compete with subsidized groundwater irrigation. However, with the crisis in irrigation there is interest in reviving rainfed systems. Although a host of social, land tenure, technical and economic constraints have been identified, some experience already exists with terrace rehabilitation and there may be scope for further support. The GSCP invested $12.2 million over seven years in terrace rehabilitation, wadi bank protection and small-scale spate improvement, water storage tanks and traditional water harvesting, at an average cost of about $2,500/ha. These costs are not affordable to poor farmers, but there is a public interest both in the environmental and the amenity functions of terraces that justifies at least partial subsidy.

In catchments affected by water shortages and by erosion and poverty, a full watershed management approach may be appropriate

A key tool in managing water resources in poor areas with severe land and water resource depletion problems is the watershed management approach, which integrates management and use of land, vegetation and water in a catchment with the objective of protecting or conserving the hydrologic services that the watershed provides and of reducing or avoiding negative downstream or groundwater impacts. Attempts have been made in the past to introduce combined soil and water conservation approaches in Yemen but there has been no programme that has scaled this up to the level of an entire vulnerable catchment. A priority would therefore be to introduce watershed management approaches in catchments where there are notable water shortages, erosion and poverty.

This section has examined reforms and programmes aimed primarily at improving the efficiency of water use in agriculture. A complementary area for improvement is to raise farmer incomes by broader productivity improvements – adapted crops and cropping patterns, improved varieties, improved husbandry techniques – and by better access to finance, infrastructure, etc. All of these can help farmers to increase productivity and so sustain incomes whilst reducing pressure on water resources. This discussion, however, lies outside the scope of this book.

The national water strategy (NWSSIP) targeted greater agricultural water efficiency and reduction in groundwater depletion

As discussed above, Yemen has been reforming agricultural water management for more than a decade. Underlying this effort is the hypothesis that, through a combination of efficiency improvements and better water management, it should be possible to maintain or even increase agricultural incomes whilst reducing the rate of groundwater mining. The NWSSIP’s three-part solution to the challenge – incentives, technical solutions, and institutions – was set out above (Box 6.5). This section assesses results to date against these three sets of measures. First, results of the measures introduced to re-orientate the incentive structure towards water conservation in agriculture are discussed (section 6.4.1). This is followed by an assessment of results of measures to increase water use efficiency and raise overall factor productivity in agriculture (section 6.4.2). Results of institutional innovations, particularly water user associations (WUAs), are then discussed (section 6.4.3). Finally, the impact of reform implementation to date on farm incomes and on the water resource is summarized (section 6.4.4).

6.4.1 Revision of the economic incentive structure to promote agricultural water use efficiency and increase income per drop

Government has been implementing the NWSSIP programme to adjust incentives. Diesel price rises doubled the cost of water in 2005 and this sent signals to farmers to change their behaviour

The elimination of price incentives to groundwater overuse (diesel price, credit price, import bans) was long held to be essential in order to improve the sustainability of groundwater irrigation. Under pressure from international institutions, government raised the diesel price from Rls 17/litre to Rls 35/litre in mid-2005. This decision produced some protests and 35 demonstrators were killed. The measure, although still inadequate to bring prices to border parity level, had an immediate effect – the cost of producing water doubled.⁴⁴ The price of water in local water sales between farmers and to tankers also went up sharply. For the first time, water seemed expensive.

Under pressure of the price rises, farmers developed coping strategies: switch to qat, and investment in piped and pressurized irrigation, sometimes associated with a reduction in hired labour

Farmers adopted various coping strategies to maintain their incomes. Those who could grow qat and were not yet doing so became more likely to do so: in many areas, qat became the only crop that could pay the cost of water. Box 6.9 describes the emergence of a qat monoculture in low rainfall areas around Ta’iz. A few farmers invested in piped conveyance and controlled on-farm irrigation, some even in greenhouses, thereby reducing water losses and increasing returns per unit of water (Box 6.10). Demand for public programmes like GSCP increased (Box 6.11). Improved irrigation was also associated with a decrease in labour use, as shorter irrigation times and automatic conveyance require less manual intervention (see Box 6.17 for a case in Tehama).⁴⁵

Box 6.9: Qat and groundwater.⁴⁶

We arrived at the village of Qarya al Guneid a few miles north-east of Ta’iz. This is a granite area with pockets of water in fissures only. The main crop here is qat. It is the low-bush variety, which produces three harvests a year. The first, al muharra, is the best quality: it requires up to 12 irrigations; the second, qatl, soon follows: it produces only small pieces and requires only one or two irrigations. The third, jumum, requires six irrigations.

The first farmer we meet grows some qat himself. He also leases part of his land and water to a ‘qat contractor’ and gets three-quarters of the crop in return. He is one of nine brothers who own wells and land: they supply water to a total of 40 farmers, and are farmers themselves. He has never heard of government programmes for supporting irrigation improvement but would consider ‘joining a qat water user association’. He has a well that is 50 m deep – it used to be 22 m. Deep drilling is needed in some parts of the area now, up to 350 m. He started planting qat 20 years ago on a pilot basis, and has extended it progressively. He has never seen an extension worker, but learned farming from his grandfather, and has learned qat farming by trial and error. He buys his chemicals – Saudi fertilizer and German pesticides – from Ta’iz, and uses a lot because they ‘make the leaves a nice shape’. There is also a green larva that he sprays against. Before, he grew other crops – maize, corn, mangoes, papaya, potatoes, tomatoes – but qat is the most profitable.

The second farmer, Nabil, is young and unmarried. He says: ‘I studied in Ta’iz, but I couldn’t find a job, so had to start on qat.’ Until two years ago, he says, anybody could drill. ‘It was good to stop it, but now it is anyway too expensive – around YR 8-9 million [over $40,000] to get to the depth needed. If there is a dispute over water […] well, there is a small department at the Local Council to complain to.’ But he doesn’t look convinced. Qat is profitable for him but margins are not always so large in this water short area. He may clear Rls 300,000–600,000 ($1,500–2,000) from his qat, with expenses of Rls 250,000 ($1,250). But if he can sell to Saudi, then he can get Rls 700,000 ($3,500). There are some landless in the village. Some can find opportunities as sharecroppers on qat. But labour opportunities are limited as most farmers try to do all their own qat work, except for picking, which is often done by women or children because they are cheaper.

A third farmer, Ahmed, has 80 rows of qat about 50 m in length, and estimated his land at half a hectare. Last season, he grossed Rls 600,000 ($3,000) from al muharra, Rls 100,000 ($500) from qatl, and Rls 300,000 ($1,500) from jumum, a total of Rls 1 million ($5,000). But since water became so scarce and the diesel price went up, he has had to abandon some of the land that he was previously contracting on. Now he is paying Rls 2,800 ($14) an hour for water, and each irrigation is 15 hours, so that his water cost alone is (2800 × 15 × 20) = Rls 840,000 ($4,200) a year, leaving him a net of only Rls 160,000 ($800).

All these farmers were soaking the earth with furrow irrigation. None of them had considered, or even heard of, drip irrigation. Plainly these farmers are coping as best they can. For them, qat is a survival strategy and they get no help of any kind. There seems to be a case to treat qat as a crop here: to do some research and extension, to help particularly on water management, even include qat in government-supported water saving programmes.

When the diesel price rose, rural people generally reduced their water consumption.⁴⁷ Those who bought water for farming became less likely to do so. In very water-stressed areas, even qat could not pay the water bill: see, for example, the case of Ahmed in Box 6.9, whose gross income from qat is Rls 1 million – of which four-fifths goes to pay his water bill. Not surprisingly, he has had to reduce his water purchases and to cut production.

Box 6.10: The diesel price rise bites in Sa’ada.⁴⁸

The Sa’ada basin committee is looking for solutions to one of the worst groundwater depletion problems in the world. They said outright: ‘The increase in diesel price helped us a lot. The rate of overdraft has slowed (the water table is now dropping by “only” two metres a year).’ They see the farmers’ salvation in modern irrigation, with subsidized equipment under GSCP. The farmers are ready. Now the problem is that GSCP cannot keep up with demand, which is ‘five times greater than what the project is doing’.

6.4.2 Technical solutions to maximize water use efficiency and raise overall factor productivity

Government’s reforms for irrigated agriculture are based on the assumption that improved agricultural productivity and irrigation efficiency can produce ‘more income per drop’. However, work on the different components of agricultural productivity is going on in Yemen but has, so far, had little impact.

Programmes to improve water productivity in agriculture are being implemented by the Ministry of Agriculture and Irrigation. The Agricultural Research and Extension Authority (AREA) and some regional development programmes are doing research and developing extension for irrigated agriculture, although attempts in recent years to develop and disseminate high productivity technology and more adapted varieties have had little result. Apart from a brief flurry of activity with ICARDA guidance in the 1990s, when a ‘rapid impact programme’ was implemented, the agricultural research and extension apparatus has had scant impact.

Our 2007 PSIA study found scant evidence of any results available at field level. See, for example, the difficulties go-ahead farmers had in sourcing advice (Box 6.11). Generally, except in the context of specific projects (see below), the study found that water management advice and high-value cropping packages that can genuinely produce more income for less water are not available, and extension outreach is very limited. Very uncertain internal and external market conditions also create risk aversion among farmers and constrain the development of higher value cropping.

Defining irrigation efficiency as the ratio between water withdrawn and water beneficially used by the plant, overall efficiency in irrigation in Yemen is estimated to be about 25 per cent – among the lowest in the world (average for all developing countries is 38 per cent, and for the Middle East and North Africa 40 per cent.⁴⁹

Box 6.11: The challenge of improving the returns to water in agriculture: farmers in Lahej explain the difficulty of getting more ‘income per drop.⁵⁰

In Wadi Tuban, we visited a demonstration farm for GSCP. The farm is 10 ha irrigated by three wells in the alluvial aquifer. The wells are about 100 m deep, deepened from 70 m some time ago. The young farmers – Yahya Mohammad Ali and his brother Arafat – share the land and are irrigating mangoes with bubbler. They say that they see the irrigation extensionist from the project every ten days or so, and the agricultural extensionist somewhat less often. The irrigation extensionist, who is from the Irrigation Advisory service of the GSCP, is present at our meeting. He says that at GSCP they have some brochures on water management but they have no prior knowledge of irrigation intervals or quantities for bubbler on mangoes, and they are proceeding by trial and error, gradually reducing the doses. The extensionist does not seem to have access to a soil moisture testing kit, so it is not clear how he is measuring whether the trees are getting the correct dose. When the farmers ask some probing questions about irrigation intervals, the extensionist cryptically replies that ‘leem (citrus) likes to be kept thirsty’. The brothers look doubtful.

The projects described in this chapter are certainly improving irrigation efficiency but much more could be achieved

As discussed above, a first round reduction of water losses and reduction of diesel costs is readily achieved, simply by improving the efficiency of conveyance and on-farm distribution. All the farmers interviewed during the PSIA study attested to very large water savings, typically 40–50 per cent.⁵¹ However, the real potential of modern irrigation technology – improved water management, higher value crops, improvements in crop husbandry and post-harvest handling, market development – has scarcely been touched on by the projects. The farmers interviewed during the study had generally done little or nothing to improve their farming, leaving the impression that, so far at least, projects were just about pipes – see for example, the case in Wadi ‘Arafa (Box 6.6).

6.4.3 Institutional measures for managing groundwater and controlling depletion

Both international best practice and the reality of groundwater management in Yemen suggest that water user associations must be an important component of the answer to the groundwater crisis

Three powerful arguments have driven planners in Yemen to focus on WUAs as an essential institutional response to water crisis, particularly the water crisis in agriculture. First is the lesson from international best practice that water is best managed at the lowest practicable level.⁵² Second is the fundamental reality of water resources management in Yemen – that government scarcely controls any water, least of all groundwater, which is managed in the real world by more than 100,000 well-owners in the country. Third is the historical tradition that suggests that Yemenis are adept at communal institutions for water resources management.⁵³

Box 6.12: Water user associations in Yemen are of various types.

• Irrigation WUAs, being promoted under the IIP, GSCP and SBWMP for both surface and groundwater. Registered as NGOs, the WUAs set up under IIP are responsible for operation and maintenance at tertiary canal level (and possibly soon at secondary canal level). Under the GSCP and SBWMP, water user groups (WUGs) were formed around a common well. In addition, a WUA is formed of WUGs or among individual beneficiaries in each area for common purposes, largely training.

• Rural water supply WUAs have long existed to manage community-based schemes and are now actively promoted under public or NGO projects.⁵⁴

• Water management WUAs, being piloted by NWRA under the Ta’iz basin plan in al Haima District. Registered as NGOs, 24 WUGs are federated into one WUA and are being trained to carry out local-level water management tasks, including awareness, water monitoring, and preparation for water conservation programmes such as GSCP.

• Full service WUAs piloted under the CWMP. Irrigation farmers in a discrete water management area were groomed for management of the water resource.

As a result, agencies and projects have been promoting the establishment of WUAs for water resources management, irrigation and rural water supply

‘Modern’ WUAs are being promoted, particularly under donor-financed projects – see section 6.3. These WUAs are of many types, differing in their roles and in the type of water resource they look after (Box 6.12).

Experience to date shows that WUAs can help farmers, particularly smaller farmers, in several ways: to access public programmes, to manage all or part of an irrigation system collectively, to act as a solidarity mechanism, and to manage water resources

At the most basic level, WUAs can act as a means of accessing public programmes for subsidized investments, training, etc. (see Box 6.13). Particularly in surface irrigation systems – springs or spate – WUAs may actually collectively manage an irrigation system, or part of one. WUAs can also act as a solidarity mechanism, allowing water users to take collective action against more powerful interests – see the example in Box 6.14 of the Falej WUA acting against resource capture (unsuccessfully as it happened, demonstrating the need for a better functioning governance and regulatory system). More evolved WUAs may act as embryonic ‘water management agencies’, beginning to set their own rules, e.g., about well-drilling and deepening, the number of pumping hours, etc. (see Box 6.13). In Wadi Warazan, a WUA official mentioned, ‘We impose a fine of 5,000 Rials on anyone taking water illegally’.⁵⁵

Box 6.13: Beit ‘Ithrib water user association.⁵⁶

At Beit ‘Ithrib in the Sana’a basin, 72 farmers got together in five water user groups and federated into a single water user association with the help of SBWMP. Only a few members have yet got subsidized water saving investments under the project, but those who have done so already report excellent results: lower costs, higher productivity and a 50 per cent saving in water. One farmer says he has reduced his pumping hours per libna from 20 to 4.5.

At first, some of the farmers are not very clear what the WUA is for, once they have got the subsidized equipment, but then the head of one WUG (jamaiyya al-mahjal) speaks up: ‘The first idea is awareness,’ he says, ‘awareness and monitoring. We shall watch each other. Expansion of the cropped area will not be allowed. We are all aware of the problem.’ Then another farmer speaks up: ‘Our objective,’ he announces, ‘is water conservation […] not expanding […] we want modern irrigation, lower costs, higher income… .’

When asked if they would all reduce pumping together, they say yes – but only when they have the modern irrigation equipment. And that, sadly, is more than a year late in being delivered to them.

There are risks that water user associations may be set up simply to access subsidies, or that they may not be empowered or useful to their members, and so may fail

There are risks – as with WUA programmes in many countries – of going too fast and of overloading these organizations. Members of the IIP WUAs (see section 6.3.2), for example, consider that the WUAs will only survive if they have some water to manage, and if institutional support continues for an extended period. During the PSIA visits, the then governor of Abyan, who had enormous experience of irrigation management,⁵⁷ said that WUAs had their limits. In spate irrigation, for example, he said they could manage the lower level system, but WUAs managing secondary canals and above would be very risky. ‘And irrigation management transfer in spate is a fantasy.’ There is a risk, too, of what the chairman of the farmers’ union called derisively ‘imaginary WUAs’, set up under projects to chase benefits. He might have had in mind the type of WUA the PSIA study saw in Wadi Siham, where the ‘member’ even forgot that it existed until prompted (see Box 6.17).⁵⁸

Water user associations also require a supportive and empowering governance and regulatory environment

WUAs would also work better – and stand more chance of becoming ‘water managers’ – if the governance and regulatory environment improves. In the case of Falej (Box 6.14), the WUA would plainly have been strengthened if its attempts to complain had been successful. The Chairman of Parliament’s Agriculture Committee told me in 2006:

WUAs alone cannot control water management. Enforcement and monitoring and supervision are required from NWRA and MWE. The police and attorney general’s office need to be involved […] cooperation is required from Security […] and support from the local councils.⁵⁹

One of the case studies on water conflict in Chapter 10 records the confusion of spate farmers in Wadi Tuban over where to complain – the Irrigation Council, the Ministry of Agriculture’s Irrigation Department, the local authorities – confirming their lack of knowledge of the rules, and their ultimate acceptance that essentially ‘water flows upstream to the powerful’.⁶⁰

Box 6.14: Water user associations and water management.⁶¹

The village of Falej in Wadi Kabir depends both on spate and on wells sunk in the alluvial aquifer, which receives some recharge from the floods each year. Thirty farmers organized into a WUA under IIP. They are concerned that the level in wells is dropping. They blame excessive upstream extraction of groundwater – the groundwater flows down the wadi just like surface water. They also blame excessive upstream spate diversion. One farmer says he has had no spate water on his land since 1997, while upstream farmers are ‘taking twelve irrigations’. When the WUA went to talk to the Complaints Committee of the Lahej Irrigation Council, the upstreamers – from Khalaf, Hussein and Habil – ‘shot at them’.

Despite these problems, WUA members are hopeful they will get spate water next season. The structures have recently been rehabilitated, and they are waiting to see if the upstreamers ‘continue to take all the water’. They believe management of spate needs backing from police and the courts – but they have little confidence that this will happen.

In addition, there are challenges when more powerful interests are joined to less powerful ones

There can be a problem of equity within WUAs: the basis of a WUA is a ‘democratic’ one, but large farmers and water resource owners may not join. This problem undermined the WUAs set up in the mid-1990s under the TWSPP (see section 6.3.2) – the sheikhs who controlled most of the wells would not join. And when larger farmers do join in, as in Wadi Tuban and Wadi Zabid under IIP, the problem is how to avoid their controlling the decision-taking. The experience in obtaining large farmer cooperation in Wadi Zabid (see section 6.3.3) suggests the possibility of reconciling large and smaller farmer interests – but it will be uphill work.

A variety of different user associations is being promoted, ranging from loose groupings for the purposes of project-related training to associations of water users that may have the capacity to become field-level managers of water resources in their area. Heterogeneity of this nature is expected – and valuable – during the first phase of WUA development. There is, at the same time, a need for cross-fertilization, learning and the application of lessons to build best-practice approaches. Ultimately, this process should lead to some alignment on common practices and to legislation or bylaws confirming the responsibilities and powers of WUAs. Close attention is needed to the purpose and sustainability of WUAs: if they are to be effective in water management, they have to provide a service that members value, and if they are to be sustained they need ongoing support.

A process of cooperative learning would help draw lessons and implement best practice

A process of monitoring and studying WUAs could be set up. The objective would be to develop a typology of WUAs based on different roles and functions for water resources management, irrigation and water supply, to define conditions for success – learning from experience from technical, institutional, management and capacity building aspects – and to draw up a methodology for setting up WUAs and supporting them. Current pilot experiences in participatory groundwater management should be monitored and built on, and a broader series of pilots or regional programmes should be built up progressively.

In addition, an enabling environment for community-led solutions and for community/agency partnership approaches is needed

The WUA model applies well to surface irrigation schemes, but less well to the inherently individualistic groundwater management. Attempts such as CWMP and SBWMP to apply the formal WUA model to groundwater have brought some positive results but it is likely, too, that a multiplicity of local-level institutional solutions will emerge. Chapter 11 looks at some of these, and at the need for mutual learning and an open enabling environment. A variety of partnership approaches looks like the best solution to Yemen’s water resources management challenge.

6.4.4 Assessment of overall results to date on the water resource and on incomes

Overall findings to date suggest that with increased efficiency of water use, farmer incomes can rise while using less water – but real water savings are not clear

The results of GSCP were reported in section 6.3.1: 83 MCM of annual savings and increases in farm incomes of up to one-quarter. These results broadly match those of SBWMP (see section 6.3.2 above), where a survey found water savings of about 40 per cent due to irrigation modernization. Piped conveyance was found to raise efficiency to 60 per cent, and modern irrigation networks to 80–90 per cent (drip = 90 per cent, sprinklers = 80 per cent). Both investments taken together were found to result in a combined efficiency of 70–75 per cent. The survey found that participating farmers were able to get more income per drop – participating farmers pumped 40 per cent less water and still increased their incomes by 10 per cent. However, neither GSCP nor SBWMP have demonstrated ‘real water savings’ (see Box 6.15).

Box 6.15: There is a need to verify water savings⁶²

Farmers’ attitudes to water conservation under the SBWMP project have on the whole been positive. However, at the conference held in 2010 to discuss the findings at the end of the project, there was considerable discussion about the extent of water savings and benefits.

A similarly mixed picture emerged at the completion of the GSCP. Despite reported savings, no impact on reducing aquifer depletion was recorded. Satellite imagery suggests an increase in water consumption.

There is a need to gather and analyse data to verify the extent of ‘real water savings’, i.e. what is the reduction in non-beneficial evapotranspiration. Satellite imagery to measure evapotranspiration needs to be employed on a regular basis in order to track water consumption, changes in cropping pattern, etc.

There is eternal doubt about what happens to the ‘saved water’. Evidence is incomplete as to whether farmers simply expand irrigation on qat, or sell the water elsewhere, or whether the water is actually conserved in the ground. There are no monitoring data to show that groundwater levels have revived or fallen less slowly in areas where modern irrigation has been installed. In GSCP, only 4 per cent of farmers thought that the project had a positive impact on groundwater levels.

Overall, there is a need for more in-depth study. Better monitoring and impact surveys are needed to determine what farmers think of the improvements, whether they will adopt them without subsidy, what they do with the extra water, whether they are changing their cropping patterns, whether yields have improved, etc.

Despite this, incentives have remained distorted towards qat, and irrigation improvement has reduced employment

The immediate result of adjusting the diesel price appears to have been a strengthening of the ‘rush to qat’, a decline in other agricultural activity, and a drop in agricultural employment. Farmers using improved irrigation report less demand for labour, as water distribution is done largely mechanically, and irrigation times are far less. GSCP farmers reported 54 per cent decline in labour requirements after irrigation improvement.

Larger farmers have the best access to means of buffering the shock of diesel price rises – for example, by accessing subsidized investments in water productivity improvements. Poorer people – typically rainfed farmers – do not experience the same direct price impacts as irrigated farmers but face increased costs of goods, and a rise in the cost of domestic water. They also face loss of employment opportunities.

A gap is the absence of accessible new technologies to improve productivity and incomes

Improving overall productivity requires adaptive research and dissemination on crops and varieties, on agricultural water management and crop husbandry techniques, and on harvest and post-harvest technology. It also requires market infrastructure – both physical things like roads, and institutional set-ups like efficient input and output marketing – as well as financial services to provide working and investment capital, and modern methods of payment and savings, etc.

Changes in the incentive structure conserve water – but risk negative impacts on production and incomes, unless accompanied by support programmes that can offset negative impacts

The Ministry of Agriculture and Irrigation is implementing programmes to improve water productivity – e.g., GSCP and IIP. These programmes are beginning to work but they are expanding at a very slow pace and the vast majority of Yemen’s farmers, particularly poorer and smaller farmers, do not have access. In addition, these programmes do not always solve the farmer’s problem, as investment in water saving does need to be accompanied by advice on on-farm water management and by agricultural packages and access to market outlets that can significantly increase ‘income per drop’ and so raise incomes.

The challenge is to increase incomes and employment while reducing water use – equitably. Experience shows that this requires a combination of measures – for example, investment in water saving, good advice about on-farm water management, agricultural packages that can raise farmers’ incomes, farmer organization in WUAs or cooperatives, market development, a harmonized approach to managing water resources at the central and governorate level, etc. In addition, there is a need to consider the points about subsidies and promoting market-based approaches that were made in section 6.3.1.

6.5 The political economy of irrigation: getting agreement to move from rapid development to sustainable management

Since the third millennium BC, Yemeni farmers have been developing practices of irrigation and agricultural water management that are well adapted to the climate and topography of the country. Simple water harvesting and flood diversion structures and hand-dug wells proved remarkably durable systems. Overdraft of aquifers was not feasible due to the simple technology used. Irrigated agriculture was largely part of a closed subsistence economy. Institutions for allocating and managing water and for conflict resolution evolved at the local level.

Only some of the larger structures built at times of economic prosperity and strong government proved more vulnerable (see Chapter 5). The fortunes of these more elaborate schemes, like the old Ma’rib dam, were dependent on the continued strength of the state. The Ma’rib scheme, which had lasted more than 1,000 years, collapsed when the Sabaean state became enfeebled – an early example of crisis in public sector operation and maintenance capability.

Technology and the explosive growth of the market economy have led to over-exploitation of groundwater and have undermined traditional irrigation governance

In modern Yemen, two significant changes have profoundly altered traditional patterns of irrigation. First, modernization disturbed the age-old balance and threatened the sustainability of the resource. Macro-and sectoral economic policies adopted by the modern state encouraged the dramatic expansion of well irrigation. At the same time, the demographic explosion and rising expectations drove the supply side (need for higher rural incomes) and the demand side (rapid development of demand for cash crops).

…while government assumed responsibility for larger spate irrigation schemes

Second, government invested heavily in the 1970s and 1980s in spate irrigation, transferring larger schemes from private to public responsibility. These government interventions in formerly private infrastructure created tensions in themselves – for example, in shifting water rights from downstream to upstream, in addition to shifting responsibility for management from users to government.

Political economy factors conditioned the course of the rapid development of irrigation in modern times

The very rapid growth of Yemen’s irrigation sector was produced by the interaction between the objectives of key actors and the factors that enabled or constrained them.⁶³

Since the creation of the modern state, Yemeni governments have pursued three objectives: to acquire legitimacy with citizens and international partners through visible development; to distribute benefits and create prosperity for as many households as possible; and to consolidate power by ensuring that influential groups have preferential access to the wealth and prestige they desire. The discovery that large reserves of natural capital could be developed through modern pumping and irrigation technology gave a unique opportunity to government to meet all three of these objectives

…and the characteristics of the groundwater irrigation sector allowed even a ‘weak’ government to exploit the opportunities in irrigation

Government in Yemen is generally considered ‘weak’, able to affect little except by agreement with powerful constituencies, and with limited implementation capacity. However, the development of irrigation shows how a ‘weak’ government was nonetheless able to achieve its objectives.

In groundwater, the conjunction of private capital and the introduction of tubewell technology was completed by the favourable incentive structure the government was able to put in place

The enabling factors in the rapid spread of groundwater irrigation were availability of capital and introduction of tubewell technology. Through touches on the macroeconomic levers that it did control – trade restrictions, diesel pricing, credit pricing and allocation – the government could subsidize the cost of groundwater irrigation and so promote the rapid exploitation of it by many farmers.

Government and donors also provided capital, and this benefited key interest groups in particular

Figure 6.9 ‘Qat palaces’ amidst qat fields (northern Hamdan district, along the Amran road). Photograph courtesy of Peer Gatter.

Through the same levers – particularly the credit mechanism – the government was able to direct benefits towards key groups important in the consolidation of its power – sheikhs and large landowners (particularly in frontier areas⁶⁴ where even loyalty to the nation – not just the government – might be at stake), and the military and commercial elite looking for profitable ventures in the farm sector (Figure 6.9). Donors supported these developments, helping to establish the agricultural credit bank and making lines of credit available.

Even the absence of institutional structures favoured groundwater development – and this, too, can be seen as part of government ‘policy’

Government also supported the unrestricted expansion of wells simply by its policy and institutional stance.⁶⁵ Groundwater development took place in an institutional vacuum where traditional rules were incapable of moderating the proliferation of wells. By failing to create a legal, regulatory and organizational framework adapted to the governance of this proliferation and of the subsequent extraction of Yemen’s water capital, government made a key, if tacit, policy choice.

Spate modernization required direct government and donor investment – and here, too, the lion’s share of benefits went to important constituencies

In the case of spate development, ‘weak’ government relied heavily on donor capital. The resulting very visible developments gave a good income boost to most users, particularly influential upstream landowners with first rights on flows. The absence of capital and recurrent cost recovery – equivalent to ‘free water’ – limited tensions, creating appreciation by both ordinary farmers and by the elite upstreamers.

This set of policies led to the rapid development of irrigation and boosted incomes, particularly for important interest groups

This strategy for irrigation development was successful, both economically and politically, modernizing agriculture and bringing self-sufficiency in many high-value products like fruit and vegetables. The benefits were spread across a large section of farmers. Important interest groups benefited, and this has helped the government to consolidate its rule. Donors were willing partners, seeing it as visible and productive development.

From the 1980s, the government also pursued a ‘virtual water’ policy that ensured food security while promoting higher value domestic production

As the population grew, Yemen lost its former self-sufficiency in cereals – nowadays Yemen produces only a quarter of its cereal needs; the balance is imported. Fortunately, the government was able to ensure food security through access to cheap, often subsidized, cereals imports. This allowed the government to pursue a water pricing and agricultural strategy that promoted development of higher value-added production rather than supporting production of lower value staples. Government control over imports also simplified the management of consumer subsidies on cereals. Thus the government’s water pricing policy fitted as well within its overall food strategy as it did within its agricultural development strategy and its political arrangements.

The incentive framework did foster domestic production of cash crops, improving farm incomes and stimulating agriculture. However, the restrictions created distortions that had unfavourable economic repercussions. The increased production of fresh fruit and vegetables came at the expense of consumers in the form of higher prices and lower quality of produce. A captive local market has meant little incentive for producers to improve quality or compete in export markets, even where there has been underlying competiveness.⁶⁶ The administration of subsidies through the agricultural credit system prevented the credit bank CACB from becoming a viable financial institution, and stifled the emergence of competitive credit markets. And of course there was an increasingly adverse impact on groundwater depletion.

Whether the government’s implicit strategy was pro-poor is open to debate. Certainly a policy that encourages efficient production of high-value crops and the availability of basic foodstuffs at low prices could be said to be pro-poor. However, several elements raise doubt about whether the outcomes of government policy were actually pro-poor:

• First, inequity in access to water resources and to the income derived from water became marked. The laissez-faire policy towards groundwater allowed the elite to drill deeper and faster whilst maintaining the legal and institutional system to prevent weaker people gaining access to the resource. Draining of water (particularly springwater) through tubewells that had previously been equitably shared led to a shift in ancient water rights towards the better-off.

• Second, the privatization and conversion to irrigated production of run-off land, to which the poor had previously had access, again favoured the more influential.

• Finally, the incentive system tilted the terms of trade away from rainfed cereals production, and this was reinforced by the absence of public programmes to support traditional agriculture on which the poor depended.

During the 1990s, political economy forces were readjusting, with stakeholders re-positioning themselves in the light of changing ideas and perceptions, and of new economic realities

The 1970s and 1980s presented unique opportunities for water resources development. Many farmers prospered, and the government achieved key policy objectives. However, by the 1990s, negative economic and distributional effects of groundwater development emerged, and both farmers and government became aware of the need for water conservation. At the macroeconomic level, a fiscal crisis reduced the scope for public subsidy, and a shift in development strategy led the state to begin moving away from subsidized regimes towards a revised model of development with more emphasis on poverty reduction and on new development approaches like decentralization, partnership, user involvement, cost recovery, local initiative and sustainability.⁶⁷

Other stakeholders also became more likely to be motivated by conservation or by desire for better services

Where farmers previously looked on groundwater as a limitless bounty, it became clear for many during the 1990s that further development of groundwater was a negative sum game. Users began to see the merit of limiting further extraction – provided that existing rights were assured and incomes were protected. A constituency for groundwater recharge emerged, driving the contentious small dams programme. In potable water, consumers started to consider that paying more for better levels of service could be preferable to poor quality subsidized options. The powerful donor community, after two decades of financing expansion, began promoting financial and water resource sustainability.

By the late 1990s, the combined factors appeared to have moved Yemen’s water policy from its ‘unregulated development and expansion phase’ to its ‘management phase’

The natural resource constraint, the crisis in the public sector and the change in the view of the role of the state together moved the focus from a preoccupation with augmenting supply alone to an increased awareness of the need for demand management. The water shortages in agricultural areas, and the dwindling availability of water resources for urban supply, emerged as powerful drivers of acceptance of the need for reform.

Awareness and reform have proved slow maturing, but have been helped along by ‘decisive moments’

This readiness to consider change took many years to emerge, and is still far from complete. Generally, this is consistent with experience in other countries, where from two to three decades have elapsed from the first sign of a problem in water management through an awareness – often jolted by some ‘decisive moment’ – to a final stage of effective and decisive action to deal with the problem. The slow and reluctant development of awareness in Yemen fits this global experience.⁶⁸

Vested interests that benefited from the earlier fast development of water emerged as potential losers from the changes

The groundwater boom resulted in a massive process of ‘resource capture’ by the powerful constituencies on whom government depended, and who consolidated their gains with profits derived through the low diesel price and protected markets. By contrast, poorer farmers and the rural landless benefited less or not at all. Government and politicians were also definitive ‘winners’ in this ‘development and expansion phase’ as the allocation and development of water were sources of power and patronage. The subsequent ‘management phase’ has proved politically much less attractive, as it has involved price rises and restraints on use applied to those who had captured the resource. In addition, where government actually began to implement pro-poor programmes such as the Social Fund for Development (SFD) and Public Works Project, this reduced the scope for patronage. Demand management and pro-poor programmes were not first choices for an unconstrained political establishment in Yemen. However, the constraints outlined above were there, and government did begin to adopt demand management policies and to surrender some of its mechanisms of patronage.

By 2000, Yemeni decision takers recognized the need for demand management – but by then the challenge had become extreme

The changing view within government of its role, together with the realization that water was not a limitless resource and with growing constraints on the public purse, combined to nudge thinking in Yemen towards conservation and demand management. The problem was that water was already over-allocated, and Yemen lacked the governance structures to recover control and rein in the extreme levels of groundwater mining that were by then occurring. And even if there were some wise heads opting for conservation, there were also powerful interests in opposition. The next section examines the forces for and against the policy reforms which came to be proposed in water for agriculture after 2000.⁶⁹

6.5.3 Stakeholders in irrigation: political will, constraints and implementation capacity

By the year 2000, the government began to act on the by now obvious problems of the water sector. A new water law was passed, a dedicated water ministry was set up, and a national water strategy – NWSSIP – was approved, with the target of efficiency, equity and sustainability in resource allocation and use. The current section analyses stakeholder interests – winners and losers – to try to assess what the political economy underpinnings of the reforms were, and to explain why some reforms went ahead and others did not.

This section describes stakeholders with significant influence over the water resources and irrigated agriculture reforms, and assesses their possible support or opposition. Stakeholders include political leaders and parliamentarians, central and local government, traditional leaders, NGOs, the private sector, the media, farmers and domestic water users. As Yemen remains dependent on external support, donors are also stakeholders. Donors are in fact powerful agents for change because of their investment resources and the accompanying ability to influence what the government does.

Parliamentarians represent, to varying degrees, constituencies that uphold traditional values: individual rights to exploit water in line with custom; the rights of the tribe and the community to arrange affairs without state interference; and the expectation that the state will provide financing for capital projects for water resources development (rather than demand management). Despite this background, parliament has consistently supported the rather weak government agencies proposing water reforms, and has passed water reform legislation. The Agriculture Committee of parliament actually travelled to Lahej and Abyan to see the work of NGOs there. They wrote a report, which they summarized succinctly as: ‘Multiply WUAs!’⁷⁰ Thus, despite conservative, populist and potentially rent-seeking tendencies, parliamentarians have been, by and large, a positive force. One important caveat exists, however: parliamentarians in Yemen are largely reactive and not very powerful: parliament will not initiate reform. It seems, however, that it will normally support it.

These individuals have benefited considerably from the opportunities to capture and exploit water resources, and generally oppose reform attempts

Tubewell-farming sheikhs and other large landowners have benefited from resource capture under the status quo, and also from the explosion in the qat market. The integration of this group into the ruling establishment has given them added influence. They oppose most reforms as a means of protecting their vested interests. In particular, they resist surrendering control under decentralized and community management approaches. They resist water user associations unless they control them (see section 6.3.2). They seek to manipulate or avoid regulation; and they struggle to replace rent lost through elimination of the diesel subsidy by greater access to subsidized water saving programmes (e.g., Box 6.17). In most cases, this opposition is in discreet ways – in simple non-compliance, or in cornering shares of publicly subsidized programmes. In some cases, however, opposition has been overt (e.g., Box 6.14), and ‘might is still right’ in many areas. Box 6.16 illustrates a straightforward refusal by the most powerful to respect good practice or fairness.

Box 6.16: Powerful interests in Abs.⁷¹

In the Abs mango growing area, where many top leaders have invested, a proposal by the local development agency to curb groundwater over-extraction was met by a riposte from the highest level that ‘water should be transferred from the next wadi’. In fact, the water in the next wadi is already fully used by small farmers.

The general run of smallholder irrigating farmers has, by contrast, a keen interest in increasing water use efficiency and reducing groundwater overdraft rates

The behaviour of irrigating farmers crucially affects water resources and the implementation of water sector reforms to conserve water. While large farmers are keen to retain the status quo, small farmers, who are equally water users, are aware of the dangers of over-abstraction and are willing to test modern, water-saving irrigation techniques and to develop or join communal institutions to moderate over-abstraction. Chapter 11 provides illustrations of this readiness.

The Ministry of Water and Environment (MWE) has the mandate for water sector reform – but lacks the authority and implementation capacity to push it through

The MWE is the government ministry responsible for designing and implementing water reform. Top management are well versed in the issues and are committed to reform. The MWE receives strong support from the planning ministry, Ministry of Planning and International Cooperation, but is otherwise somewhat marginalized in power circles, and does not have influence enough to sway decisions on major economic issues such as diesel price increases or agricultural trade liberalization. Up to now, the MWE has had erratic, often tense, relations with the Ministry for Agriculture and Irrigation, which represents the interest of 93 per cent of water users, and despite many attempts to forge a common strategy, the MWE has been unable to mobilize more than nominal support from the Ministry of Agriculture and Irrigation for reforms.

Aside from a lack of authority, the biggest problem with MWE is implementation capacity: although the ministry has several top officials of high calibre, they have virtually no staff to support them. As a result, MWE has very limited ability to plan, implement and monitor outcomes in the sector, and its top staff spends an inordinate amount of energy struggling to get other agencies to cooperate on NWSSIP or dealing with donor requirements.

The NWRA is responsible for water law implementation, under MWE supervision. The agency faces enormous challenges with slim resources. Its main problem is lack of implementation capacity. For years the agency struggled to absorb its legacy staff, each with their own institutional culture, and has been dogged by a top-heavy and rather inert headquarters and lack of management vision or capability. The agency’s comprehension of the reform challenge, its commitment to reforms and its implementation capacity are all limited.⁷³

The Ministry of Agriculture and Irrigation (MAI) represents the biggest and most powerful water users but has moved progressively towards a water management and conservation perspective

The Ministry of Agriculture and Irrigation represents the interests of irrigated agriculture and for a long time channelled large subsidies to the expansion of irrigated farming. Until the 1990s, the ministry was responsible for water resources allocation and management as well as development, and still retains staff and the mindset of a resource manager. In recent years, in line with the change in the government approach (see above), the ministry has moved more to promoting resource conservation, water productivity and user associations (see section 6.3 for a description of these programmes).

The Ministry of Agriculture and Irrigation has generally regarded water sector reform negatively

The Ministry of Agriculture and Irrigation participated only marginally when the national water strategy, NWSSIP, was first developed in 2003–4, and it was reluctant to support key provisions such as raising the diesel price or acting on qat. Behaviourally, MAI regarded the strategy as a threat rather than as a credible national programme in which it should participate. This is probably due in part to institutional rivalry – MAI looks on the new water ministry as a menace to its power. This is put nicely as: ‘MWE is a son who has forgotten his family.’ The Deputy Minister of Agriculture also commented that the water strategy was ‘not prepared with enough consultation or involvement of MAI’, etc.⁷⁴ In reality, the Ministry of Agriculture and Irrigation sees the national water strategy as ‘loss’, both at the level of the rural economy – MAI senior officials say that the strategy is ‘all about reducing agricultural water use, but what about farmers’ livelihoods?’ – and at the level of public and donor resources – ‘donors are now leaving us for NWSSIP’. As one senior official commented to our 2007 PSIA study: ‘Agriculture has 93 per cent of the water – but only 8 per cent of the water investment budget.’ The result was a general breakdown in relations: a cooperation agreement between the Ministry of Agriculture and Irrigation and MWE hammered out at staff and deputy minister level through months of negotiation went unsigned for a year and was then shelved.

The Ministry of Agriculture and Irrigation does support the idea of WUAs as farmer groups but sees integrated water resources management as a threat

The Ministry of Agriculture and Irrigation, as a representative both of its own and of farmers’ interests, takes issue with specific elements of the national water strategy. While it is in principle supportive of decentralized management and stakeholder partnership – and is actively promoting WUAs (see section 6.4.3 above) – it is intensely suspicious of the IWRM approach,⁷⁵ which it understands as surrendering control over agricultural water to the new water ministry, MWE. In focus group discussions during our PSIA study, it was clear that the Ministry of Agriculture and Irrigation staff often see NWRA, Basin Committees and other institutions as at once ineffectual and hostile to the interests of farmers. This attitude of the Ministry of Agriculture and Irrigation has been reinforced by the constant hectoring by the water ministry and donors that ‘agriculture has to give up water’, and by the comparative lack of support for programmes that will provide compensating increases in rural incomes.

The Ministry of Agriculture and Irrigation has also seen adjustment of the diesel price as a negative change in the terms of trade for agriculture

The Ministry of Agriculture and Irrigation has also not supported the increase in the diesel price – although it has been powerless to stop it – because it has understood it as a reduction in subsidy to the rural sector without any countervailing means of improving productivity and so restoring rural incomes.⁷⁶

Recent years have, however, seen a recognition that the water resources problem cannot be resolved unless the problem of irrigated agriculture and rural incomes can be solved – and vice versa

When the NWSSIP Update was prepared (2007–8),⁷⁷ the MWE took pains to reassure the Ministry of Agriculture and Irrigation that NWSSIP was not about transferring water from poor farmers to rich towns but rather about increasing rural incomes while improving the sustainability of irrigated agriculture. This new cooperation resulted in a strong irrigation reform strategy and investment programme in the NWSSIP Update, and subsequent institutional restructuring of MAI.

Donor support for – and expectations from – reform in water for agriculture have been high, but tangible results are required if interest is to be sustained

Donors, particularly the core group of Germany, the Netherlands and the World Bank, have strongly supported water sector reform and are financing a number of projects supporting sustainable groundwater management. Donors are tracking progress, and the stakes are high: donor support will diminish if the reform programme falters, or if no tangible results are forthcoming.

6.5.4 Ways to move reforms ahead ‘with the grain’ of the political economy realities

Dealing with the ‘political economy of reform’ requires time, institutions, dialogue, opportunism, incentives, and leadership

Clearly, an understanding of the political economy of reform provides considerable explanatory power about what will and will not work where reform of water for agriculture is concerned. It also indicates ways to increase the chances of reform passing successfully. The following key political economy factors need to be factored into drawing up, deciding on and implementing the reform programme:

• Reform requires support from stakeholders, and support requires both a learning process and time. Education, transparency and patience are indicated.

• Reform needs national institutions with analytic and implementation capacity and the ability to monitor and report on reform implementation and results. Building such institutions is vital.

• The role of catalysts and educators is important. Donors can contribute materially in this role.

• The twin parents of change are necessity and opportunity. Reformers need to wait for the ‘decisive moment’ and then exploit it to the full.

• There is a certain ‘adaptive capacity’ in every community, more or less pronounced and powerful. Understanding the adaptive capacity is key to predicting and promoting change.

• No one will act against their will, and the correct incentive structure is essential.

• Leadership is imperative. No reform programme can work unless there is at least one powerful reform champion.

Clearly, negotiating and implementing a reform strategy for agricultural water in such choppy waters is a challenge. Some practical ways forward are discussed here. The first is to promote an informed national debate, leading to a culture of conservation.⁷⁸ In a pluralist, argumentative society like Yemen’s, the best entry points are knowledge – what is happening and how disaster may result, what lessons international experience may have to offer, and so on – and structured debate, in the press, in academia, in the endless conferences to which the administrative and intellectual classes are partial.⁷⁹

Box 6.17: The rich get richer….⁸⁰

In the downstream area of Wadi Siham, we visit the farm of Sheikh Abdul Karim. He has invited some of the neighbouring farmers, and also workers from his own farm. The farm is about 30 ha, irrigated by wells. The crops include tobacco, sesame, cotton, okra, hot peppers and watermelon. He employs up to 50 people in the season.

He has GSCP conveyance pipes on 10 ha around one well. The well is 100 m deep and was last deepened about 20 years ago. He says he has no problems of quantity or quality. Factors which impelled him to seek GSCP aid are the rising costs of diesel and – he says – of labour. He pays women about Rls 200/day ($1), and men Rls 300 ($1.50). He is very satisfied with the pipes because they cost only one-third of what equivalent pipes retail for locally – Rls 800 ($4) from the project against Rls 2,800 ($14) from the local stockist. He is also happy because he is saving on fuel: where he was buying 400 litres a week he is now buying only 200 litres. He says that water which used to take three hours to reach a field – here he points out a field about 100 m away – is now there immediately.

The service he got from the GSCP was good, he says. He applied, was told to form a water user group, a surveyor visited his farm, he signed the papers and made his deposit, and then he got the pipes. He has no problems – except to request more pipes for the other 20 ha. He mentions this repeatedly throughout the discussion. He says he has had no advice from the project about water management or irrigation scheduling, and no advice about cropping – he learned long ago and does it all in the same old way, he says.

When we ask about his water user group, the sheikh at first does not recall it. But when prompted, he remembers that to get the pipes they formed a group called Majma’ Suleimaniyya. Recollecting, he says vaguely that they will have some training activities soon.

Here the pipes certainly save water, and are reducing consumption. He is pumping with less diesel. The project, however, is quite limited: really just a survey and some pipes. The sheikh has not adopted any different irrigation methods or changed his cropping pattern. It is not clear what has happened to the ‘saved’ water.

There is clearly no poverty reduction impact, as he is the biggest farmer around and comfortably off. He is reducing employment with his water saving, and his labourers are certainly poor. It is not clear whether such a farmer should really be the priority for a public subsidy.

A second approach is to ‘empower the weak’. This is already underway with the promotion of WUAs, but much more can be done to equip these associations with knowledge and resources, to help them to organize, manage and federate, to demonstrate success in water saving and income improvement, and to have a voice. Here, a critical opportunity is the growth of spontaneous collective water management (see Chapter 11). Much can be done to support and develop these movements and to tap into the ‘adaptive capacity’ present in every community.

A third component is to develop and support champions and institutions that can promote a vision of reform and catalyse its realization.