The Resource Allocation Strategy as a Key to Irrigation Project Design

Abstract

The way in which resources (mainly land and water) are allocated within an irrigation project is crucial for determining whether the project's social and production objectives can be reached and also if the project will be sustainable in the long run. This article presents three examples of preparatory phases or feasibility studies for irrigation projects in Latin America and Africa. These examples provide a practical example of how resource allocation considerations are taken into account and what repercussions they have on overall project design.

Background

This article is based on professional field experience in the design of irrigation projects in developing countries, i.e. projects which are usually co-financed by development banks. One decisive phase of irrigation project design is the feasibility study, which has to bring together a wide range of different disciplines in order to assess the economic, financial, social, technical and environmental viability of a project. In a very abbreviated form, it can be said that the main task is to arrange a "happy marriage" between socio-economic and technical design aspects. Ensuring that the system allows for cost-effective as well as secure water provision and that an organisation capable of good management is in place are key ingredients for sustainability. This task, as easy as it may sound in theory, constantly proves to be a difficult one in practice. One reason is that the different disciplines and stakeholders concerned tend to speak different languages shaped by their accustomed ways of thinking and/or specific interests. Although this can be overcome by good leadership and willingness to communicate constantly with each other, there is still a need for tools and instruments that can help to formulate shared visions. This article presents one possible approach to this end.

Case-oriented introduction to resource allocation strategies in irrigation design

Irrigation projects usually pursue a mix of productive and social objectives. Whether these objectives can be met usually depends on how resources are distributed,  or how they can be accessed by different (potential) users.  The main resources to be considered within this context are usually land and water. Sometimes other scarce resources such as credit or production quota (e.g. of sugar) may also be important.

Based on this orientation, the initial studies during the feasibility stage will consider the classical topics (geology, topography, hydrology, soils, agricultural production and markets, local institutions, etc.). Additionally, they should put special emphasis on understanding present resource distribution, its history, resource market mechanisms and market performance, as well as implications of these aspects for future developments in scenarios with-project and without-project.

The development of the resource allocation strategy then follows the steps listed below:

1. Clearly define the objectives of the irrigation project.
2. Clarify the principles to be respected while pursuing these objectives (e.g. to follow a participatory, consensus-oriented approach involving all major stakeholders in the development of the project concept).
3. Lay out the different elements of the resource-centred strategy in terms of initial allocation and mechanisms for future "market"1 operation.
4. Check for consistency between strategy elements and objectives, as well as between the strategy elements themselves.
5. Plan technical layout of irrigation system, irrigation organisation and support structures for project implementation in tune with the resource allocation strategy.

Figure 1 summarises this process in a visualised form. Brief descriptions of practical examples - concentrating on the study phase of three irrigation projects - are used to illustrate the practical application of the approach and to allow the reader to assess its usefulness.

Figure 1: Elements and sequence of the approach

Case I: Bolivia - Sacaba Irrigation Project

Case I involves the improvement and enlargement of a traditional 1,500 ha scheme, with co-financing from the KfW (German Bank for Reconstruction) . The main objective is to reduce seasonal migration to the coca production areas by offering alternatives for making a better living in the Sacaba valley. The main principle applied was to define the project design in a participatory manner, with due respect to technical, financial, legal and environmental limitations. The resource allocation strategy was thus developed over a 12-month period, through an iterative process of discussions with more than 20 communities. The great majority of these are located in the Sacaba valley. In addition to holding repeated discussions, technical (feasibility) studies were carried out on about 10 traditional dams in the mountains. Main elements of the emerging resource allocation strategy were:

• The project concept foresaw allocating the major part of "new" water to the dryer, southern part of the Sacaba valley. Most families in this area were members of the existing irrigation organisation, but held few shares. This concept was accepted by the water user organisation. However, its members also requested including some villages in the mountain area, i.e. near to the water sources, in order to legalise the water theft at present 'traditionally' practised by these communities, and turn them into members with full rights and obligations. After all participants had agreed to these general guidelines, the project elaborated a detailed proposal for the distribution of new shares among communities, based on the number of families, the area and the present water availability. This proposal was immediately accepted, perhaps not so much because farmers understood the underlying, somewhat complicated formula, but rather because the resulting distribution pattern simply seemed "just" to them. A negotiation that ran parallel to the discussion of water distribution dealt with the contributions to project cost required from the users for acquiring the new water rights.  A consensus was reached in this field too - albeit a rather demanding one from the farmers' point of view.
• Since various different water sources were available and were to be developed by the project, the project team introduced the idea of reserving different sources for different groups of communities. One important criterion was the relationship between land characteristics (especially gradient) and water flow. Thus communities in or near to the mountains should be connected to sources with low flow rates (in order to make the water easier for them to manage). Communities further away in the valley should be connected to sources with higher flow rates (in order to reduce losses). This concept was rapidly accepted and actively promoted by the water user organisation.
• Since water rights in the watershed belonged partly to farmers, partly to the city council (for drinking water), an effort was made to strengthen the municipality as a full-fledged stakeholder in one of the water sources, in order to allow for future mobility of resources between irrigation demands and the urban population's (growing) demands for drinking water.
• Finally, the project design included measures for structuring water markets and rendering them more flexible by setting up rules for lending and selling of water rights. It also included measures to create the necessary infrastructure for allowing transfer of water rights from surplus areas (e.g. some communities near the city where agricultural land is lost to urbanisation) to deficit areas in the Sacaba valley.

This project concept has now been successfully implemented.

Case II: Swaziland - Lower Usuthu Smallholder Irrigation Project

This project, co-financed by the European Union, aims to develop approximately 11,000 ha of irrigation area near the Usuthu river. Water from peak summer flows will be stored in a new 160 million cubic meter off-river reservoir and made available for irrigation by a mixture of pumping and gravity irrigation.

In this case, land belongs to the Swazi Nation and is managed on its behalf by traditional chiefs, meaning that there are no formal, individual land titles. The potential irrigation area is a loose, patchwork pattern of plots cultivated by households inside large grazing areas and covers part of six chieftaincies. Apart from some smallholder groups using pumps and sprinklers for sugar cane production, the area has no irrigation. The existing smallholder groups have different types of legal arrangements (cooperative, association, company), according to the decisions made by their members.

The main objective of the project is to alleviate poverty in rural areas. This involves the complicated task of assigning both land and water resources. The main guiding principles for doing so were to provide access to the resources in a just and equitable manner, while avoiding setting up rules that are well-meaning but too complicated to be controlled or enforced from "above".

A second main element was to work with user groups. One reason is that this model is already being successfully applied in smallholder irrigation in Swaziland. Furthermore, based on worldwide experience over many years, forming decentralised, resource-user groups is broadly regarded as a key factor in achieving low transaction costs in natural resource management, specifically in irrigation.

The main strategy elements derived from the above mentioned objectives and principles were:

• All households living in and adjacent to the command area are eligible to become members.
• Irrigation units of between 55 to 110 ha are to be identified in the field based on technical (e.g. shared infrastructure), natural (e.g. soil characteristics) and socio-economic (e.g. clan boundaries) characteristics. These units are to be demarcated in the landscape and will thus quickly become visible for the population.
• People living in and around each unit are invited to join the respective water user group until a target ratio of approximately 3.5 ha per household is reached. This average area per household was determined on the basis of a detailed farming systems analysis. Each household is permitted to be a member of one group only. The process of identifying irrigation units and regrouping households is to be managed in close coordination between project team and local (traditional) authorities.
• The newly formed groups, and not the individual water users, will later be the "clients" and "shareholders" of a lean, central scheme management. Land and water rights are to be assigned to the group which will manage them on behalf of its members (while the individual members' rights and obligations are clearly defined in the respective group's statutes and bye-laws).
• All further decisions concerning internal organisation and resource allocation are made by the groups themselves, with technical assistance from the project preparation and implementation team -  as long as they are in line with certain quality standards, e.g. concerning legal requirements to be met or aspects that need to be covered by the statutes and bye-laws.
• Results of the different processes going on at group level are to be made public throughout the scheme.
• Water and land transfers inside a water user group are to be  managed later by the group itself. Groups should also be given the chance to save water in the dam for the following year. However, savings will be lost if the dam overflows. The overall scheme management has to be involved for water transfers between groups.

The proposed strategy has one further interesting feature: although it aims at fully involving traditional authorities in the creation of the irrigation organisation, their role will probably be weakened in the end, since the new irrigation organisation will be of paramount importance for social networking and income generation.

Because of the relatively high investment needed for implementing the project, the first hurdle that needs to be overcome is to assure the financing. This process has now been largely concluded, and the resource allocation strategy described above has already played one important role - as a project feature that contributed to convincing potential donors.

Case III: Kenya - Kimira-Oluch Smallholder Irrigation Development Study

This is the the most recent and also the most advanced example of a resource allocation approach in irrigation systems design. As the study (financed by the African Development Bank) has not yet been  completed, and discussions on the best approach are on-going, it serves here as an example to illustrate the basic, underlying theory for the development of a resource allocation strategy.

The basic project idea is to provide gravity irrigation to approximately 1,500 ha in two neighbouring areas (Kimira and Oluch) by diverting water from two rivers that flow into Lake Victoria.

Again, the main goal of the project is poverty alleviation. The study team has added the more specific objectives of improving and securing the local population's access to key resources.

It had originally been intended to  tackle the resource allocation issue by dealing with land tenure.  This would involve carrying out "participatory" land use planning and - where necessary - land redistribution, in order to reach an equitable situation and assure efficient resource use by matching labour, land and water availability at household level. However, the study team soon realised that the population did not want to discuss land issues with the project. Land rights in this region have been made official following a lengthy process of land registration and preparation of land titles (still being completed), and people have a very strong sense of ownership. Rudimentary land markets are in place (leasing and selling). Plots are being fragmented due to heritage laws. This occurs in particular along rivers, the one area where some limited irrigation is already possible. Many families have incurred debts, often  due to HIV/AIDS, and land is used as a security for obtaining informal credits.Thus, alongside the fragmentation of plots, cases can be observed where richer households start to accumulate land because their debtors are unable to pay back in cash.

Taking this into consideration, the team posed the following questions: even if the project were to make an effort in the face of strong resistance to enforce farm or plot sizes ideal for irrigation, the same driving forces would return as soon as the project were handed over. This would lead to inequality in land ownership similar to that now observed, as well as the danger of inefficient water use. Would it not be a far more intelligent approach to use the dynamics observed in order to achieve near-optimal resource allocation, while trying to strengthen the position of households with only a small amount of land? In other words: the team's objective is not to "struggle against the tide", but to "ride the waves". Three potential answers started to emerge: do not touch the land (with exceptions aiming at voluntary agreements in areas where land reorganisation or cooperation between neighbouring plot owners is absolutely essential for making irrigation possible); assign water quotas on a fixed unit basis per farming household (i.e. a quantity which is sufficient for irrigating 0.5 ha); and make sure that water remains scarcer than land. Only then can the negotiation power of households without much land (which thus acquire a  relatively large amount of water) be empowered to develop shared land- and water-use models. Furthermore, water scarcity provides a strong incentive for efficient use. However, relative water scarcity means that the command area has to be bigger than the technical optimum would dictate, which will tend to increase cost for infrastructure.

Other complementary options for keeping water scarce could be market based. This could be done, for example, by distributing only 80% of available water in terms of water shares to the individual households while keeping the remaining 20% for annual or bi-annual auctioning. However, this approach would have repercussions on the technical design of the system and would tend to increase cost, since flow flexibility needs to be assured by increasing canal size.

Another consequence of the strategy is the need for a support structure which accompanies and supervises the processes launched, until the "markets" are fully operational, meaning amongst other things, that they are transparent for all participants.

This is of course an extremely simplified presentation of the discussions that took place during the elaboration of the feasibility study and are still on-going. Many other factors which concern future water users (e.g. soil qualities in the project area) have to be taken into consideration.   Further points of discussion are  the question of eligibility for becoming a water user, the role of decentralised water-user groups within irrigation management, or the need for accompanying measures for supporting those households which have run into debt.
One key instrument developed for structuring the complex discussions is the so-called Consistency Matrix, which is presented schematically in Table 1. The idea is to relate the individual elements of the resource allocation strategy to the project objectives and assess how they contribute (or fail to contribute) to their achievement. Furthermore, the inter-relations among the different strategy elements themselves are analysed, in order to detect and eliminate contradictions (i.e. cases where one strategy element poses a hindrance for the achievement of another.)

Conclusion

Developing a resource allocation strategy as a basic orientation for irrigation systems design, either for rehabilitation/improvement of existing schemes or the creation of new ones, helps to focus the task at hand. Based on technical and socio-economic studies and oriented towards meeting the objectives of the project while respecting basic principles of how this is to be done, the resource allocation strategy translates into technical and organisational designs as well as into a task description for the support structures needed for implementing the project and operating the scheme. 

However, the approach does not constitute a standard recipe that can be applied automatically. It rather delineates an arena for discussion, proposing questions to be asked and interrelationships to be analysed. In this way, it aims to achieve a consensus of opinion between the different disciplines and stakeholders involved.

Table 1: Schematic example of consistency matrix for checking proposed resource allocation strategy