Management and maintenance - perennial problems

Introduction

It is generally agreed that the performance of most large-scale irrigation schemes in the Third World has been very poor. Dr. Anthony Bottrall of the Overseas Development Institute, for instance, notes:

"Planners' targets are rarely met: the overall productivity of water is much lower than might be expected and, especially on large surface-water delivery systems, the pattern of its distribution is often extremely inequitable, with farmers in the head-reaches receiving far more than those at the tail, whose supplies (if they get any at all) tend to be sparse and unreliable." [1]

So too, Bruce Stokes, then at the Worldwatch Institute, warns, "unless farm level management improves (in irrigation schemes) efforts to rapidly increase food production by expanding irrigation will founder". [2] Indeed, in India, vast areas of irrigated land lie idle for want of the appropriate management and maintenance.

Asit Biswas also notes the poor performance of irrigation schemes. He quotes a 1981 World Bank Study of 30 irrigated projects in 15 countries. The study concluded:

"Overall, water management ... was found to have received inadequate attention. Insufficient provisions for the systems' operations and maintenance were made at appraisal, and insufficient action was taken during implementation. Analysis of water management issues in completion and audit reports - as well as in the appraisal reports - tended to be incomplete or superficial; quantitative data was sparse and fragmentary ... water supply proved inadequate in 10 cases, unreliable in 6 and inequitable in 3; water losses proved excessive in 5." [3]

Significantly, Biswas goes on to comment,

"from my personal experience in the Africa, Asia and Latin America, the criticisms Bank study are unfortunately commonplace."

Rightly or wrongly, he attributes much of the problem of salinisation to poor management.

Likewise, James E. Nickum, Professor of Asian Studies and Agricultural Economics at Cornell University, attributes the spread of the secondary salinisation in China to poor management. He points out that the National Press is constantly alluding to the problem: in particular, "emphasising construction to the neglect of management" and "only grasping construction without regard for effectiveness" are two failings which receive regular criticism. [4] Indeed, as early as 1963, an editorial in the People's Daily estimated that some 6.7 million hectares of irrigated land were going uncultivated due to poor management.

More recently, the management of the giant Bishihang Irrigation Department in Northern China has been taken to task for its failure to see that the scheme is properly run. Thus a report on the performance of the Irrigation District noted,

"there is no unified, authoritative, sound management for the irrigation district as a whole, making it impossible to do a good job of managing this kind of large-scale irrigation district."

The authors of the report went on to list sane of the more obvious management failures:

"Technical personnel are lacking and some canal structures and projects have no-one to manage them once they have been brought into service. The people who do manage some of the projects actually damage them because they do not understand their technology." [5]

In a similar vein, agronomists in Henan Province complained that that

"the necessary ancillary projects had not been added to 10 large reservoirs built in the 1950s, or to the recently installed 110,000 tube-wells, leaving the province's irrigated acreage 2.7 million hectares below its potential." [6]

More generally, Nickum learned from colleagues at a recent symposium on Long-Distance Water Transfer Schemes that poor management in China had led to the following problems:

1. Farmers in the upper reaches use too much water while those in the lower reaches do not receive enough.
2. Though good results are achieved in experimental areas they are nothing like as good elsewhere. The peasants do not adopt the methods proposed. They regard drainage as too expensive.
3. Poor management has led to salinisation in some parts of the North China Plain - principally through excessive seepage from irrigation canals. That rate is indeed high - reaching 45 percent in the People's Victory Canal system, for instance, and over 50 percent in the channels of the Shijin Canal system.
4. Too little water is used to irrigate too much land. This means that water in the main canals remains there for 300 days a year keeping the water table high. [7]

That those problems should have been encountered in China is particularly significant - not least because the Chinese agricultural system has frequently been held up as a model for the rest of the world to imitate.

The problem of inefficient management is now so widespread in water development schemes, that Dr. Robert Chambers of Sussex University has even talked of the need to "manage those who manage the water". [8] A recent workshop on irrigation management, run by the Overseas Development Institute, also noted the poor management of irrigation schemes. In particular, it was argued that such schemes suffer from

"inefficient finance and staffing for operation and maintenance; minimal training of irrigation staff in operational procedures; grossly inadequate provision of agricultural extension staff; and slow up-take of water by farmers leading to much wastage." [9]

Participants in the workshop also reported a high incidence of water disputes in large-scale irrigation schemes. In the Chimborazo irrigation district of Ecuador, where such disputes are common, it was reported that water use efficiency was as low as 27 percent. Similar problems were also evident in the area served by the Ganges-Kobadak project in Bangladesh.

Management and maintenance: a low priority concern

One of the most obvious reasons for the poor performance of water projects is that little attention is ever paid to the problems of management and maintenance by those who design and construct such schemes. Thus Bottrall notes,

"The planning and design process tends to be dominated by technical experts (especially engineers), with economists being called in to calculate the costs and benefits of what is technically (but not necessarily socially or administratively) feasible." [10]

Moreover,

"In their justification of a project, planners will assume certain future cropping patterns; but (with a population of numerous independent farm operators) what administrative mechanisms will be used to induce widespread adoption of these cropping patterns? What procedures will be used for allocating water? What specialised staff-training programmes will be required to apply these procedures and provide farmers with advice on field-level water management? Few planning reports supply more than vague or routine answers to these questions." [10]

This does not seem to be unusual. By way of illustration, Chambers cites a multi-volume planning report for a major irrigation project in Sri Lanka which contains no more than one and a half pages on organisation and management. [11]

Having spent millions of pounds on constructing an irrigation scheme, surely it makes sense to ensure that it is properly operated and maintained? Why then do the promoters of large-scale irrigation schemes - and the governments which back them - consistently ignore the problem of management?

Sadly, we are drawn to the conclusion that the promoters of such schemes are only interested in the political and economic capital to be derived from actually building a dam or irrigation project. Once it has been set up, there is little kudos to be gained from seeing it is properly run and the temptation is thus to view management and maintenance as the responsibility of the next government to come into power.

Bureaucratic ignorance

Another reason for poor management and maintenance is the basic ignorance of the bureaucrats who are called upon to assure the smooth running of irrigation schemes. Carl Widstrand, for instance, points out how the assumptions underlying the administration of large-scale irrigation works "are never based on sound knowledge". [12] He provides an interesting illustration from the Gezira scheme's Handbook of New Personnel, published in 1951 and which, surprisingly enough, was still being used in 1980. The Handbook recommends:

"Before any irrigation is embarked upon, a number of particulars must be ascertained beforehand. Having decided it is possible to grow certain types of crops, the volume of water required to bring these crops to maturity must be determined with due regard to all circumstances; this is known as the water duty ... It may be that the precise water duty in the Gezira is not yet unanimously agreed upon, but an overall figure of 30 cubic metres per feddan per day seems fairly safe for design purposes (of canals).

"Broadly speaking the agriculturalists have found that a volume of 400 cubic metres of water is required in the Gezira on each feddan of cotton every month during the dry season. Also that, on average, each feddan of cotton should be watered every 15 days. Though this applies primarily to cotton, it suffices for other crops such as dura and lubia."

That quote was first singled out for comment by Tony Barnett in an article on the Gezira scheme in 1977. The operative phrases, says Barnett, are "for design purposes" and "broadly speaking". [13] Widstrand goes even further:

"No-one obviously knows anything about the known behaviour of cotton and, if they do, it is not put at the service of the cultivators. Irrigation technicalities, rather than agricultural realities, seem to determine decisions about watering." [14]

The contrast between modern methods of 'management' and the running of a vernacular irrigation system could not be sharper. Indeed, Widstrand admits, "the peasant has very much more knowledge of local conditions than the local administration". For instance, the peasant often "knows that the new methods will not work" and often refuses to accept "recommendations based on the latest scientific knowledge". This

"creates an instant conflict between the cultivator, who knows his environment and who knows how to manipulate it, and the government extension, who does not understand that the peasant lives by his wits and not by his hands alone."

The peasant realises, in particular, that the rules set up by the administration are crude and inflexible - indeed, they have to be, in order to be implemented by distant bureaucrats with no knowledge of local conditions. Again, the Gezira scheme provides a case in point. Thus, water is allocated according to what Widstrand calls "the average principle". The result, he says, is

"that 'the average' farmer gets an 'average' amount of water for an 'average' crop over the year. Everybody gets water over the year but not necessarily at the precise or necessary moments. This concept is closely related to the idea of 'normal rainfall' and other peculiarities in the 'Folklore of the normal' that simplifies administrative thought."

By contrast, in those traditional societies where irrigation is practised, water tends to be distributed to farmers in accordance with the requirements of the land they farm. The system is flexible enough to make sure that their actual crop, not just their average crop over the years, is watered. The irrigation cycle is also carefully synchronised at a village level in such a way that the water is provided exactly when and where it is required, thereby enabling the system to deal with periods of abnormal rainfall. Indeed, a traditional system is, above, all 'subtle'. Small wonder, perhaps, for it is run by people who have been imbued since childhood with the traditional knowledge required for its operation and maintenance.

Lack of accountability

A further failing in the management of modern irrigation schemes is the lack of accountability. If anything goes wrong, no one in the administration is held responsible. Excuses are always found. Bruce Stokes of the Worldwatch Institute considers that this failing is particularly apparent where irrigation systems are state-owned:

"With no individual accountability, those at the head of the canal often withdraw more than their share of water, adversely affecting all the farmers in the system." [15]

That lack of accountability also favours corruption which unfortunately seems to be a feature of bureaucratically run irrigation systems throughout the whole World. The most obvious form of corruption is when farmers who are members of the ruling political party are favoured to the detriment of those who are not. This, we have been told locally, is very much the case where new land has been brought under irrigation in Sri Lanka.

Corruption is often very much more blatant. As Bottrall points out:

"Even in the best managed irrigation projects, operating staff are likely to be subjected to strong pressures from farmers to supply them with more water than they are entitled to; and these farmers may often have powerful local political backing. If the operating agency is to meet its objectives, staff must therefore be able and willing to impose discipline on competing pressure groups by applying rules impartially and penalising those who break them.

In fact, staff often give way to the pressures placed on them, in the absence of strong commitment at high levels to meet the agency's (stated) objectives, and the result in the worst cases can be a degeneration into 'water anarchy', in which disproportionately large amounts of water go to farmers who are more influential and/or favourably located near the head of the system. The staff's difficulties are compounded when controlled operation of the water delivery system is hampered by design deficiencies." [16]

Widstrand also notes:

"Irrigation staff are ... subject to strong pressures, threats and inducements. This leads to what we, in our Calvinistic approach to life, call corrupt practices."

Indeed, corruption is not only common, it is sometimes so necessary for survival that it has become an accepted part of life. Thus:

"The gate keeper may be persuaded to open the gate to let more water through; the gauge reader may under-report the amount of water taken; the overseer may install large watercourse outlets. The assessors (patwaris) in the Pakistani system assess each farmer twice a year for each crop. They are supposed to reduce the possibility of corruption, but as the assessment procedures and the water rates to be paid are so mysterious, the person who cannot fill in the right forms or does not know his rights, is very much in the hands of the patwari." [17]

The influence of large landowners over tenants who require water can be considerable. According to Ashfag Mirza, for instance, farmers in the Punjab are so used to being victimised that they consider 'fairness' in dealings with an official to be a favour in itself. [22]

The problems of maintenance

If the management of water development schemes in the Third World is nearly always deficient, so too is their maintenance. There are obvious reasons for this: one is the difficulty in finding the appropriate technicians, another that of obtaining spare parts. Those difficulties are well illustrated by Tanzania 's experience with simple, small-scale water supply schemes. [19] Indeed, as Widstrand reports, some of those schemes actually came to a complete halt "because the technical advice available could not serve every scheme at the time the service was needed". Thus

"there was only one technician whose visits were constrained by lack of transport to reach the schemes and to deliver the materials, and by the necessity to spend two or three days with one scheme before entrusting the work to the local mechanics. Also, materials could not reach the scheme in time, because they had to be purchased in Tanga or Dar es Salaam." [20]

It has also been pointed out by Drs. Heijnen and Conyers that pumps are often out of order and that repairs take an inordinately long time because of the shortage of qualified technicians and of finance. [21] To overcome that problem, Widstrand suggests, the organisation of maintenance work should be decentralised. But it is difficult to see how decentralisation will solve the shortage of technicians or, for that matter, how it will help overcome the problem of finance which is less likely to be available at the local level than in the large cities.

Dr. N. A. De Ridder also suggests that minor repairs and maintenance work should be left to the villagers. Although he accepts that villagers will have other preoccupations, De Ridder nonetheless insists that they must be taught how to use their wells.

"Unless this is done and some form of maintenance is organised, there is little hope for a long lifetime for the wells provided by the project." [22]

Outright vandalism has also proved an intractable problem in many irrigation schemes, further undermining the level of maintenance. In East Africa, for example, taps, valves and pipes are often stolen during construction to be used elsewhere or sold in the market. In India too, reports Salisbury, the brass fittings are often stolen from pumps since they are useful for making pots and pans. [23] Such house-hold items obviously play a very much more important role in the lives of the local peasantry than the largely irrelevant machines from which the brass is removed.

In the Indian example, the authorities responded by making pumps and other equipment out of cast iron - a material with which the local blacksmiths were unacquainted. The result was that 80 percent of the 50,000 cast-iron wells in drought-prone, hard-rock areas broke down and ceased to produce water. [24]

Widstrand describes a common situation in an East African village:

"It is a new and hot, shimmering day. Nothing much happens in the village: there is a smell of stale beer and dust. Suddenly on the road large dust clouds: a procession of Landrovers approaches the village. Out of the vehicles spill a crowd of county councillors, district officers, local politicians, and sweaty, but competent, experts.

Someone gives a speech, telling the villagers that now development has at last caught up with them and they shall have water, the toil of women shall finish and health, economy and what-not, shall sprout. Competent technicians unpack the machines and a day of infernal noise begins. The surprised laity invite their guests to a meal of beer and goat meat, more speeches, departure of the caravan, but for some experts, who in a few days' time produce water out of a pump. They depart: 'Here is your water, look after the installation and phone us if something goes wrong.'

Phone! After some time the diesel runs out, the pump disintegrates, and everything is as before. After all, no one really liked that clear cold water which hurt their teeth. Water should be brown, muddy, tasty and more filling. And, by the way, during the rainy season it is much closer to go to the river than to the pump." [25]

Co-operation and maintenance

The above examples illustrate, above all, the basic failure of the authorities to obtain the co-operation of the peasants who must operate water development schemes. This could not contrast more starkly (as we shall see) with the way tribesmen or peasants closely co-operate with each other in the operation and maintenance of traditional irrigation systems.

But why should tribesmen and peasants co-operate with the authorities? In effect, they are being asked to behave in a way which is totally contrary to everything they have learnt since childhood. Indeed, it is only through systematically destroying the cultural pattern of peasants and tribal farmers that the latter can be introduced - with any conceivable measure of success - into the market system.

We have already seen, in our chapter on resettlement, how the Sri Lankan government is using the Mahaweli scheme to transform its peasantry from subsistence farmers into "modern agriculturalists". Another example of the same trend is provided by the government of Zaire's plan to create a "paysannat indigène". In that instance, the government was quite explicit about its determination to destroy traditional cultural patterns. Thus, as George Kay points out:

"It was decided ... that, in addition to normal extension work, an 'impact scheme' should be introduced whereby progressive individuals could be relieved of the economic and social difficulties of village life and equipped with the land, capital and technical assistance necessary to enable them to became advanced commercial farmers in a very short time. The object of the scheme was to facilitate the development of commercial farming and rural prosperity by creating a nucleus of proficient peasant farmers who would, by their example, inspire others to improve their farming and seek their living from the land." [26]

That programme - like many similar ones - failed, the failure being blamed "on the stupidity of the farmers who did not want to become progressive". [27] The peasants were also seen as incompetent, lazy and incapable of following government instructions.

In fact, the whole scheme might be said to have been doomed from the start. What the Zairian government referred to so disparagingly as "the economic and social difficulties of village life" were, on closer analysis, the very social and cultural customs which for years had governed the behaviour of the local tribesmen. Those customs were essential for maintaining the integrity and stability of both the tribesmen's society and their environment.

To ask those tribesmen to abandon their customs was thus to ask them to abandon the basic a prioris of their culture - to violate, in fact, every law previously enforced by their elders and sanctioned by their customs. Small wonder, then, that the tribesmen of the area refused to co-operate in the scheme. To have done so would effectively have meant accepting the disintegration of their culture; breakdown of their families and their communities; and their own reduction to the status of despised and badly paid members of a depressed rural proletariat.

Viewed in those terms, their obstinacy was more than justified: it was essential to the survival of their way of life. Indeed, it is a sad reflection on the tenor of modern economic thinking that the majority of development theorists still look upon the cultural constraints of traditional societies as nothing more than intolerable restrictions on individual freedom and, in particular, on the free inter-change of goods and services.

The conflict between food and cash-crops

Quite apart from being asked to accept the total destruction of their culture, those peasants and tribesmen uprooted by water development schemes are also being asked to jeopardise their very security. As Widstrand points out, "government is interested in marketable surplus". Farmers, on the other hand, "may be more concerned with survival and thus prefer to grow food crops or pay more attention to other crops than the cash crops of the scheme or programme". [28]

Egypt provides a case in point. Thus, notes Barnett, a conflict has arisen "between the government's cotton and the farmer's bread". [29] If the peasant co-operates with the government and produces cotton, he will have to sell to the government at a price that is only a fraction of the world market price. How, then, can he acquire food for himself and his family - food which he is no longer allowed to produce himself? Moreover, to produce cash crops, he needs both to buy fertiliser and pesticides and to pay for irrigation water - all of which are often beyond his means.

The history of development in the Third World is replete with similar examples of the problems which are caused by forcing tribesmen and peasant farmers to enter the cash economy. Thus, John Gretton of the London-based Earthscan organisation, describes a large-scale agricultural scheme in Senegal in which the peasants were asked to grow rice under irrigation, rather than the more traditional millet. The scheme failed dismally. Only a third of the rice fields were harvested, reports Gretton:

"Though rice is more productive than millet, the peasants were unused to cultivating rice and resented the additional work involved. In addition, there were the problems of switching from a subsistence to a cash-crop economy. People found themselves unable to pay even the cost of fuel for the irrigation pumps. They had to borrow money, which could mean an ever-growing debt." [30]

The same sad tale is being repeated throughout the Third World today; what is more, it has happened wherever the market economy has been imposed on subsistence farmers.

The advantages of traditional irrigation systems

Anthony Bottrall, who perhaps more than anyone else, has studied the deficiencies of management in water development projects, suggests that the only solution to the problem is to use traditional irrigation systems as a model for new schemes:

"In contrast to the large publicly operated systems, the communal systems are generally well organised: man-management is not one of their problems." [31]

He refers to the work of E. Walter Coward who points to three important features common to most indigenous communal irrigation systems:

"(i) They have an 'accountable leadership': leaders of each local group are selected by members of that group, their performance is periodically reviewed by them and they are compensated for their services directly by the group.

(ii) The systems, although themselves small, are usually sub-divided into smaller sub-units, each with its own leader: 'management intensity' is very high. Moreover, each of these 'mini-units' corresponds to a discrete physical sub-section of the larger system.

(iii) The Systems are rarely coterminous with village boundaries: they are channel-based, not village-based. Attempts by outside agencies to organise local irrigation groups on a village basis may, therefore, often run into difficulties." [32]

Coward considers that indigenous irrigation systems can be improved in many ways - although he admits that

"it has been a common experience ... that governments' attempts to incorporate communal systems into large projects tend to lead to the disintegration of previously dynamic water users' groups, because they lose most of their original responsibilities in the process and are given no new ones to take their place." [33]

However, there are examples in recent times of water-users' associations being set up very successfully - in Taiwan, for instance - although, unfortunately, efforts to introduce them in other parts of South-East Asia have failed. [34] It is suggested by Nickum that the Chinese have achieved certain successes in this field, but the most recent book on large-scale water development in China does not seem to bear this out. [35]

One of the problems is the difficulty of creating an efficient association of water users out of nothing. Widstrand admits that "groups cannot be conveniently structured or produced, and blanket government decrees to start such groups do not work". [36] He cites as an example "the death rate" of government introduced co-operatives in East Africa. Nevertheless, it appears that such schemes are still being tried out in Bangladesh, India, Indonesia and Malaysia.

Here we come to the crux of the matter. As Carruthers writes,

"To be successful, planned irrigation technology transfer must include not only the hardware but also the software of services, institutions and attitudes." [37]

Those attitudes must include the motivation, the ability to co-operate and the know-how to manage and maintain an irrigation system - and these qualities have, historically, only been provided by the cultural pattern of a traditional society. Indeed, where no traditional cultural mechanism exists, either to assure co-operation among farmers or to provide them with the necessary know-how to practice irrigation agriculture, it becomes very difficult to ensure the smooth-running of an irrigation system - let alone its proper maintenance.

René Millon points out, for instance, that once the Hacienda de Arroyo de Enmedio near Guadalajara, Mexico, was divided up in 1927 amongst farmers from a number of different communities, the farmers "were unable to agree to act in concert to keep these lands under irrigation". As a result,

"the irrigation system was allowed to fall into disrepair. The villagers involved were unable to cope with the problems posed by the necessity for inter-village co-operation." [38]

Another example of the inability of peasants with no cultural tradition of irrigation to manage an irrigation system is provided by Wolf Roder. Thus, irrigation technology was first introduced to the Shona of Eastern Rhodesia by European settlers at the turn of the century. However,

"the absence of a tradition of customary concepts of water allocation and of community co-operation in maintaining canals, hampered the development of this form of agriculture. As is so often the case, diffusion of technological innovation outran the ability of the social traditions to adjust." [39]

Many similar examples can be cited.

The question, then, is how to learn from the experience of traditional irrigation societies? We shall look at five such societies. As we shall see, all of them have proved themselves capable not only of managing and maintaining their irrigation works but, most important of all, of doing so without precipitating the ecological and social destruction caused by modern water development schemes.

References