Professional development workshop on dryland management

-M WORLD B-N - MECOR POLICY AND RE~CH STAFF Environment Departmeht Professional Development Workshop on Dryland Management Jon Martin Trolldalen June 1990 Envi~oment Working Paper No. 33 Thisiser has bee prepared for fnt~mal use.'Tbe views and nsherin a th o of, as oun a o he d d Bank, to is affiatd ogenådm or to This report summarizes the proceedings of a Dryland Management Workshop held in 1989 as an integralpart of the Bank'swork proa on desertification and drlndmanagement. The workshop was organized primarily by Ridley Nelson, and the contributions were compiled by Jon Martin Trolldalen. The Government of Norway provided major support to the program. This paper reports on findings that are particularly significant inthe context of the overall work program of the Environmental PoHcy and Research Division. Departmental Working Papers are not formal publications of the World Bank. They present preliminary and unpolished results of country analysis or research which is circulated to encourage discussion and comment. Citation and use of such papers should recognize their provisional character. The findings, interpretations, and conclusions expressed in this paper are those of the authors and should not be attributed in any manner to the World Bank, to its affihated organizations, or to members of its Board of Executive Directors or the countries they represent. Because of the paper's informality, and in order to release the research without dela, the tpescript has not been prepared in accordance with procedures appropriate to official texts. The Wrld Bank accepts no responisibilfity for errors. CONTENTS Page Preface Kenneth Piddington, Director, Environment Department, World Bank.............................. v PART I - MAIN ISSUES Jon Martin Trolldalen, Environment Department, World Bank.................................. I PART I - INTRODUCTION 2.1 The Extent of the 0Desertification" Problem and How to Measure It Prof. Stein Bie, University of Tromsoe, NORAGRIC, Norway.................................. 7 2.2 How Farmers and Pastoralists Adjust to Land Degradation. How Can beors Help? Prof. Michael Mortimore, University of Birmingham, U.K....................................... PART II - REVIEW OF EXPERIENCE 3.1 What Has Been the Experience with Dryland interventions? Juan Seve and Bruce Ross-Sheriff, Consultants E/DI, International Resources Group, USA....................................... ............... I5 PART IV - RANGELANDS ISSUES 4.1 Experiences with Pastoral Associations In West Africa -- What Should Be Our Future Strategy? Poul Sihim, AFTAG, World Bank.................................. ....................... 23 PART V - WATER AND SOIL FERTILITY ISSUES 5.1 Strategl las fr Initiating Sustainable Moisture Conservation Programs In Semf4ifld West Africa Chris R44J, Free University, Amsterdam, Netherlands ...............................................33 5.3 Dryland Management: Water and Soil Fertility Issues Raymond Meyer, USAI, USA ...................................... 39 PART VI - POLICY ISSUES FOR IMPROVING LAND MANAGEMENT 6.1 Why and How Do Land Users Make Land Management Decisions? How Can Policies Encourage Land Users to Make Thems Effectively? Prof. Piers Blaikie, University of East Anglia, UIK................................................ 45 PART VII - OFF-FARM EARNINGS AND MIGRATION ISSUES 7.1 Income Diversification of Rural Households in Burklna Fase Thonas Reardon and Chris Delgado, IFPRI, USA......................................................49 7.2 Temporary and Permanent Migration In West Africa - Some Early Findings from the Onchocercasis - Freed Areas Study: What Support Should the Bank Give to Migration? David Brokensha, Institute for Development Anthropology, USA..............................55 PART VIII - PARTICIPATORY ISSUES 8.1 The Aga Khan Foundation's Approach to Participation and Its Relevance to Bank Projects Andrew Weir, Aga Khan Foundation, Switzerland...................................................... 57 8.2 Issues of Local Participation In Bank Projects David Richards, World Resource Institute........................................................6........1 PART IX - WRAP-UP 9.1 Wrap-up Professor Stein Bie, University of Tromsoe, -NORAGRIC,Norway ........................... ..... 65 Annexes Annex 1: Workshop Agenda (including the Working Groups) Annex 2: List of Participants PREFACE by Kenneth Piddington, Director Environment Department, World Bank On May 10-12, 1989, specialists from around the world gathered with Bank staff at the Watergate Hotel in Washington to discuss desertification and the Bank's role in dryland management. Two years earlier Mr. Barber Conable, President of the World Bank, had identified desertification as one of the Bank's major environmental concerns. Many managers and decision makers believed that desertification was a problem of "marching sands," and the solution was simply to spend money to stop it. Although that view was held mainly by non-specialists, it remained an important influence in discussions about global cooperation. The Department's Drylands Task Force has tried to redirect thinking towards the more widespread and, in many respects, more difficult problem of the subtle land degradation in drylands not located at the desert fringe. Although moving sands rarely overtake these areas, degradation generally has greater consequences in human and economic terms. The workshop focussed predominantly on this question, as well as the broad range of related socio-economic issues. The presence of outside speakers from the scientific and NGO communities reflected the Bank's desire to work with and learn from these groups. Participation from the UN Sahelian Office (UNSO), Resources for the Future, American Association for the Advancement of Science, International Resource Group, Institutional Food Policy Research Institute (IFPRI), and others contributed greatly to the success of the workshop. The Bank was especially interested in translating information into practical operational guidelines. This topic was addressed by the International Resource Group's report, Experience on Dryland Intervention. The discussion revealed the complexity and the low replicability of the different interventions in various agro-ecological zones. I conclude that research for successful intervention in drylands should be a continuing priority for the World Bank. Other issues of discussion included* Experience with Pastoral Associations; Promising Technologies -- or Research Directions for Increasing Productivity and Improving Sustainability in Rangeland; and Strategies for Initiating Sustainable Moisture Conservation Programs. This workshop was a step towards making Bank staff more sensitive about environmental aspects in dryland management. Working with scientists and other experts, the Bank will use similar initiatives to continue the learning process. I want to thank the Chairman, Mr. John Hayward, for his excellent guidance through the many areas, and Ridley Nelson, Jon Martin Trolldalen, and Surinder Deol for arranging the workshop. I also want to thank Lawrence Mastri for his help in editing and preparing this document. V PART I Main Issues by Jon Martin Trolldalen, Environment Department Part I - Main Issues This section outlines the issues and ideas in the papers and subsequent discussions.' Part II - Introduction The ODesertification" Problem and How to Measure it (2.1) Professor Stein Bie opened the workshop with an overview of the international perception of dryland degradation. Recent measurements (satellite imagery and airpliotos) in the Sudan have challenged the popular notion that the Sahara Desert is moving southward. No general advances have been recorded, making the basis for this assumption unclear. Over the last eight years, satellite imagery has revealed that the northern limit of green bio-mass may move between 50 and 250 km per year. Given thos'fluctuations, a long-term time series to record desert spread and evolution of the environment is n4w essential. Although resiklence is variable, the natural environment is probably more resilient than expected; therefore, identifying highand low resilience areas will be important in setting priorities. But there are great regional variations in degradation and, in some places, it is severe and irreversible. In Africa, degradation of forestry resources seems more critical in the Sudanian zone (500-800 mm) than in the Sahelian. Donor communities should not reject dryland projects in the <800 mm rainfall/year zones a priori; they should conduct or request studies of resilience and productivity in support of project design and reviews. How Farmers and Pastoralists Adjust to Land Degradation. How can Donors Help? (2.2) Professor Michael Mortimore said that uncertainty is a fundamental feature of dryland environments but, by adapting to long-run degradation and short-run drought and famine, farmers and pastoralists have created a set of risk-reducing strategies. Policy should pursue intensification in conjunction with diversification. Intensification is a land-conserving strategy, and diversification, if it takes ressure off the ecosystem, has the same effect. If drylands are to continue growing without adverse environmental or economic consequences, agricultural, livestock, and forestry interventions must be supplemented with appropriate policies for informal sector employment, migration, and urban development. 'A wrap-up was presented by Professor Stein Die at the last session (see page 65)/ -2-. Two areas of reseaA should be given priority: - Adaptive capability: Existing studies are selective and infrequent, and patterns are probably highly variable between place-and time. Adaptive capability needs to be charted for specific communities and, in particular, for facilitating structures in the macroeconomy. Degradation, management, and tenure: Intensification in dryland ecosystems must be pursued through integrated resource use. Conservation management is linked to resource tenure, which usually determines the future security of access rights. The nature of customary resource tenure, its relation to social deprivation and accumulation, and changes in the production systems require urgent investigation. Part M - Review of Experience What has Been the Experience With Dryland Interventions? (3.1) Juan Seve and Bruce Ross-Sheriff of E/DI (International Resource Group) attempted to answer this question (see working paper series, Environmental Policy and Research Division). They concluded with six principles as crucial to dryland interventions: 1. Require dryland project designs to state objectives and benchmarks in terms of process, and do so clearly. Make defining development success a clear priority over building institutions or structures. 2. Design and implement projects that make clear the commitment to the long term process (at least 20 years). 3. Invest more -- or influence the development assistance agencies to invest more -- in helping countries develop strategic planning expertise for natural resource management. 4. Invest a greater proportion of project resources (time, funds, expertise) in monitoriag and evaluating biophysical ionditions and process changes. Develop innovative ways to ensufe that feedback occurs from monitoring and evaluation through adaptation of project objectives and methods. 5. Sponsor training for project implementors at all levels; sponsor improved methods for project planning/monitoring/management. 6. Invest a significant proportion of each project's resources in on-site adaptive research to improve technologies. Part IV - Rangeland Issues Experience with Pastoral Asociations in West Africa - What Should Be Our Future Strategy? (4.1) This study by Poul Sihm triggered intense debate on what, if any, technologies exist to deal with this issue. Although some basic problems in rangeland/livestock development were identified, most agreed that agro-sylvo-pastoral systems are essential for the survival of semi-arid areas., Issues included-. - uncertainties about available technologies; - pressure from outsiders makes existing land tenure inadequate; - producer groups are disorganized and therefore have little or no political influence; - lack of integration (crop/livestock/natural environment/social/economic) in our understanding and approach. There are different views on the Bank's involvement. Some expressed concern that rangeland programs have too long a time span to justify substantial involvement. Others emphasized that there are important areas for assistance, including: - some limited level of physical investments; - improved donor coordination; - increased coordination of technical and social science assistance; - improved legislation and public policy; - promotion-of pastoral/producer organizations; - development of pilot projects; - - research on. range rehabilitation, social organizations and decision-making processes; supplemental forage sources; soil flora and fauna; animal disease/preVeition/control. The Holistic Resources Management (HRM) approach was described as a promising, though not yet fully developed, strategy for management of rangeland projects. The HRM has not been sufficiently supported by scientific evidence to-gain widespradacceptance. Its guiding principle is process rather than-product. -Because of the tight organization required, there is some doubt about HRM's feasibility in a common property rangeland situation. 'Part V - Water and Soil Fertility Issues (5.1 and 5.2) Papers by Chris Reij and Raymond Meyer suggested that several new and promising water and soil conservation technologies have appeared. However, the dissemination of knowledge is weak within the scientific and donor communities, as well as within t*e_countries. Moreover, the availability of information and data on such-technlogies is limited. A number of promising technologies for improving land productivity are available or are being tested. But the sustainability of such systems Is often poorly understood by the donor community. The discussion revealed uncertainty about the phosphate status of dryland soils. There is some hope that leguminous trees and plants may fix valuable quantities of nitrogen. Phosphate may have to be applied to avoid this becoming the limiting factor. More investigation is required. -4- The following principles were defined: - In the short term, productivity change in the drylands is difficult to measure because of large seasonal fluctuations. Investments must focus on the process of change and thus be long-term oriented. - An integrated approach is necessary -- look at the whole farm system. - Small scale (with the potential to be scaled up) and low-cost technologies are preferable. - Traditional production systems need to be examined, built upon and, if possible, improved (reconciled with new technologies). - A dynamic-adaptive approach to technologies (i.e., understand the farmer's motivation) should be applied in pilot projects. - -Water harvesting techniques are promising, particularly catching surface water. Under certain -conditions, the exploitation of ground water may be preferable. - Research should have a long-term focus (water, soil, and climate). Part VI - Policy Issues for Improving Land Management Why and How do Land Users Make Land Management DeelsionsTHow Can Policies Encourage Land Users to MAke Them Effectively? (6.1) Professor Piers Blaikie maintained that any policy designed to reduce land degradation must be based-on a comprehensive explanation of how land users make, or fail to make, decisions. There are five basic steps in formulating water and soil conservation policy: (i) identification and measurement of the physical processes of degradation; (ii) identification of those affected by land degradation; (iii) an explanation of why land degradatiidn is happening and identification of which technical and social determinants can be changed through policy; (iv) a review of government and local institutions ability to change land user practice; and (v) the formulation of a policy which matches (iii)-and (iv). The following points summarize some policy implications: - The land user's perception about land degradation and its importance may differ from the official perception. - The land user may often calculate water and soil conservation techniques in the context of the household budget as a whole, not just per sq. km of land. (The impact of non-tarm income is important here.) - Policy change must be matched to the capabilities of governments and complementary institutions. - Land tenure, access to land, labor,and capital within the household are important structural issues in policy formulation. There is a problem achieving common objectives, i.e, between the Bank, the borrowers, and the beneficiary population. Part V II- Off-farm Earnings and Migration Issues (7.1 and 7.2) Although, this section was based on studies from West Africa (Burkina Faso), by Thomas Reardon and Chris Delgado and by David Brokensha, the findings probably apply to other areas* - Non-cropping income is very important. It comprises 52 percent of income in the Sahel zone and 57 percent in the Guinean zone; however, it comprises only 26 percent of income in the Sudanian zone. - Local non-farm activities are most important in the best agricultural zone -- the Guinean -- since the areas with the greatest,lversification options are found there. - By contrast, the Sahelian zone's income diversification is derived more from other zones. - Off-farm income is used to compensate for variable cropping outcomes; There are still some uncertainties about the environmentul impacts of off-farm income (potentially negative: charcoal/firewood and mat production; potefitially positive: handcraft, repair service). Temporary and Permanent Migration in West Africa (7.2) David Brokensha outlined some early findings from the Onchocerciasis Preed Areas Study and focussed on what support the Bank should give to migration. The diversity of social, cultural, economic, and ecological conditions offers few generalizations, and site-specific approaches are necessary. Nevertheless, the presentation and the discussion emphasized the following principles: The rural-urban integration should be improved by information and transport - systems, e.g. roads, buses, and mobile markets. Seasonal and temporary rural-rural ,migration should be stimulated, especially between drylands and areas of high) agricultural potential. Likewise, as a general conclusion at this stage of the study, rural-urban migration should not be discouraged by the Bank. Part VIII - Participatory Issues (8.1 and 8.2) The presentations by Andrew Weir and David Richards, and the ensuing discussions, revealed the importance and complexity of local participation. Important because participation is vital for successful dryland intervention when marginal lano and marginalized people lack access to official channels. Complex because the issue is intermingled with institutional structure, organization, and politics (which sometimes exceed the mandate of the Bank). Beyond strengthening national and local extension services, it was unclear from the presentations and discussions what role the Bank can play in enhancing the capability of nongovernmental organizations and farmer groups. -6- Villager Involvement in planning and implementation Improves the effectiveness of interventions by increasing the information available to planners. This encourages site-specific flexibility and local responsibility. The Bank should apply appropriate emerging techniques more often, such as "Rapid Rural Appraisal," to improve local participation. General Issues The discussions and working groups established the need for and complexity of multi-sectoral approaches to dryland interventions. A clear division between the sectors, although In many ways artificial, proved useful in pinpointing crucial elements of project design, review, and implementation. Some common statements follow: - The productivity of the drylands is relatively low, therefore the economics of intervention must be carefully reviewed in a national context. However, investment must not only consider direct economic returns -- it must look at the social and economic costs of doing nothing. - Health and education are often significant incentives in even traditional rangeland and agricultural projects. Such components enhance the local people's confidence in the project, which may lead to increased participation. - Donor coordination must be improved -- in research, planning, and implementation. The Bank's role, however, is unclear since donor coordination is predominantly a borrower responsibility. - Multi-sectoral interventions must Pe developed, even though the experience has been somewhat negative with complex ind overly-ambitious initiatives. - The impact on environmental degradation of the price of capital (e.g., interest rates), commodity pricing, and subsidies for agricultural products remains uncertain and is a matter for site-specific investigation. Despite unreliable data on rates and causes of soil degradation, soil improvement seems crucial. - Land tenure and land security have a major influence on dryland interventions. It is, however, a complex issue that calls for careful study before moving ahead. - The role of women in traditional agricultural production is underestimated when developing assistance programs, and incentives should be targeted directly to them. - Small, successful demonstration projects are essential for "diffusion" and adaptation. PART II - INTRODUCTION 2.1 Drylan Degradation Assessment Techniques - A Summary by Professor Stein W. Bie University of Tromsoe/NORAGRIC Introduction The culmination of a prolonged drought in the Sahel in about 1985 was the start of a public debate on desertification. A general view appeared that Irreversible dryland degradation had taken place and was probably ongoing. Speculations arose about the causes of this alleged degradation, and its alleged irreversibility gave rise to doubts about the wisdom of investments in the drier areas. A bistory of debate Closer scrutiny of the literature indicates that similar debates have taken place on a number of occasions following droughts recorded in the Sahel during the last 100 years. There are ample indications - also from paleobotanic and geological records - that dr6ught in the Sahel is a recurrent phenomenon, and that drier and wetter periods occur. So far, there is no clear periodicity revealed. Recent measurements in the Sudan have challenged the frequently held notion that the Sahara desert is moving southward at an approxiiate speed of 5 km/year. No obvious general movements have been recorded, and the basis for the statement is uncertain.' Use ofsatellite_ imagery over the last 8 years has revealed that the northern limit of green bionass may move between 50 and 250 km per year. 'There is a need for a long-term time series to record desert spread. A debate obthe causes of famine There has been a wide-ranging debate on the possible causes of the famines that have arisen in the Sahel. Food shortages appear to-coincide with below-normal rainfall years, and there is general agreement that climatic fluctuation is a factor. Inadequate rainfall data make statistical analyses difficult. Similarly poor archival data on crop andlivestock losses and human suffering make past reports of famines difficult to compare with the two recent and better documented famines in the 1970s snd 1980s. Other proposed causes of famine include: higher birth rates and lower mortaity; changes in land use patterns; loss of alternative income-generating activities (e.g., the caravans); new trade patterns; and the erection of new national frontiers. 7 - Famine and land degradation Although the occurrence -- if not the severity -- of famines is well-documented, there is A scarcity of data linking general dryland degradation to famine. However, there is little doubt that local land degradation has taken place. There is therefore a need to develop better methods whereby land degradation can be assessed. In the past, inaccurate methodologies have had major impact on donor attitudes towards the occurrence, frequency and severity of dryland degradation. Desertlcation and dryland degradation The study proposes to define desertification as a subset of dryland degradation, indicated by very low vegetative biomass production over a large area over a prolonged period. Dryland degradation may also take other forms (e.g., salinization, alkalization). Measures of dryland degradation Two assessment factors of dryland degradation are proposed: 1. resilience 2. productivity. Resilience is a measure of ability of the land to withstand and recover from shock, i.e. a sudden impact (for example, drought, overgrazing, change in specified land use). Resilience is measured by changes in productivity of specified commodities, normally vegetrtive biomass of specified characteristics. If land, when subjected to shock, returns quickly to previous level of productivity (quantitatively and qualitatively), then the land is resilient and less prone to dryland degradation. Loss of resilience leads to dryland degradation and -- in special cases -- desertification. Permanent loss of resilience leads to irreversible land degradation. Resilience may be measured by combinations of satellite surveys, aerial photographs, field surveys and statistics. The value of historic air photos and satellite images is particularly stressed. The use of remote sensing allows for rapid and low-cost collection of resilience data when supplemented by field observations or yield records. Productivity records the level of production under a given land use. Loss of productivity of a commodity may be an indicator of land degradation. Loss of productivity of one commodity need not be mirrored by reduced productivity of others. Productivity is normally assessed by field methods and historical yield records. Extrapolations in space may be done by remotely sensed images (satellite and air photo techniques). Relations to Other Assessment Techniques Other methodologies of dryland degradation assessment and desertification have been proposed (e.g., FAO/UNEP provisional methodology, a Soviet methodology, a modified FAO/UNEP methodology tested in Kenya 1988/89, and others). The methodologies reviewed appear to rely on large sets of input data. Many data sets require detailed climatological, pedological and socio-economic studies. Although their scientific merits are not disputed, these assessments are generally both costly and time-consuming. They are therefore less attractive for direct, rapid, inexpensive and decision-oriented surveys for project identification and evaluation as-required by the World Bank and many other lenders and donors. 9 Conclusion The resilience of many areas of the Sahel is demonstrated through,the rap4 recovery in prqductivity following greater rainfall in most areas from 1986. Fears of irreversible dryland degradation in the form of desertification appeat to be overstated, although qualitative changes in biomass may at least temporarily have taken place. There seems no reason for lender and donor communities to reject dryland projects in the <800 mm rainfall/year zones a priori, but to conduct or request studies of resilience and productivity In support of applications for project funding or for evaluation purpoges. It is felt that such studies may give more direct decision-oriented data more rapidly than other currently proposed or implemented assessment techniques of dryland degradation. n1 2.2 How Farmers and PastoraHsts Adjust to Land Degradation: How Can Donors Help? by Professor Michael Mortimore University of Birmingham 1. Introduction 1.1 Terms: The term adaptation is preferable to adjustment because the latter implies fine- tuning of existing productive systims in order to maintain an equilibrium, whereas the scale of uncertainty and variability in the dryland ecosystems is such that structural changes within systems and, more important, major reallocations of labor between systems are frequently necessary. This should make it plain that I am looking at degradation as a process having multisectoral implications. 1.2 Experience: This paper is based on experience in the Nigerian sector of the West African drylands. I am not in a position to guarantee its applicability elsewhere. 1.3 Method: Can adaptations to land degradation be empirically discerned? To do so requires making the following distinctions: (a) Adaptation to long run degradation from adaptation to short run drought or food crises. Annual rainfall variability masks longer term changes in land productivity. Harvest failures obscure longer term trends in yields, for which reliable data are scarce. Years of poor grass growth (or exceptionally good) obscure longer term changes in rangeland. Uncertainty is a fundamental property of (African) dryland environments. (b) Adaptation to land degradation from adaptation to economic or demographic changes. Longer run changes in land productivity may cause and/or effect changes in the economic status of resource managers. Such changes of status occur within the community (in individual or family poverty/wealth), or between communities (changes in the internal terms of trade). 2. The vulnerability stereotype The view widely propagated in the media and accepted in some scientific literature and public statements by governments and donors is that dryland populations tend to be the passive victims of forces beyond their control. Observable exhaustic-a of food stocks and starvation, loss of assets including livestock, flight to relief camps or urban areas, and extreme dependency appear to support this view. To the extent that suffering is caused by rainfall failure, a lack of control is indisputable, given present technical capabilities. To the extent that it is due to warfare, political deprivation, or administrative corruption, deplorable as they are, control lies beyond the scope of technical or management prescriptions. But to the extent that this view assumes the absence of alternative livelihood options, or failure to use them, it is contentious if not misleading. 11 12 3. Adaptive capability A number of empirical studies of adaptive behavior in the African drylands confirms the existence of alternative livelihood options either in situ or elsewhere in the national or international economic space. These options are exploited energetically, even though (1) labor returns may be low, in the village or the city; (2) the migrants themselves carry all the costs of inter-regional labor transfers; and (3) they forego housing and every type of comfort. But during food crises, such options are valued against survival not cumulative objectives. The effectiveness of alternative livelihood options in the 1970s and 1980s in West Africa may be assessed from: (a) Rates of human mortality. Excepting some pastoral groups, there is no evidence that human mortality reached the proportions of the major famines remembered in the past. In northern Nigeria, it is hard to establish evidence for any increase. Where such an increase did occur, it tended to be amongst those unable to move. (b) The proportion of the population that ended up in relief camps. Again, possibly excepting some pastoral specialists, this was small (even in Ethiopia?) - though very visible. What were all the rest doing? The fact, and the means, of survival of so many people in the African drylands during the famines of the 1970s and 1980s are inadequately recognized and understood. The term resilience best describes such capacity in human systems to absorb shock without permanent fundamental change. 4. Policy directions The policy directions implied by the vulnerability and resilience views may well diverge. 4.1 Vuinerability: Forces are beyond local or community control, and people cannot easily help themselves. They become the object of rescue operations. Solutions are sought and imposed 'top-down' or from outside the system (new technologies, 'rational land use', compulsory migration, social reorganization). But interventions may have disruptive implications for livelihood systems which have proven survival value. Food aid may damage self-reliance and local market production; transforming farming or livestock systems may reduce diversity or flexibility, increase input dependency, or undermine the subsistence priority; social reformation may damage community autonomy. 4.2 Resilience: Proven adaptive capability provides a starting point for supportive policies designed to (1) enhance diversity in the economy, (2) broaden the structure of alternative livelihood opportunities, (3) facilitate movement, inter-regionally on (more controversially) internationally, and (4) strengthen community autonomy in resource management and labor mobilization. An analysis of adaptive behavior will indicate appropriate directions. S. Systematizing adaptive capability The following analytical breakdown is proposed: 5.1 Aapting to drought: This category comprises adaptations within productive systefis to counter the effects of reduced rainfall. In farming systems, it includes changes in soil management (e.g., fertilizing practice), in labof organization (e.g., labor hiring), in land use (e.g., fallowing practice), in cropping systems (e.g., in variety selection), and in technology adoption (e.g., ex-ploughs). In livestock production systems, it includes changes in herd composition (e.g., from cattle to small livestock), in grazing circuits (e.g., southerly displacements), in the timing of movements (e.g., in relation to water supply), and in fodder sources (e.g., from crop residues to purchased feeds). 5.2 Adapting to poverty: This category includes three types of behavior. (1) exploiting alternative livelihood options (e.g., trading, mat-making, laboring), in situ or elsewhere, by means of inter-sectoral labor transfers and mobility; (2) liquidating assets (e.g., livestock, personal property, land); and (3) mobilizing social networks based on kinship, clientage, or almsgiving (transfers in the form of loans or gifts of food, cash, productive capital or inputs). Poverty-adaptive capability can be strengthened by policies to preserve or enhance opportunities for diversification in the economy (e.g., the urban informal sector) and to protect freedom of movement (nationally or internationally). Such measures may support type (1) behavior. Type (2) ought to be restricted, in the best interests of the poor, while type (3) lies beyond the scope of policy. 5.3 Adapting to hunger This category of adaptive behavior follows from the exhaustion of household food stocks and consumer inability to turn to the market, or market scarcities. It consists in the intensified use of vegetable foods from farm trees and the natural vegetation, some unpalatable or even toxic without careful processing. A wide range of such famine foods has been identified in addition to those normally esten. Hunger-adaptive capability is furthered by diversity in the eco-system, the protection of natural vegetations, and the planting and protection of farm trees. 6. Adapting to Land Degradation 6.1 Long run adaptation: Short run adaptive behavior, as shown above, responds to short run stimuli and objectives. Land degradation is a long run process. This affects adaptive response in three ways: (1) the time-scale of adaptation is longer, (2) individual decisions relating to resource management and labor allocation have to sustain a pattern over a number of years, although each year tends to be treated as a separate accounting unit; and (3) productive systems as a whole tend to evolve in the same direction, including whole communities or regions. A productive system may become unable to support its dependent population either because primary productivity is declining (land degradation), or because the dependent population is growing relative to the land available. 6.2 Alternative pathways: Empirical observations of African dryland societies suggest that, in the long run, three forms of adaptation are discernible: (a) Negative adaptation: increased frequency of land use with fixed technologies and labor inputs: land degradation, with declining incomes from primary production. (b) Positive adaptation: 1. Land productivity held constant or improved by successive increases in labor inputs and/or new technologies: intensification. 2. Incomes maintained or improved from alternative livelihoods sought locally or elsewhere: diversification. The degradation (a), intensification (bI) and diversification (b2) pathways are available to communities either as alternatives or in combination. 14 6.3 Policy Implication: The degradation pathway is unacceptable. The scope for intensification, demonstrated in the triple system of agro-forestry with livestock found in the Kano Close-Settled Zone of Nigeria, is inadequately proven in respect of other dryland ecosystems, and may not be fully able to support population growth (as in the KCSZ). Therefore the policy objective must be to pursue intensification in conjuntion with diversification. Intensification is a land-conserving strategy and diversification, if it takes pressure off the ecosystem, has the same effect. The implication is that, if dryland populations are to continue growing without adverse environmental or economic consequences, agricultural, livestock and forestry interventions need to be supplemented by appropriate policies in relation to informal sector employment, migration, and urban development. (They will continue to grow even if there is a downturn in fertility rates and continued selective out-migration.) 7. Research needs 7.1 Adaptive capability: Existing studies are selective and infrequent, and the patterns are probably highly variable between places, and dynamic with respect to time. Adaptive capability needs to be charted for specified communities and, in particular, with reference to: (a) Blocking and facilitating structures in the macro-economy; and (b) The effects of poverty on adaptive behavior, especially alternative livelihood options available, perceived and used. 7.2 Degradation, management and tenure: Intensification in dryland ecosystems must be pursued through integrated resource use (farming - where climatically appropriate - livestock, wood-cutting, and collecting). Conservation management is linked to resource tenure, which controls social access to resources and the future security of access rights. Many interventions have sidestepped issues of tenure with respect to land, trees and water. This is no longer acceptable since intensification policies cannot be soundly based on ill- defined or fluid access rights. The nature of customary resource tenure, its relations with social deprivation and accumulation, and contemporary trends in response to economic change require urgent investigation. PART Ill. REVIEW OF EXPERIENCE 3.1 What Has Been the Experience with Dryland Interventions? by Juan Seve and Bruce Ross-Sheriff E/1DI, International Resources Group I. Abstract -- World Bank Drylands Management 1tudy A. The Drylands Management Problem B. Relevant Criteria for Successful Interventions C. The Drylands Problem D. Issues 1. Economic 2. Social 3. Institutional and Managerial 4. Extension and Training 5. Policy and Legislation 6. Technology and Research 7. Geographic E. Areas Requiring Further Study U. Abstract/Outilne of Presentation What Has Been the Experience with Dryland Interventions? This presentation is based on a study currently being conducted by the Internationa Resources Group on behalf of the World Bank. The study is based on a review of the relevant literature, including a substantial number of reports dealing with approximately 100 drylands projects and interventions. The literature review has raised a large number of Issues, among which the following are considered of crucial importance for the future of dryland management interventions. The presentation focuses on the findings of the study and on hypotheses formulated with respect to these issues. A. Technology. Are there enough technically proven and economicalli attractive natural resource technological management approaches available to halt the decline of rural production systems in drylands? Even if this is the case in soil management, forestry and agroforestry, is it so for range management? What can and should be done to take advrtage of technological approaches that show promise? 15 16 3. Economics. Importance of Including various types of economic incentives for beneficiaries as an Integral part of projects and/or interventions. Of particular importance are access to markets and development of marketing organizations, resource tenure rights and various forms of subsidies. Is access to markets a real possibility, or do dryland populations practice subsistence economies? How important is the resource tenure issue and what reforms could and should be implemented? What should be the role of subsidies in solving drylands problems? Where can and should the subsidies come from? How should they be applied for their maximum lasting effectiveness? Are there any trade-offs between economic development and environmental stability in a drylands context, or are they mutually reinforcing? Planning and Management. Is the traditional project approach effective in drylands? Is the time span of a project sufficient to provide lasting solutions to dryland problems? Should projects and other interventions be part of a longer term process (say 20 years) during which donors and governments remain committed? What adjustments would donors have to make in order to implement this long term process approach? Are small projects (NOO type) more effective than larger projects (World Bank type)? To what extent is lack of coordination (mainly involving donors and host government agencies) responsible for lack of success in dryland management activities? Is the direct involvements of beneficiaries crucial to success in dryland management intervention? Where should it start and where should it end (design, planning, implementation, etc.)? Should "replicability" be a basic concern in project design? 17 World Bank Drylands Management Study Abstract Despite decades of substantial investment by the World Bank and other donor organizations, populations living in dryland areas of developing countries have experienced virtually no gains in well-being. Perhaps the number of development projects has been insufficient to counteract the factors which result in economic deterioration. But a general consensus exists that most dryland development projects have failed, i.e., they have not resulted in sustained gains in economic productivity or improved sustainability. However, enough successes have occurred to reaffirm development professionals' belief that project interventions in the economic systems of dryland areas can foster sustained development only if the interventions can be designed and implemented correctly. The discrepancy between frequent failures and occasional successes implies that development professionals do not know all they should about designing and implementing projects in dryland areas. In 1986 and 1987, the Club du Sahel, USAID, and the World Bank each sponsored studies of successful interventions in rural dryland areas of Africa. In early 1989, the World Bank began a review of these studies and other literature on dryland interventions to determine if the necessary and sufficient conditions for successful dryland intervention are known and, if not, what needs to be done to learn the conditions. This abstract summarizes the results to date of the literature review stage of that effort. Most of the successful interventions identified in the 1986-1987 studies and in other studies were relatively recent and unsustainable levels of inputs from external sources were continuing in most cases. Thus, although the study reports usually state their conclusions as facts, most of the target audience for the reports -- development professionals in the World Bank and other agencies - - are likely to consider the study conclusions about the conditions that determine success to be unproven hypotheses at best. This is not unusual for development planning and implementation - - development professionals generally choose among opinions, not among facts. However, to choose among opinions, the decision maker needs to know which criteria should be applied to evaluate the validity of an opinion. This implies a degree of analytical rigor, i.e. a clear and explicit statement of the conditions that must be satisfied to cause an intervention to be successful, and an explicit definition of the measurements to be taken when evaluating the necessity and sufficiency of the conditions specified. Such rigor is lacking in studies reviewed thus far. The most difficult and controversial steps in gathering knowledge that will change development professionals' decision making are deciding the criteria for success and-the measurements for demonstrating successful approaches. As a first step in defining the necessary and sufficient conditions for success and in deciding the measurements required to validate the definition, we have identified a number of substantive issues in the literature review. These have been grouped into the following categories: economic, social, institutional/managerial, extension and training, policy and legislation related, technological and research, and geographic, - However, before addressing specific issues, it is pertinent to attempt a definition of "success in the context of natural resource management interventions. For this study, a successful intervention will be one that contributes to economic growth, improves the well-being of the beneficiaries, enhances environmental stability (or at least avoids environmental deterioration), reduces the risk of drought effects, and is economically and socially sustainable in the long run. In 18 this context, getting momentum started In a type of activity that has long term economic potential can be considered successful. *Successe does not have to be necessarily measured against a project's goals or terms of reference. Successful activities can be unexpected and different from what was originally intended. On the basis of the experiences examined, it can be stated hypothetically that, in order to be successful, dryland interventions will have to satisfy at least the following conditions: 1. involvement of beneficiaries in decisions and implementation; 2. market incentives; 3. business, commerce, or market oriented organizations at the village or small community level; 4. interactive extension, including technology transfer, and S. long term and evolving commitment on the part of governments and donors. The literature reviewed suggests that, in spite of the severity and complexity of dryland management problems, sufficient technically proven and economically attractive natural resource management approaches are available to halt the decline of rural productions systems in these areas. Better technological packages can and should be developed, but technology does not seem to be the major limiting factor in improving drylands productivity through natural resource management. Investment in development of dryland resources need not attend a technological breakthrough in invention of profitable, sustainable technologies for dryland agriculture, although such a breakthrough would, of course, be a boon. From the literature review, it has become clear that success cannot be attributed to technical methods alone. The right socioeconomic incentives have allowed marginal techniques to succeed, while their absence has caused excellent technical approaches to fail. In addition, a healthy, or at least non-deteriorating environment is key to economic growth in drylands. The evidence examined does not indicate a conflict between environment and development. On the contrary, profitable land management schemes may provide the best means to stop environmental degraJation in drylands. The evidence reviewed indicates that a major reason why rural development efforts, particularly in drylands, have not had lasting effects is the lack of economic incentives facing the ultimate beneficiaries. Improved economic incentives can be achieved in a number of ways including: 1. direct cash payments 2. food aid -- 3. subsidized crop prices 4. subsidized inputs and services (fertilizers, trees, crop seeds, transport, etc.) 5. subsidized capital investments (soil and water conservation works, equipment, tools, fencing materials, wells, etc.) 6. cash compensation for specific activities (protection of seedlings) 7. establishment of commercial and marketing organizations 19 8. tax exonerations 9. macro-economic policies favoring exports (devaluation) From an economic standpoint, the challenge for the donor community resides in projecf design and economic project appraisal. The literature reviewed suggests that few projects q interventions have been designed with the initiation of a sustainable economic growth process atia specific objective. Projects and interventions should not be concerned only with physicil accomplishments and institutional strengthening -- they should aim too at starting economic growth processes at the level of local communities. An important element in the design of these interventions will be their long-term commitment. From a social standpoint, the main problem seems to be the lack of analysis of social structures and dynamics in project preparation and design, which leaves interventions with a narrow technical focus. The literature describes a few examples of successful or promising interventions in which beneficiaries have participated actively. In general, sound preparation, years of study and discussion, and devoted management have been at the root of these successes. The importance of beneficiaries' participation is generally recognized by development professionals, but few projects have made the effort necessary to guarantee local input in project design. Another major social issue is the role of women. In a large number of societies living on drylands, women constitute the bulk of the labor force, while men usually control the decision process. Yet to date, development professionals have given too little attention to women's problems and potentials. Conflict resolution is an important issue. As the environmental crisis in drylands has worsened, so have conflicts between competing interests. Among these, attention must focus on conflicts between rural populations and their governments, between farmers and herders, between various producers for access to fertile lands, between agriculture and forestry, and between human and wildlife populations over land use. Donor and local government involvement in conflict resolution should take on increasing institutional importance in coming years, but it remains under-emphasized and under-funded. To a large extent, problems related to rural development in general, and to drylands in particular, occur in the managerial and institutional areas. On the whole, difficulties of dispersed population, highly variable climatic conditions, lack of trained staff and lack of infrastructure, among others, call for project designs that are flexible, avoid great complexity, do not rely heavily on capital intensive physical and institutional infrastructures, and aim for modest achievements. These attributes are seldom found in dryland management projects. On the contrary, one often finds overly ambitious project goals, emphasizir4 the size or number of interventions introduced over their quality, adaptability and long term impact. Typically, projects attempt to do too much at one time and reflect a lack of understanding of what will and will not work. From the standpoint of project implementation, the main problems deal with various aspects of coordination at all levels, namely between benefticiaries and implementing agencies, between beneficiaries and local governments, between various government agencies at a given level, between donors and governments, between donors, and between different offices of the same donor. This lack of coordination ii its varied forms has wasted time, effort, funds and other resources, and may be contributing to the persistence of underdevelopment through expensive failures. A third basic managerial issue is the monitoring and evaluation of project or intervention management. This is seldom a priority, especially in regard to social factors. Monitoring and control systems should be an integral part of each project or intervention, not only in an accounting or budgetary sense, but also to ensure that the intervention corresponds to objectives and work plans, that it remains strategically consistent, that it continues to respond to the needs of the beneficiaries, and that it signals the need for changes and adjustments if events so require. In the area of extension, the main problem is the lack of interactive approaches. Extension activities acquire more importance and require more time and dedication as novel initiatives are introduced. Many techniques, even if they have low technology input, often require a long period of extension to ensure acceptance. On the other hand, top-down extension methods have proven to be less successful than interactive approaches, especially when themes conflict with local knowledge. Another issue related to extension and training is the diffusion of knowledge acquired through a project or intervention. In mbay cases, a great deal of training is imparted by project staff to counterparts and other project-related people with little, if any, linkage to national or regional training centers. Finally, the demonstrated effect of visibly successful interventions needs to be exploited. Of course, it must not be assumed that a given activity will magically spread because of demonstration effect. Viewing a successful operation is only a good starting point which normally requires dedicated and patient follow-up. From a policy and legislation standpoint, the most important issue is the uncertainty of resource tenure rights, particularly of land and trees. The environmental crisis has created the need to invest in land to protect it from degradation, but the insecurity of tenure has reduced the incentive to make such investments. This is especially acute in several African countries where property rights have traditionally been ambiguous and where, in many cases, rural lands became national domain at independence. Perhaps the most valuable policy changes are those which allow local communities the freedom to organize natural resources management systems within general resource sustaining guidelines. Essential to these efforts have been regulations transferring land and tree tenure from the state to rural communities. This allows for the creation of organizat;ons outside of state institutional structures, and empowers local authorities who can better monitor changes in soils, water and plants. In addition, conflict resolutipd must be generalized. This will help resolve the basic cleavage between public and private motivation regarding resource stabilization in dryland areas. Another major point of debate about drylands management is whether there are sufficient technological elements to implement sustainable development interventions. While no one knows if existing technological elements are sufficient and adaptable to all situations, a considerable number of interventions examined indicate that lack of appropriate technologies has not been a major factor in the numerous dryland management failures. This begs another question: if there are sufficient technologies, what then is stopping their application? In any event, it must be pointed out that technological change, with or without external influences, is not absent from dryland environments. A number of examples suggest that there is sufficient range of technically proven and economically attractive technological elements available to halt the decline of rural production systems in dryland environments. These can be grouped into water conservation, soil conservation and forestry/ agroforestry. In many cases, effective technologies have built upon existing land use practices. This generally leads to lower costs and better acceptance. 21 The literature review suggests that research in drylands should focus on two major topics resource inventory and improvement of applicable technologies. In addition, all dryland management research activities, whether they focus on resource knowledge or technologies, must explicitly consider the capricious nature of rainfall. From a geographic standpoint, regional differencesbetween Africa, Asia and Latin America, as well as between regions within these continents, are important In explaining why some approaches succeed and others do not. Afterthis literature review, it is clear that genera!zing about approaches, designs, or technologies is dangerous. There are fundamental differences between dryland situations in Latin America, North Africa, Sub-Saharan Africa, India and Southeast Asia. Key variables include resource tenure Institutions, the degree of peasantization, importance of Irrigation, demographic and ethnic characteristics, kind and magnitude of state bureaucracies, degree of 4ecentralization, degree of sophistication of markets and marketing channels, lanrastructural development, and post-colonial trajectory. The impact of these variables, critical as they may be, does not seem to be sufficiently incorporated into analyses of project activity or project design. Additionally, these variables are often an obstacle to the replicability of a given'Intervention. PART IV. RANGELAND I$SUES 4.1 Experiences with Pastoral Associations in West Africa: What Should Be our Future Strategy? by Poul A. Sihm, AFTAG Wqld Bank Summary The paper sets the stage for the crisis in the arid, semi-arid, and sub-humid pastoral areas with some statistis4nd an outline. It concludes with the statement that the establishment of pastomal associations (PAs), the allocation of the natural resources they traditionally exploited, and the training of Its members in resource management and marketing wil arrest .the degradation of natural renewable resources in those 'reas. In addition, it will eventually provide governments with an effective partner in environment rehabilitation and management. A preliminary definition of a pastoral association is advanced, and the paper then tries to see these associations from the point of view of the pastoralist, the government bureaucrat and the development bureaucrat. The paper then reviews the Bank's experience with projects whose main objectives include the establishment of pastoral associations in the following West African countries: Zaire, Senegal, Central African Republic, Burkiii Faso, Niger, MMI and Mauritania, as well as the intentions of two new projects in Chad and Cameroon. The description and datp?n the projects have been included inan appendix. The paper then opens a discussion of (a) whether the pastrl production system is viable; (b) what is required to make PAs successful; (c) what is required to establish a pastoral association; and (d) why the Bank should be involved in the financing of such projects. The author concludes the following. a. Ifriven the right policy environment, PAs are viable, low input production systems indispensable for the survival of marginal agriculture in the arid, semi-arid andsub- humid zones; b. Necessary ingredients for successful pastoral assodiations include: a national policy; the organization of the pastoral population into producer groups; the allocatiofi'-of usufrqct of water and grazing; help with marketing and fair pricing; technical advice and training them to manage their affairs and their eprironment; ~~1 ~23 ). 24 c. Time Is the most important requirement; then training in literacy and self- management, as well as acquiring inputs and providing their own services; further requirements include technical assletance and initial support in setting up security stocks, a revolving fund, and a grant to be used as seasonal credit for the benefit of the poorer members of the group; d. The Bank must assist governments in establishing the policy environment in which the associations may thrive. Since this form of development is not replicable from country to country, the Bank must also be involved in testing the model for each country over a long period of time. The paper concludes that (i) the success of PAs is still a matter of faith simply because they appear to be the only solution to the dilemma; (ii) economists need to look further ahead than financial discounting will allow and contemplate the cost to future generations of doing nothing; (iI) there is great promise in long-term low input development as explained in the paper, and (iv) the Bank must be involved because: (a) its intervention is needed to create the right policy environment and to test models, and (b) PAs eventually may become the implementers of environmental management in their areas. This may have implications outside the PAs' immediate areas. I. Introduction This paper is a first attempt at examining the World Bank's experience with pastoral associations in West Africa. On the basis of personal knowledge of most of the projects and reports, the author reviews seven projects in seven different countries, compares their components, their performance, and some of the reasons why they have developed with different speed. The paper discusses a few pertinent questions relating to PAs and tries to draw some conclusions. The pastoral area of Sub-Saharan Africa (SSA) in the context of this paper comprisesA 9 Mill. square km of Tsetse free lad in the pastoral mode or 40% of total SSA land area. 57 Mill. active agro-pastoral population or 25% of total SSA population. 72 Mill. TLU of livestock or 53% of total SSA livestock population. The 40% of the total land surface of Africa which is in the pastoral mode plus approximately another 10% which should not be in the agricultural mode can only be used for extensive livestock production, game production, national parks, or recreation. This is a description written by a livestock specialist. A forestry specialist would likely have a different description. The 72 million TLU which presently exploit these areas will, on a very low assumed offtake rate of 10%, produce 7 million TLU or 1.7 million tons of meat/year, 5 million tons of milk, as well as 7 million hides and skins. 'All figures are based on Hans Jahnke's "Livestock Production Systems and Livestock Development in Tropical Africa," Kiel, 1982. The figures presented above were arrived at by assuming that the land of the arid zone south of the Sahara and half of the land in the semi-arid and sub-humid zones are in the pastoral mode of production. To arrive at a population figure, the author has, considering the participation of women and children in livestock production, counted 60% of the total population as active (and not the 38.5% used by Jahnke for the ME-man equivalent). The TLU (or UBT) is the equivalent of an animal of 250 kg liveweight. 25 Much has been written and said about the crisis in the arid and semi-arid zones of Sub- Saharan Africa. The demands of a growing human population and the availability of more manpower have generated a growing livestock population. Pastoral zones have decreased in area and quality of grazing, as well as in security. This is due to an ever expanding cultivation which has not only devoured the dry season grazing areas of pastoral groups, but is even breaking the fragile wet season grazing for the promise of one to two low yield crops in five years. This may be economical for the cultivator since the social cost is carried by the pastoralists and the economic cost by the country. The pastoral societies have tried their best to survive. Their former power has been broken by colonial and post-independence governments that opened the grazing areas to everybody through the construction of public water points. The pastoralists are now without the insurance of their wet season grazing resources and necessary inputs. They also lack the political power or legal means to defend themselves against the invasion of their communal pastures during the wet season. The terms of trade have turned against them: the price for their livestock in drought years is not in balance with the price of the cereals they need. Moreover, officials are harassing them with legal and illegal demands for payments. This imbalance in prices has led to a major change in the ownership of livestock which, during each drought, is purchased from the pastoralists by farmers, traders and government officials. These buyers use livestock, especially cattle, as a savings bank or insurance against a failed crop year. During the growing season all the "investors" return their herds to pastoral areas which ate often herded by the same pastoralists from whom the herds were bought. There they compete with the remaining pastoralist herds for the grazing secured for them by their access to public water points. The combined pressure this creates on a dwindling grazing resource during the wet season leaves little standing hay for the dry season. The result is degradation of the grazing areas and increased cutting of trees and branches - often by pastoralists wanting to construct their traditional wells away from public ones in order to secure them and the grazing they command. The superstructure of a traditional well uses about 300 branches and, if the well is in a sandy area, about 150-300 trunks are necessary to prevent the wells from caving in. In order to stop the degradation of the pastoral areas we must reduce the pressure on the grazing land, roll back the limits of cultivation, and rehabilitate grazing space. To achieve this, we must allocate grazing to the traditional user, put a stop to the annual free grazing by the "investors," and redress the balance of cereal and livestock prices. This is easier said than done and, politically, very sensitive. Many economists believe that the rehabilitation of pastoral areas is not economic. If one considers the economic cost of production foregone of food aid and social services every time drought breeds disaster in the Sahel -- as well as the national cost of 25 million pastoralists begging in the towns of the Sahel and of restoring the environment 20 years hence -- then the economics look different. This paper maintains that, if we create PAs and train them to take care of themselves and their environment, it will cost less than going from crisis to crisis, and it will eventually reduce the cost of emergency aid to zero and render pastoral production systems productive. This, however, requires a consistent, long-term policy that has the durable support of donors and governments. H. A Pastoral Association (PA) and Its Purpose A PA could be defined as: a legally recognized, voluntary association of pastoral livestock owners who use a common area for grazing and frequent the same water points for the purpose of managing natural resources (water, grazing and other vegetation), for the procurement of inputs and services and the selling of products. As time goes on, more elements may be added to this definition. 26 The purpose of a PA is to ensure the security of the community and its livestock, and to promote the welfare and wealth of the group by better managing common resources of water and vegetation for optimum livestock production. It supplies animal production inputs and provides first aid in animal disease and seasonal credit to poorer members. The association can enter into partnerships with donors for well rehabilitation or construction, and for environment management projects. The pastoralist desires food and forage security, human health services, secure waterpoints, access to grazing, free movement in case of drought, availability of veterinary drugs, adult literacy training, schools for their children, and a reasonable price for their livestock and the commodities they buy. In the minds of government bureaucrats, the PAs are meant to afford them complete control of people and animals through the interventions of the extension services. The aim is to bring order into a seemingly disorderly situation, bring social services to the population, and to get animals into the market in order to feed urban populations and to ease the pressure on the pastoral environment. Meanwhile, most government officials privately want to use pastoral livestock as a savings bank. The development bureaucrat, depending on how much he understands the pastoral production system, is concerned about the environment, the cyclical drought, and human starvation. For many years the development agencies acted on a number of myths about the pastoralist largely furnished by government bureaucrats since, at the time, the dialogue with pastoralists was not as developed as it is today. We are finally beginning to study the production systems and understand how they function. We now realize that the pastoral livestock owner, if properly trained and assisted, may become both a supplier of meat and, in some cases, milk. But, if allocated the land and water points he uses, he becomes manager of these resources and, eventually in his own interest, the protector of the environment. This is possibly the most important side effect of the creation of PAs. If a union of PAs is organized and given control of natural resource management at the district and the regional level -- as is happening right now in Niger -- and eventually at the national level, it is bound to have a political impact which may finally be able to change the ingrained attitudes of politicians. HI. Bank Experience with Pastoral Associations The paper only deals with PAs in IDA-financed projects in West Africa. Later it is the intention to expand the paper to include our experience in East Africa, notably the Kenya Group Ranches, as well as the experience of other donors. In the appendix to this paper there is a description and summary of the various projects in West Africa financed partly or entirely by IDA credits involving the establishment of PAs. In chronological order, credits for this purpose have been or are effective in Senegal, Zaire, Central African Republic, Niger, Burkina Faso, Mali, Mauritania, Cameroon and Chad. Not all started out with the creation of pastoral associations as their main objective, and not all have followed the same development path or are today at the same level of development. Senegal and Zaire are the most advanced in the establishment of PAs, although the latter seems to suffer setbacks in the transition to self management. Niger has moved quickly once the allocation of resources issue had been solved. Mali was well ahead on implementation but lost ground because of the lack of political support from government. Mauritania, although a latecomer, promises to move quickly. Burkina failed, and it is too early to report on Chad and Cameroon, where the IDA credit has only just become effective. When we look at the advancement of PAs, it seems there are common factors important to their development, and which we consequently need to monitor -- namely, leadership, legislation, land tenure, waterpoints, food security, health services, literacy training, marketing, working capital and the concern for the environment. 27 The leadership of PAs is important. Like elsewhere it makes a difference whether the leadership is the traditional group of elders or a change of generation has taken place. In most PAs the traditional and hierarchical leadership of the successful pastoralists govern the less successful ones -- who have little regard for equity but a great concern for the common welfare of the group. Wealthy members of the group who have acquired wealth in non-herding occupations are often the initiators and movers of the establishment of the PA -- digging the well, installing the pump, etc. The traditional chief and the Islamic priest are important but often exert their influence very discretely. Legislation is important. PAs perceive early in the process that it is needed to give them security in their allocated area. When one considers the absence of land tenure regulations in West Africa, it is no mystery that clear legislation has not yet emerged 'or the all9cation of natural resources to the PAs. What is equally clear is the need for legislation. irst, the eed for legislation to confirm the allocation of grazing and waterpoints and the duty of the it al s thorities to enforce it if they are called by the PAs to do so; second, the need for legislv AP to annul earlier laws establishing public access to waterpoints constructed by the state; and thard, the need for detailed legislation to confirm the institution of the PA itself and its relationship to local institutions and to government. Legislation has advanced the process mostly in Senegal. There the existcnce of a law on *pioneering zones " (Zones Pionieres) -- establishing Common Economic Interest Groups (GIEs) - - allocated the management of land and waterpoints to the groups and offered them protection against the intrusion of outsiders. Niger stayed with the cooperatives concept when it created "Groupement Mutualistes Pastorales" (GMP) and "Groupements a Vocation Cooperatives" (GVC). In the other countries, governments decided to use whatever legislation was already available until it was clear what specific legislation would be required. it is now obvious that -- apart from the cancellation of the public's access to waterpoints constructed by the state and a law on the allocation of grazing, waterpoints and trees -- a law is needed which enables groups to invoke the help of the local authorities against invasion by outside groups. Waterpoints determine the access to grazing. Animals depending on their species and the season can only graze a certain distance away from the well. For lack of a well or because a well has fallen into disrepair, an area may be inaccessible to herds. So wells need to be constructed and rehabilitated. Unless one talks about the repair of a simple surface well, this generally costs more money than the association can muster. Unfortunately, part of the program in Mali was delayed because a cofinancer had insisted that the PAs to be established around 15 subsidized boreholes had to obtain a loan in the Agricultural Credit Bank (BNDA) to pay for the storage wells constructed in connection with the boreholes. The PAs claimed interest in establishment and the construction but refused to tie themselves to a loan which, they said, they would have difficulty repaying in a bad year. In another part of the project the rehabilitation of wells, financed by the PAs themselves, went very smoothly, mainly because the investment demanded was of a manageable size. In Niger it took seven years before government was ready to construct a waterpoint for a PA. In Mauritania the subsidized rehabilitation of four existing wells and the construction of one new well is part of the establishment package for a PA. In Burkina Faso the PAs under establishment could not get to the wells in their area because landless peasants had settled around the well. and the government was unable to resolve ttt problem. Food Security is high on the list of the PAs' priorities. This has been handled in various ways. Most of the associations in Burkina Faos, Mali, Senegal and Zaire are sufficiently close to local markets to take care of their own needs. In Niger, where the pastoral zone is very large, the project initially assisted by financing the establishment of Relais Pastoraux (outlets for daily needs) at points where associations were being developed. Now that the first associations are in place the members take care of the problem by establishing a shop which sells daily needs against cash. 28 In Mauritania the first activity of the new PAs is to establish and operate a small dairy unit and a small fattening unit to assure supplies of meat and milk for sale to families during the dry season. The investment is handled through the PA revolving fund, which is reimbursed from the proceeds of sales. A vegetable gardening project is also being financed by one of the PAs. Health services and human drugs often are the highest priority of pastoralists. Their preoccupation with health is understandable since they have many acute and chronic disease problems, which are particularly aggravated during the dry season. Although services and drugs for sale are being provided by regional health authorities, the PAs have given them so much priority that the projects have, in various ways, had to provide support. In Mali, the government provides training of health workers and midwives and has established small clinics through the project, although the clinic is constructed by the villagers. In Niger, pastoral centers include a dispensary, and the PA has constructed a small hospital next door. This limits the number of cases which have to be evacuated. Human first aid medicine is sold and dispensed in the camps by health workers, and training of midwives is being conducted at the centers. In the other countries, separate services also exist, but are not yet strong enough to deliver human health in pastoral areas. Literacy training. Having board members and committees who are literate seems to add to the dynamics of development. Literacy training was the cornerstone in pastoral development in the projects in Senegal and Zaire and is important in Mauritania, where all members of the Board of a PA are literate in Arabic. In Niger, PAs build schools next to the pastoral centers. In Mali, literacy training is moving very fast with approximately 2000 pastoralists being trained. Thirty to fifty new training centers are also being established, and 200 trainers are educated each year. Marketing is a concern of the pastoralists for three reasons. First, they want markets where they can conduct their exchanges without the interference of legal or illegal taxation. Second, by gathering large numbers of animals for sale, therefore attracting several principal traders, they want the best possible prices. Third they want to protect themselves against manipulation by cartels. In Mali 15 livestock cooperatives have been created, mainly in market towns where enclosures have been erected by the cooperative to which only sellers, buyers, and the agent of the project have access; thus, entry is denied to the police and other officials who have in the past used markets to collect taxes or fines legally or illegally. PAs in Senegal and Niger are getting ready to establish their worn marketing and supply cooperatives. In Senegal, one of the major follow up activities has been the establishment of smallholder fattening units (average two heads of cattle and two sheep or goats) to improve income. In the Central African Republic (CAR), where the livestock owners have organized the National Association of Central African Herders (ANEC), all supply functions and animal health activities are managed by that organization in conjunction with the Livestock Services. Working capital for communal projects and the provision of animal inputs to poor members of the groups on short term credit is indispensable and is being established in various ways. In Mauritania each member contributes, through his representative on the board, an annual sum which is employed by the board. In Mali, an initial working capital has to be deposited in a bank account before the process of establishing the association starts, and payment for well construction is based on payments from members collected when watering their livestock at the well. In all countries, cost recovery has been introduced for drugs and vaccinations. In Mauritania payment for drugs and vaccines for animals, together with a tax on private imports of drugs and feeds, goes into a special account to finance the operation of the livestock services in the field. In CAR, membership dues from the 49 PAs (Herders Associations) provide about US$ 1.0 million a year for common projects which are used for the operation of ANEC, as working capital for the purchase of drugs and feeds for sale to members, and for the purchase of members' cattle for 29 marketing. Lately, profits have to a large extent been returned to members in the form of rebates on dues and taxes. Environment is an expressed concern of PAs in all countries -- but it is the last concern. In the ranking of concerns environment rehabilitation and management appear after the benefits of all the prime concerns -- i.e. control over resources, food and health security, and control over marketing -- have been secured for the community. There are two things to remember in this context. First the environment is where the pastoralists make their living. They are so much a part of the environment and know it so well that they expect it to bounce back, as it always has, every time resources have been hammered by a combination of drought and grazing. Second, when they really get conscious that something needs to be done, the task appears too huge and beyond any means they can muster. And it is. Environment is a main concern for the newly established PAs in Mauritania where committees have been created for each sub-group of 20-30 families, and the cutting of branches and starting of grass fires is subject to heavy fines. In Senegal, Zaire, and Mali the construction of firebreaks is undertaken in some of the older PAs, and the regeneration of pastures is taking place around some villages. In all cases, pastoralists have difficulty linking environmental damage to stocking rates. Some ecologists support them, stating that the lack of hoof impact on rangelands is a main cause for deterioration.2 IV. Discussion A. Are Pastoral Systems Viable? Many scientists are watching the decline of the pastoral economy, the degradation of the rangelands, and the hunger and death brought on by severe drought, and feel convinced the pastoral production systems are breaking down and will eventually disappear. Until we have more data this cannot be sustained. A large atiount of the myth surrounding the pastoral production systems has been debunked in the last 15 years. Because we are learning more about the functioning of the pastoral systems and the effect of grazing on rangeland, we are beginning to doubt our present explanation of *desertification" as the outcome of overpopulation and overgrazing of the rangeland. We have also learned a lot more about ownership of livestock and how Pfter each drought increasing numbers of livestock, in particular cattle, are acquired by anybody whose income is not influenced by the drought, i.e. traders and government officials. Agricultural people are increasingly investing their savings in cattle. During the rainy season (growing season in agriculture) all these cattle return to the pastoral zone where they compete with the cattle of the resident livestock owners for the available grazing and water. A typical example is Niger where this is easier to study because of the delimitation of agricultural and pastoral zones and the absence of other major occupations, but it happens in varying degrees in most countries. The dumping of subsidized meat from developed countries depresses the meat price in the region. And, when meat can be brought in at half the price from surplus stocks in Europe or Latin America, markets in the coastal countries disappear for the Sahelian pastoralist. For pastoral production to remain viable, natural resources in the pastoral zone must be allocated to groups of pastoral producers who must then be made responsible for management of these resources. Services and good advice must be provided and production inputs be made available. 2AIan Savory, Holistic Resource Management (1987). 30 Pastoral areas, like the agricultural areas, will continue to have surplus populations so that, even after a rehabilitation of systems which are without animals, there will have to be a certain emigration by those who lose their livestock. Formal education, which is more and more sought, will also result in emigration of young pastoralists. The bottom line is that agro-sylvo-pastoral systems, given the right policy environment, are viable low input production systems and, moreover, indispensable for the survival of agriculture in the semi-arid areas. B. What Is Required to Make PAs Successful? A national policy must be established. Producers who exploit the agro-sylvo-pastoral zone in each country must be organized into groups, and open access or common property grazing areas must be allocated to its traditional users. Agricultural extension services must be strengthened and production inputs must be made available on sale to the producers. The usufruct of water and grazing resources must be allocated, at a minimum, during the 3- 4 months of the dry season. In some places it is possible to allokiate both wet and dry season grazing. In due course land which should not be cultivated must be returned to pastoral use, or special legislation must be created for the agropastoralists of the intermediate zone. In addition, the groups need assistance getting their marketing organized and getting a fair price structure. Savings and loan schemes must be introduced in the associations to increase internal credit possibilities. PAs promise to become an important tool in environment management, and they should be supported financially, institutionally, and by enabling legislation in the rehabilitation of their own environment. Finally, associations must be organized into regional unions in order to increase the political clout of the pastoralists. C. What Is Required to Establish a Pastoral Association? First of all, time. The preparation of the minds of the group leaders, the inventory of land and herds, the consultations with neighboring groups regarding the validity of the claims to grazing made by the group, the consultations among the group members themselves -- this all takes time. Our present experience indicates that it takes about a year to create an association. In Senegal it took another ten years to make half of the initial groups self-managing. The process of developing self- management starts with an adult literacy campaign, the training of group members as "auxilliaires" (lay animal health assistants), followed at every stage by support from the extension services. Initially, there needs to be some outside incentive which is attractive to the groups; something that makes it worth their while to establish an association: (i) a well repaired or constructed; (ii) a security stock of veterinary drugs or feed; (iii) a revolving fund; (iv) the rehabilitation of production by providing for restocking of families who lost animals during the last drought; or (v) a human health facility. The important principle is to meet any reasonable priority need identified by the community and then identify which activities are achievable by the community and which activities need outside support. Credit, in our opinion, is not an early option since the PAs are trying to survive and their repayment capacity is impaired. A grant which can be used by the group as credit internally may be a better way. Credit should be reserved for groups who are able to start a revenue earning production. D. Why Should the World Bank be Involved? The Bank needs to be involved because pastoral systems are viable given appropriate policies, and there is a need to keep up pressure for policy reform through Sectoral Adjustment Lending (SECAL). The policies needed to bring about the equilibrium in the pastoral areas are not going to be popular with the politicians and authorities because they curb the rights of the agriculturists and traders and, more importantly, the rights of the government officials themselves. 31 Coordination of donor involvement is required, as well as the establishment of a national program which sets out the *rules of the game." Since a model for a national environment program is not replicable from country to country, we need to complete an initial model for the agro-sylvo-pastoral zones of each country. We also need to see the necessary reform and legislation through its first ten years -- the approximate time it takes to establish the necessary policies and legislation and for the first associations to mature to self-management. V. Conclusion Knowledge about pastoral associations is developing very slowly. We are all anxious for answers to our many questions. The Bank is particularly anxious to know if we are making progress in pastoral development and environment management. Whether the PAs will fulfill our expectations is still largely a matter of faith. Until now, the technician has been able to say: I know of nothing else that will work, so have patience. Whereas it may be too early to proclaim the success of the PA model, it is not too early to imagine what will happen if nothing is done in the pastoral zones of Africa. Most economists using present methods of financial discounting say pastoral development has little economic impact. Others say we have to look further ahead and contemplate the cost to future generations of pastoralists begging in the cities and of rehabilitating the degraded grazing land. Some ecologists despair while others offer bright visions of rehabilitation of rangelands through concentration of animal impact. In countries like Niger, animal export is the cash earner which provides outside income to keep subsistence agriculture in business. In all countries soil fertility maintenance depends on the integration of livestock and agricultural production systems. In resource poor countries it is not possible to ignore the contribution livestock can make. Personally, the author sees great promise in sustained low cost interventions of (i) organization, (ii) allocation of resources, and (iii) relevant training in resource management and financial support for pastoral groups to achieve: (a) literacy, (b) a reasonable health standard, (c) education, and (d) the rehabilitation of their environment. The individual lending operation on the regional and sub-regional level may be too modest in volume and too complicated in implementation to be the type of operation which we at the Bank should lend for in the long run. But I do not see that we can be excused from putting in place a national program and policies, and to demonstrate that it works at the field level. By promoting policies of orderly allocation of resources, we might well be the only organization which can halt the land grab going on in African pastoral areas. This fits well with our approach to natural renewable resource management -- environment management as we call it in the Sahel -- and it leaves plenty of room for bilateral donors. Financing of the main program, once proven viable, thould preferably be by grant funds rather than loans. As indicated elsewhere in this paper, the pastoral groups may indeed become the most important implementers of such national programs. On the basis of limited experience, we may expect that if nothing is done about the anarchic exploitation of natural resources, the oscillations between boom and bust in the pastoral zone will increase dramatically. Each time, recovery would be longer, and each 4rought would make more people refugees. This would eventually result in the collapse of traditional pastoral societies and cause emigration to the towns -- where traditional social security systems are already breaking down under the weight of an increasing number of landless and unemployed. The financial, social, and - ecological cost of such a scenario would be enormous, and some countries would find their viability and existence threatened. PART V. WATER AND SOIL FERTUM,,!SSM 5.1 StrategIes for Inltlating and Acceleratle - Moisture Conservaton Prograsns In Sed-arld We Africa by Chris Reij Free University, Amsterdam A Top-Down Versus a Participatory Approach to Soil and Water Conservation Basically, there are two contrasting approaches in soil and water conservation: a top-dOwn approach and a participatory approac. The former, which is still predominant, is charactefized by* the use of comolex techniques which cannot or will not be replicated by farmers; a reliance on the skills of conservation experts and technicians; - a low priority for training farmers; - villagers (men, women and children) employed as laborers and paid with cash or food; - often, but not always, a considerable reliance on machinery. The participatory approach is characterihed by: - the use of simple, replicable techniques; - a reliance on the skill-and motivation of farmers; - a high priority for farmer training. Which Approach to Choose? An analysis of soil and water consirvation (SWC) projects in Sub-Saharan Africa has shown that most projects characterized by a top-down approach have failed because the conservation works, which were constructed -at considerable cost, have subsequently not been maintained by the beneficiaries and, as a result, usually have degraded within a few years. Furthermore, the beneficiaries never applied the conservation techniques voluntarily to their own fields, which were not treated by the projects. -Some major SWC projects in Sub-Saharan Africa are still heavily top8own butiiven the historicalevidence, there is sufficiUnt reason to have doubts about the sustainability of these projects. The best que can hope is-that thiconservation works will tmaintained which, as experienceshows, is quite unlikely. The area treated during the project phase will also not be expanded in the post- project phase. The only way forward eemsto bea more participatory approach to WC. We should 33 34 aim at the involvement of hundreds of thousands of farmers who treat part of their cultivated land each year. It will be clear that this is a major challenge. Do We Have Replicable SWC Techniques? If we want farmers to apply SWC techniques to their own fields, they should have the following characteristics: a. they should be simple enough to be mastered by the farmers themselves; b. they should be low-cost, otherwise they cannot be applied on a scale of a hundred thousand hectares and more; c. they should be efficient, which means they should have a considerable impact on yields, preferably in the first year, and permit a yield in years of low and erratic rainfall; d. they should require minimum maintenance, otherwise total maintenance requirements will be higher than what farmers are able or willing to spend on them. Conservation technology responding to the criteria mentioned is at present available for-areas with stones. Stone contour bands were introduced in 1980 by the OXFAM funded Agroforestry Project in the Yatanga Region of Burkina Faso. In 1982/83, about 150 hectaree in this region were treated by farmers with stone contour bunds and the "guesstimate" for 1987/88 is about 5000 hectares (Reij, 1988). The picture is more gloomy for areas without stones. Ear bunds have generally not been maintained by farmers in Burkina Faso and Niger. Demi-lunes and trenches have only been constructed on a modest scale by some projects, mainly in Niger. Tied ridges have been tested on research stations, as well as on farmers' fields, but they have nowhere been adopted by farmers on a significant scale. Are Simple Techniques Always Repicable? Simple techniques are not always replicable. The lack of adoption of demi-lunes in Niger is a good example. Demi-lunes are easy to construct, they are efficient in that they make crop cultivation possible on lands that-would otherwise not be cultivated or, in the case of tree plantations, they increase survival rates and growth rates of trees. Nobody has really done researci_ on why demi-lunes are not replicated by farmers. A major reason could be that labor requirements for construction and maintenance are considerable. Can We StW Use Traditional SWC Techniques? In regions where traditional techniques (stone lines, stone bunds, pits, etc.) are still maintained and expanded, such as the Plateau Dagon in Mali and the Ader Doutchi Maggia in Niger, one could build conservation programs on the already known traditional techniques. The efficiency of traditioial conservation techniques is at present often too low to cope with the rate-OU environmental degradation. Stone lines have been constructed without any regard for contours; planing pits have been dug, and the excavated earth has been put upslope rather than downslope of the pits preventing runoff entering. Traditional techniques can be improved by introducing simple -survey techniques-and by providing minimum technical training to the farmers. 35 How Successful Is the Kelta Project? The Keita Project in Niger is a most interesting SWC project and is deliberately selected here because it is considered by many to be very successful. Apparently, there are people in the Bank and in other agencies who would favor funding of similar type projects elsewhere-in Niger or in other Sahel countries. There is no doubt that the Keita Project is highly innovative. It has introduced new conservation techniques in the region and has demonstrated flexibility by constantly adjusting and improving its technical package. The project has introduced water harvesting techniques to rehabilitate degraded plateaus using trapezoidal bunds, and it closely associated tree planting with bund construction. The project has shown flexibility by adjusting its techniques systematically. At first, long earthen bunds partially covered with stones were constructed to rehabilitate degraded land. - -'To reduce the substantial damage due to lateral flow and outside runoff, the length of the bunds was reduced to 100 m. with the tips of the bunds extending upward and the space between thd bunds treated with stone spillways. The bunds are now entirely covered with stones to reduce maintenance requirements to a minimum. Why Is the Future for Kelta-type Projects Gloomy? a. The design of earthen contour bunds is too complicated for villagers who cannot handle design parameters such as rainfall intensity, runoff coefficients, size of catchment area, etc. b. The quantities of earth and stones that have to be moved to treat one hectare are bigger than what farmers are able and willing to do voluntarily. It is no coincidence that not even one meter of bund has been constructed by farmers on their own fields. c. Similar conservation works elsewhere in the Ader Doutchi Maggia (Operation Maggia 1963- 1971; Projet de Developpement Rural Badeguicheri 1972-1982) have never been adequately maintained by the beneficiaries, and there is no reason why the people in the Keita region will suddenly maintain the works constructed by the Keita project. d. The cost of the package is about US$1000 per hectare, which makes their large-scale application unlikely. Treatment of 10,000 hectares would cqst USS 100 million. How Successful Is Stone Contour Banding In Burklna Faso? In the Yatengsa region, where thousands of farmers have been trained in the use of water tube levels and in basic-principles of stone band construction, the area treated annually increased strongly between 1982 and 1988. The technique is simple, and farmers have developed considerable skills. There is no doubt that even if all conservation projects would withdraw from the region, stone contour bunding would be continued by the farmers. Replicability in practicel A major question is: will the area treated in the Yatenga continue to increase as strongly as it has done over the last six years? The answer is most likely not. In many villages, stones now have to be transported over longer distances. Unless farmers are provided with donkey carts or wheelbarrows to facilitate transport of stones, the area treated annually cannot continue-to4frow so strongly, Although numerous SWC projects on the Plateau Central promote stone contour bunding, there is nothing that resembles a systematic support program for the farmers. Another possibility would be to reduce transport requirements by creating stone quarries close to each village. Dynamite couldt be sed to create them. There is considerable euphoria concerning stone contour bunding. It may be useful to remember that the adopters of the technique are still in a minority (less than 10% of the farmers). Despite numerous examples of perceptible yield increases on treated lands and the successful 36 rehabilitation of degraded land, the majority has not moved yet. It is surprising but, so far, nobody has undertaken research on the socio-economic characteristics of adopters and non-adopters. Finally, it should be mentioned that thousands of farmers on the Plateau Central have, during the last 5-10 years, spontaneously constructed stone lines and stone bunds on some of their fields. They have not received any training, and the quality of their work leaves much to be desired. They could be the target for a more systematic expansion of stone contour bunding in other parts of the Plateau Central. Steps to Initiate and Accelerate SWC Projects 1. Make an inventory of what has already been done in the field of SWC in a country or region and analyze why traditional and modern conservation works have or have not been maintained and expanded by the beneficiaries. 2. Practice shows that the need for land improvement is most strongly perceived by people in areas with high population density and strong environmental degradation. 3. In areas with stones, concentrate on stone lines and stone bunds to treat cultivated areas, as well as to rehabilitate degraded lands. Build small check dams and rock dams to control and rehabilitate gullies. Use dynamite to open quarries close to villages to ease access to stones and to decrease their transport costs. 4. In areas without stones* - test earthen contour bunds with spillways; - experiment with vegetative barriers (direct seeding of Andropogon gayanus). S. Set up a multi-tiered training program: - for decision-makers; - for extension agents; - for men and women at village level, and accessible to all villagers interested in acquiring skills in surveying of contours, bund construction, raising of seedlings, management of available equipment, etc. 6. Design an ambitious support program which enables villagers to carry out conservation work when and where they want to do so (pick axes, wheelbarrows, donkey carts, etc.). 7. Provide transport equipment and technical inputs for works which cannot be done by villagers with available tools. Example: rock dams. 37 Success depends on: - whether people feel the need to Improve their land; - the selection of the right technical package; - a major effort in extension and training (including use of mass media); - an adequate long-term (10-20 years) support program; - a minimum harmony in modalities of implementation between donor agencies; - an adequate sequencing of development and conservation activities based on local priorities. Final Remark: What Is the Potential for Vetiver Grass I Semi-Arid West Africa The potential for vetiver grass in semi-arid West Africa seems limited. First, vetiver grass needs to be planted in the beginning of the rainy season, when all available labor is used for planting and weeding. Second, direct seeding of vetiver grass Is not possible. Third, because of the scarcity of irrigation possibilities and direct competition with other Irrigated Crops, the introduatiln of irrigated vetiver grass nurseries is hard to imagine In semi-arid West Africa. Fourth, a ptrennial grass like Andropogon gayanus Is well known; its economic value is appreciated and direct seeding is possible. Literature: IFA (1986). Soil and Water Conservation in Sub-Saharan Africa: Issues and Options. Prepared by the Centre for Development Cooperation Services, Free University, Amsterdam. Reij, C. (1988). The Present State of Soil and Water Conservation in the Sahel. Paper prepared for Club du Sahel/CILSS Regional Workshop "Gestion des Terroirs Villageois" (Segou, Mali, May 21- 27, 1989). 5.2 Dryland Management: Water and Sol Fertilty bsues by Raymond E. Meyer Current Status While the status of dryland systems may vary worldwide, dryland systems are generally low in productivity and susceptible to degradation of the resource base. Soil fertility and soil water issues play a dominant role in both productivity and degradation. Many of the soils are inherently low in plant nutrients, and current farming systems are essentially a mining of this fertility capital. Current farming practices are also a major factor in the decline of soil organic matter and resulting susceptibility to erosion, whether wind or water. The decline in structural stability of the soil and its fertility status, due to current practices, begn a vicious cycle of less favorable water/fertility relationships, resulting in less output -- resulting in less management investment and higher risk - - resulting in less output. This cycle must be broken. There are some very productive dryland agricultural systems that can provide examples of principles that must be followed. Needs A detailed understanding and knowledge of dryland agricultural systems in general, and the specific one under consideration, are essential for obtaining solutions for improving the productivity and stability of the system. Effective decisions for investment in development also require solid understanding of the farm-level economic environment. The tone of dismay or fatalism about semi- arid regions needs to change. Since the Green Revolution successes, the central focus of agricultural technology development has been on breeding. In the frequently harsh environments of semi-arid agriculture, it is probably not realistic to ask breeders to develop new cultivars for a multitude of different conditions involving both water and plant nutrient deficiencies and stresses. If moderate improvements can be made in the agronomic conditions first, i.e. more water available during the crop season and slightly higher soil fertility, then breeders can concentrate on yields, quality characteristics, disease, and insect resistances. A moderately improved agronomic environment would not only lower stress levels but also reduce the differences between sites, enabling new cultivars to have wider ranges. National policies must recognize that contributions of dryland agriculture are vital to the national economy and to the well-being of the people living in the dryland regions. National tax, currency, and wage/price policies must treat the dryland agricultural regions fairly. In extremely risky environments, dryland farmers cannot assume all the risks of production if access to stable markets, credit, storage, and purchase of food are not assured. New policy alternatives must be explored to avoid cereal price collapses in good rainfall years, such as 1986 in Africa. Developed countries recognize the importance to farmers and researchers of not letting product prices collapse. Developing countries also must prevent the disincentives from price collapses of cereals in good weather years. Because of climatic and yield extremes, individuals and governments cannot count on a given production figure for the coming season. Consequently, managing dryland agriculture is, in reality, adopting procedures to cope with the dry, variable climate, including dealing with the economic risks of possible crop failure one year and bumper harvests the next. In either case, farm income may be devastatingly low. 39 40 Climatic Resource While climatic or agroecological zoning is useful and even necessary at some levels of planning, they are not always useful at the farm level and may be detrimental to achieving success. Climate must be considered as a resource rather than just a constraint. The climate is the primary determinant of water management and must be understood from a historical basis as well as a real time basis. Dominant features of rainfall in dryland regions are: its limited amount, temporal and spatial variability, and unpredictability. High rainfall years raise the long-term mean more than low rainfall years lower it. Consequently, there are more years below the mean than above the mean, with the degree of skewness inversely related to amount of rainfall. This fact makes it imperative that more emphasis be placed on probabilistic analysis, with less attention given to mean values for designing crop calendars. To match a crop to an area, the length of the growing season required for the crop must be matched to periods with favorable temperature for crop growth and a "reasonable" probability of adequate rainfall to produce a yield. Rainfall distribution in space is also irregular and, as a result, neighboring localities can have very different fortunes in the same year. Finally, rainfall intensity is extremely variable, and high intensity events, even when the amount is relatively low, can result in substantial runoff and soil erosion; so a crop management system must protect the soil resource. Soil Resources The majority of dryland soils have serious problems of texture, depth, fertility, or susceptibility to erosion. Limited water-retention characteristics of many dryland soils are a major constraint, as is surface crusting which increases runoff. Increased soil organic matter could alleviate the severity of all the major controllable soil problems, but most practices used in dryland areas tend to reduce organic matter levels. Devising practical and economic means to utilize soil resources in a sustainable way is a great challenge for researchers, farmers, and policy makers in dryland areas. It is difficult to withstand the pressure to overdevelop and overuse the resource base -- which can make the soil problems more serious. Only by knowing the soil characteristics, potential productivity, and the extent or frequency of distribution of the more important soils can reasonable estimates be made of country production possibilities. Economic Relationships Development of improved systems for dryland farming is limited by both biological and economic factors. The relation of the biological yield distributions and the decision-maker's risk- bearing ability are crucial. The imperfect integration of biological and economic information results in biological researchers not being guided by a clear understanding of economic constraints and the economic analysis not accurately reflecting the biological facts. Macroeconomic decisions are frequently made with unrealistic information from the farm level. Capability to manage both agronomic and economic data at the farm level with eventual capacity to aggregate to national levels is essential for valid and effective policy decisions. Because each farmer has a unique situation concerning his economy, capital improvements, crop/non-crop acreage, animal units, water supply/quality, etc., it is impossible to prescribe a single strategy that will fit every situation. However, there are degrees of similarities, and it becomes important to find those common parameters that can be used for stratifying and providing a framework for analysis. The tradeoff between current production and the future productivity of the soil resource characterizes the economic nature of the problem. Analytical tools and methodologies must be developed to understand the current status and identify what changes are desirable and possible within the constraints of the farm system resources. The development of whole farm models and economic analysis of technological inputs for improved investment decisions in dryland systems can provide an analytical process to improve investment decisions at both the farm and aggregated levels. 41 Possible laterventions There is enough available knowledge to indicate that improved productivity of the soil resource in dryland systems will require the addition of plant nutrients to improve soil fertility, and practices to improve the water retention capability of the soil system. Agronomic Innovations to achieve these objectives must precede varietal development. At the very low soil nutrient levels in much of Sub-Saharan Africa, inorganic and organic fertilizer are complements, not substitutes. There are no proven low cost substitutes for chemical fertilizer. Potential complements such as manure, residues, crop rotation, local rock phosphates, nitrogen fixation, and mycorrhizae are worthy of continuing research; however, for Africa over the next decade, cereal yielding gains are now possible and profitable from water retention and chemical fertilizer. The productivity level of the soil resource must be improved for many dryland areas before input-minimizing practices can be used effectively. Specific observations include: - Chemical fertilizers should be used where available at a cost commensurate with the need for production in the area, equity considerations, and specific nutrient constraints such as phosphorus. - Broadcasting of fertilizer, which is the most common method of application in many regions, may lead to inefficient use of expensive fertilizer inputs. New technologies such as placement between paired rows to "hide" the fertilizer from access by weeds and to decrease the nutrient immobilization, in the case of phosphorus, could have significant effects on increasing fertilizer application efficiency, decreasing weed competition, controlling costs, and improving the crop response function. A new technology of an inverted T for seed and fertilizer placement could have major implications. - In certain phosphorus deficient soils, one-time application of a large dose of phosphorus fertilizer has been found to be more effective than three time split applied in small doses in producing total grain yields over three years. - Where water infiltration is a problem, due to soil degradation, stone and mud dikes made on the contours are effective -- particularly if combined with animal manure. - Water retention devices such as tied ridging or improved land preparation can be used if labor is not constraining or if animal traction is available. - Biological fixation of nitrogen should be used whenever feasible as a means of improving soil fertility and reducing the need for chemical fertilizers. Incorporation of legume crops into cropping systems with the proper rhizobium and adequate phosphorus is extremely important, but may be difficult in some farming systems. - Because of the positive interactions, agronomic practices are more effective when two or more are used in concert. Added fertilizers will not be wholly effective without improved water conservation, and even then they will usually not be beneficial without weed control, increased plant populations, or if proper nutrient balances are not maintained. - Despite its potential, -package" adoption is not evident in the typical sequential adoption pattern of subsistence farmers. A farmer with limited resources will find it difficult and risky to simultaneously adopt several new techniques that require a shifting of household resources. Moreover, learning a new practice thoroughly may extend over several seasons and have uncertain future payoffs. For example, if a farmer has the perception that increased planting density increases his risk in a dry year, he will not have sufficient plant density to receive all the benefit from fertilizer or water retention practices, and thus does not see an adequate return. F9r these reasons, technology adoption often proceeds slowly despite the potential benefits demonstrated at a research level. 42 - More information is needed on constraints of risk management and the priorities of farmers rather than the "lack of money" per so. A farmer in western Niger views the use of fertilizer without certain rainfall to be an excessive risk. He has to pay loans, tax, and social obligations after harvest when grain supplies result in lowest prices and, therefore, returns to investments in fertilizers are often marginally profitable, if at all. He will, however, pay cash for manure, i.e., nomadic cattle on his land. There is a need for more insight into the decision-making of farmers and the criteria used to make decisions and what risks are acceptable. Feedback between farmers, extension, and research is needed. Improved analysis of historical climatic data can allow for better crop calendars that more fully use the climatic resource. A certain level of predictability of seasonal rainfall by time of on-set will allow for adjustment of agronomic practices to more closely match potential production and reduce risks. Strategy A strategy to address the fertility/water problems of dryland agricultural systems must address the following: - how to increase the productivity and stability of cropping systems under dryland conditions; - how to increase the retention and storage of water from natural existing precipitation, thereby minimizing runoff, erosion, and evaporative losses; - how to increase water use efficiency of crops through improved soil, water and crop management practices; - what are the economic impacts of improved rainfed crop and livestock production practices; - what public policy and program options would stimulate adoption and dissemination of improved production practices; and - what are the benefits and costs of additional investments in dryland or rainfed production systems. A scientific strategy for improving dryland production in areas with different levels of capability would involve the linkage of the scientific communities in those countries with common research objectives. This would lead to a better understanding of dryland resources, the production potentials of'those resources, and a sufficient understanding of the economic system to simulate possible production systems. Primary steps would include the development of: - a database of existing research information concerning the soil and water resources and the production systems; - workshops with participants from the host country, international research organizations, and both developed and developing country scientists to outline priorities for research activities; and - research activities conducted in developi g and developed countries with common research objectives and methodology for improved understanding and extrapolation of results to areas of need. 43 The scientist-to-scientist linkages developed by these steps would provide a powerful network of researchers resulting in a synergistic increase of research information and utilization. Doing any one of these steps without committing the financial resources to complete all three provides only a limited impact on dryland development. Certainly, science and technology alone will not result in the elimination of poverty, environmental degradation, or inadequate food production. Dryland recovery will also depend on political and financial commitment by donors and governments. Equally essential are infrastructure and institutions for the provision of basic education, extension and technical services, research, and appropriate governmental policies. Ultimately we may find there is more plasticity in the sociocultural and economic aspects of agricultural systems than in their biophysical aspects. Operational Implications There is simply too much pessimism about agricultural development in dryland areas. Much of this is the result of inadequate understanding of the scientific basis for dryland agricultural systems. The following are offered to assist in alleviating this lack of an informational base for decision-making: - More support should be given to develop a database of the current status of land resources in these areas. This would include land resource inventories linked to remotely sensed base maps and a geographic information system. Preferably this would be done within country and involve training people and developing national institutions to maintain for monitoring purposes. Soil surveys at appropriate scale should be provided before developing new areas. - Since the water resources in these areas are primarily the climatic resources available, more attention should be given to collecting, analyzing and using climatic information for agriculture. Again this should be done within country with the training of personnel to maintain the system over the long term. - More emphasis should be given to the development of analytical tools and methodologies for evaluating costs and implications of interventions, opportunity costs of taking no action, and for providing alternative scenarios at farm and aggregated levels. - Increased support must be given to research in soil and water resource management for productive and sustainable utilization of the dryland resource base. National research and extension institutions must be strengthened. - Project preparation studies should include input of people with the scientific knowledge and dryland experience to understand the system. In-house expertise is needed to assist in designing, managing and evaluating projects. - Seminars/conferences/workshops should be organized every 18 to 24 months in different dryland areas within the developing countries with the explicit purpose of improving communication among developing country scientists and developed country scientists working on similar dryland problems. - Increased resources should be allocated to the training of people and to the development of technical institutions in dryland areas. Farmer education and training is essential. Achieving long-term sustained growth in the productive capacity of dryland areas will require sound decisions based on accurate assessments of resource problems and potentials, and on careful analysis of alternative policies, programs and projects. Hence, the capacity of governments and donor agencies to make the necessary technical evaluations and rigorous analyses must be improved if thfe desired changes are to occur. PART V. POUCY ISSUES FOR IMPROVING LAND MANAGEMET 6.1 Why and How Do Land Users Make Lamd Management DaeIons? How Can Poldes Encourage Land Users to Make Them Ffectively? by Piers Blakie Overseas Development Group University of East Anglia Norwich NR4 TJ England Almost all land is capable of being managed to proviqe sustainable yields indefinitely FAO list over 30 distinct groups of soil and water (SWC) conservation techniques, and there are usually countless regional sub-types, at least some of which are familiar to land users. Yet in many places, land is being degraded. Any policy designed to reduc land degradation (LD) must be based on a correct ad comprehensive explanation of how land users make, or fall to make, these decisions. There are five basic steps in policy formulation for SWC. These are (i) identification and measurement of the physical processes of degradation; (ii) the identification of those for whom land degradation is or will be a problem; (liI) an explapation of why LD is taking place and the identification of which of the technical and social determinants are amenable to change through policy; (iv) a review of thq ability of government and local institutions to change land users' pradtice; and (v) the formulation of a policy which matches (iII) and (iv). None ofl these steps are straightforward and misidentifications at any point in the policy-making process can lead to inappropriate or uninplementable policies. This paper focuses on (iiI) and (iv). An explanation of why LD is taking place can be made at different levels of abstraction using a variety of disciplines. The attached diagram and notes are abbreviated from a forthcoming article in Land Degradation and Rehabilitation 1:1, they outline a way of approaching an analysis of any particular case of LD. Readers should scan Annexure I first and then return to this text. Having Mead Annexure 1, the reader may feel the explanatory scheme starts with a specific instance of Li and ends by trying to explain the whole world, and that such a project is impossible and, in any case, unnecessary. My reply would be that it should be used as an interpreted checklist of factors which may be relevant in formulating a realistic SWC policy. A more conventional approach concentrates on th site-specific ang technical aspects of SWC and ignores the structural reasons for land degradation. A review of 17 FAf projects which included a substantial element of $WC were reviewed by the.author (reprinted in Blaikie 1985 p61f). Almost all had failed to reach their central goal of reducing soil erosion (although peripheral goals such as mapping, buildings, training, etc., were more , often attained). Clearly, one of the reasons was that the wider circumstances in which LD takes place and the capabilities of the administration were not specifically considered. 45, 46 One of the most important elements in the chain of explanation concerns the ways In which land users adapt to the consequences of LD. It received a couple of paragraphs in Annexure I and is pursued in more detail in Annexure 2, which outlines a decision-path for land users. Basically, it is a perception-diagnosis-action sequence where, unfortunately for policy-makers, there are plenty of opportunities for land users to stray from the path leading to effective changes in land use practice. Summarizing the policy implications for this decision model a. The land users' perception of whether land Is degrading or not and whether it matters may differ from the official one. For example, bush encroachment in the Communal Areas may not necessarily be viewed by Zimbabwean farmers as degradation since the increased browse may provide food of the last resort in a drought when grass has long since disappeared. The cases of the 11 Chamus in Annexure 2 is another case in point. b. There are numerous cases where expert conventional wisdom has failed to make an impact on land users because it was palpably wrong. The reappraisal of "what is wrong with the Himalaya" at the Mohonk Conference and summarized in a special issue of Mountain Research and Development (July 1987) is a cautionary tale of those experts who were convinced of anthropogenic erosion on a massive scale. The problem is now seen to be much more complicated than previously thought and geomorphic/orogenic processes are now given much more weight. What follows from this is a new direction for SWC away from the prevention of soil erosion and towards nutrient conservation and fuelwood provision. In Zimbabwe, careful historical work by such workers as Wilson and collaborates (1986) throw some doubt on the long established official view of inexorable increases in stocking densities in some areas, and there are numerous other examples in drylands. c. There are many instances where economic analysis of degradation and SWC takes land as its object of analysis and calculates the enhanced income streams and costs of SWC over time. However, the land user may calculate SWC decisions in the context of the household budget as a whole, not just per ha. of land. The impact of non-agricultural income is most important here (viz., the case of Lesotho and Botswana in Annexure 1). d. Policy must be matched to the capabilities of government and other complementary institutions. In Sudan, it is clear that SWC can only be encouraged using local leadership and local inputs, perhaps involving the provision of a "menu! of SWC activities appropriate to a particular area, but leaving choice up to the capabilities of individuals. Figures I and 2 in Annexure I provide a general approach to understanding LD and the role of government in increasing/reducing it, but there are sometimes shortcuts to this approach. This is simply to identify a role model -- an institution in governinent, a project, a village council, or even a category of land users -- which has been able to reverse LD. The World Bank's use of such projects as the Nepal-Australia Forestry Project as a *stalking horse" is an example. Systematic research of and learning from other agencies' experiences are useful empirical back- ups to the more analytical approach in Annexures I and 2. 41 . e. The issue of land tenure and its interface with the contrOJ and access to land, labor and capital within the household are usually impdrtant structural issues in SWC policy. It is ikot Just a matter of official land law, lMportant though this is, but also the operation 6f customary law and parallel systems controlling rights to land. The problem is that sa outside aency can only addresstself to official land law, not to its local implementation. TIAe Sudanese case outlined in Annexure I is a problematic example. SWC in commoproperty management systems, is a more-general expression of the official/local interface between paper and practice, government and people. Agencies such as the Bank may be able to advise on the former, but have little actual, leverage "to modify the latter even in the medium term. There seems to be no substitute for local empirical research into how official and customary law in land questions actually operates, and to be prepared to adapt SWC policy to these local variations. o- PART VIL OFF-FARM EARNINGS AD MIGRATION ISSUES 7.1 Income Diverahilictin of Rural Households In BurMus Fase by Thomas Reardo International Food Policy Research Institute, Washington, DC and Christopher Delgado' International Food Policy Research Institute Washington, DC In debating the development and sustainability of Sahelan agriculture, rural households are often treated solely as farmers. Price incentives' to produce more cereals and to invest in sustainability are thus seen as automatic for economically rational households. Yet Sahelian households do not always behave as policy makers anticipate they will. In Burkina Faso, farm-leveU investmenVs in productivity and sustainability of cereals production have not appreciably increased following a sustained upward trend in real cereals prices. Nevertheless, households are feeding themselves, even during drought years, and even in z6nes where one would expect that hunger would abound, such as in the Sahelian zone. The reason is that rural households purchase substantial portions of their food (Reardon, Delgado, and Matlon, 1987). These purchases are financed by a diversified Income base Sahellan rural households are not merely farmers. In this brief, evidence from Burkina Paso shows that most income springs from non-cropping sources, in both the relatively arid northwest and the relatively well-watered sou%ltvest of Durkina Faso. The results come from IFPRI/ICRISAT collaborative work using data from the ICRISAT baseline survey conducted by Peter i6iion. It covered four harvest-years -- 1981/82 to 1984/85 (a variety orgood and poor harvests) and was conducted in three zones of the country: a) the Sahelian zone in the northwest (agroclimatically, a very poor zone, with low rainfall, poor soils, extremely viriable cropping outcomes); b) the Sudanian zone in the Mossi Plateau (a poor to intermediate zone, with low-pedium rainfalL poor soils, moderately variable cropping outcomes); c) the Guinean zone iithe southwelt (a moderately favored zone, i.e. medium-high rainfall, good soils, relatively stable cropping outcomes). tlon (1988) gives details. 'The authors are grateful to Garry Christ6nsen and two anonymous journal reviewers for their comments. They also gratefully a9knovledge the excellent research assistance of Peter Mation. The views expressed are solely the authors. 49 50 Table I shows levels and composition of household income using 1981-85 averages. There are sever.l striking points concerning income levels: Rural households are not always worse off than urban households: The richest 1/3 in the Guinean zone had household income per adult equivalent of 94,000 FCFA (- about $185 per capita); the richest 1/3 in the Sahelian zone had income per capita of about $138 per capita. These figures are greater than the average income of the poor tercile in Ouagadougou in 1984/85 -- about $135 per capita (Reardon, Delgado, Thiombiano, 1989). Household Income Is not closely related to the agroclimatic level of the zone: The worst and the best zones have the highest incomes: the Sahel average income is 42,000 FCFA/AE, while that of the Guinean zone is 55,000 FCFA/AE. The Mossi Plateau's average is only 29,000 FCFA/AE. Despite greater purchasing power, more food aid was targeted to the Sahelian zone than to the Mossi Plateau in the 1984/85 drought year. The targeting was based on crop production outcomes, not purchasing power (Reardon and Matlon, 1989; Reardon, Matlon, and Delgado, 1988). Income Is not evenly distributed* The gini coefficient of income per adult equivalent is about .3 in the zones. Hence income distribution is unequal despite relatively easy access to land. Although less unequal than South Asia, this is fairly representative of rural West Africa. This suggests the need to look for sources of income other than cropping which differentiate households. There are several striking points concerning income composition: Non-cropping Income is very Important It comprises 52% of income in the Sahel zone and 57% in the Guinean zone; however it comprises only 26% of income in the Sudanian zone. In general, the higher the income, the lower the share of cropping In total household Income: The poorest zone (the Sudanian) relies the most on cropping. The poorest terciles in all zones rely the most on cropping. Reardon, Delgado and Mation (forthcoming) show for the same period that hunger increases as incomes decrease in these areas. Hence those most dependent on cropping alone -- with least diversified incomes -- are the hungriest. There is no necessary link between own production and food security at the household level in the Sahel. Local non-farm activities (Intersectoral lInkages) are most important In the best agricultural zone -- the Guinean: Local non-farm income (food processing, cottage industry, commerce, etc.) is highest in the Guinean zone -- 38% of income. These activities are closely tied to local agriculture. This substantially boosts the income of women relative to those in other zones (Gabre-Madhin and Reardon, 1989) and demonstrates the potential for "intersectoral growth linkages (Mellor and Johnston 1984). Households in the other two zones also have local non-farm activities but they are less important, and the poor participate in them less than in the Guinean zone. By contrast, the Sahellan zone's Income diversification, Is more externally-based 11% of income derives from non-local or migration sources versus 3% in the Sudanian and only 1% in the Guinean zones. The potential for intersectoral linkages is reduced by the weaker and more variable agricultural base. Thus, in the Guinean zone, the demand base for income diversification is local; it is much more external in the Sahel. : Off-farm Income Is being used to compensate variation In cropping outcomes: Table 2 shows the coefficient of variation of four yearly averages of cropping and total income of households per zone. Cropping income is 2.6 times more variable than total income in the Sahelian zone, 1.7 times in the Sudanian, and only 1.4 times in the Guinean. Hence, the ratio of the variability in cropping' income to total income decreases as the agroclimatic level increases. The riskier the cropping enterprise, the greater the benefit to income diversification. S1 These results raise a host of policy and research issues that must be pursued to better inform policy debate. What are the consequences of rural household Income diversification for sustainability? Is the Sahelian household's keen interest In income diversification, which requires time and cash investment, at odds with its willingness or capacity to invest in the sustainability of the agricultural base, e.g. construct bunds? (Reardon and Islam, 1989) What are the consequences for improved technology to Increase productivity? Do Sahellan households want to plow their non-cropping Income back into cropping -- to buy animal traction equipment, fertilizer, etc.? Or, do the more stable and possibly higher returns in the non-cropping sectors attract and divert their cash? Is this a key factbr in holding-up the modernization of Sahelian agriculture? What Impact do Sahellan household-lcome strategies have on the evolution of trade patterns -- between the Sahel and the world, and between the Sahel and the coastal countries? In the context of current debates, is risk aversion a major factor in consumption shifts towards imported rice? Towards Ghanaian corn? (Reardon, 1989) How can policy best use Intersectoral growth linkages to raise rural Incomes? What are the relative roles of infrastructure and technology investments? Many of these issues have a direct impact on the single most important long-run agricultural policy issue in the region: What commodity priorities would get agriculture moving? 52 References Gabre-Madhin, E.Z., and Reardon, T. "The Iaportance and Determinants of Women's Non-Agricul- tural Income Strategies in Rural Burkina Faso", Select Paper, AAEA 1989 Meetings; mimeo March 1989. Mation, P.J. 198. The ICRISAT Burkina Fas Farm-Level Studies: Survey Methods and Data EUM, ICRISAT, August 1988. Mellor, John W., and Bruce F. Johnston. "The World Food Equation: Interrelations among Development, Employment, and Food Consumption," Journal of Economic Literature, Vol. 22 (June 1984), pp. 524-5.1. Reardon, T. "Cereal Demand in West Africa: Implications for Sahelian Regional Protection", in the Proceedings of the CILSS/Club du Sahel Seminar on "Regional Cereals Markets in West Africa', Lom6, Togo, November 6-11, 1989. Reardon, T., Delgado, C., and Matlon, P.M., Coarse Grains in Household Income Strateaies and Food Security in Burkina Faso Draft IFPRI Research Report, forthcoming. Reardon, T., Delgado, C., and Mation, P. "Farmer Marketing and the Composition of Cereals Consumption in Burkina Faso", mimeo from IFPRI/ISRA conference, Dakar, July 1987. Reardon, T., Delgado, C., and Thiombiano, T. Substitution by Urban Sabelian Consumers Between Coarse Grains and Imnorted Rice and Wheat: The Case of Ouaadonenn Submission to Publication Review Committee, IFPRI research report series, November 1989. Reardon, T., and Islam, N. "Issues of Sustainability in Agricultural Research in Africa", in the Proceedings of the Symposium on the Sustainability of Production Systems in Sub-Saharan Africa, at NORAGRIC, Agricultural University of Norway, As, Norway, 4-7 September 1989. Reardon, T., and Matlon, P. "Seasonal Food Insecurity and Vulnerability in Drought-Affected Regions of Burkina Faso" in D. Sahn (ed.) Seasonal Variability in Third Word Amdculture: The Conseuences for Food urity, Johns Hopkins Press, 1989. Reardon, T., Matlon, P, and C. Delgado. "Coping with Household-level Food Insecurity in Drought- Affected Areas of Burkina Faso, World Develonment. Vol 16, No.9, pp. 1065-1074, 1988. TABLE I INCOME SOURCES BY ZONE AND INCOME STRATUM 1981-5 AVERAGES (In percent of total income) crop Livestock Locul Non-tocal toed Oher Total ncome Ime Noa-forma Non-farm Aid Tranfms Tncom lncom. Income.i CIA/AB SARELIAN Poorest 1/3 66 17 6 5 3 4 22,566 Richest 1/3 36 J7 35 9 1 1 70,545 Overall 48 15 24 11 2 1 42,205 N-45 SUDANIAN Poorest 1/3 82 5 8 2 0 3 15,660 Richest 1/3 69 6 16 5 0 3 48,616 Overall 74 6 14 -3 0 3 29,295 Nu44 GUINEAN Poorest 1/3 57 18 22 2 0 3 29,886 Richest 1/3 31 19 48 1 0 01 95,629 Overall 43 17 38 1 0 1 55,261 N-47 Source Reardon, Delgado, and Matlon (forthcoming) 54 TABLE 2 - INTER-YEAR VARIATION FOR 1981-85 IN CROPPING INCOME AND TOTAL INCOME -- AVERAGE OVER HOUSEHOLDS Coefficient of variation (CV) of yearly averages over households Cropping Total Variation of Cropping Income Income Income Variation of Total Income SAHELIAN .64 .37 1.73 - SUDANIAN .48 .37 1.30 GUINEAN .50 .32 1.56 Source: Reardon, Delgado, and Mation (forthcoming) 7.2 Temporary and Permanent Migration in West Africa Some Early Findings from the Onchocerdasi-Freed Areas Study: What Support Should the Bank Give to Migration?* by David Brokensha Institute for Development Anthropology 1. Heterogenelty The areas that attract migrants are characterized by diversity -- of physical environment (rainfall, soils, vegetation); access to markets; infrastructure; 'commitment to development" (Arhin); production systems; ethnic composition; types and patterns of settlement; population density. This means that few generalizations can be made and that site-specific approaches are necessary. 2. Institutions What support should be given will depend to a large extent on the types and efficacy of national and local institutions. Contrast, for example, the AVV in Burkina Faso and the National Onchocerciasis Secretariat in Ghana. The Bank should, where practicable, work with existing "line agencies" rather than encourage new organizations. At the local level, political party branches may be the best way of approaching local people, as in Mali; or non-government organizations, such as the Catholic and other churches in Northern and Upper Regions of Ghana, may be effective in reaching the people. 3. Regional Integration Attempts to promote regional integration should be encouraged. 4. Local Production Systems Existing production systems (which include not only crop production but also livestock management, fisheries, and off-farm activities) should be built on and not destroyed. It should be recognized that many people will continue to depend on traditional forms of agriculture, and also that wage labor will continue. 5. Intensive Agriculture Despite the fact that many will practice traditional agriculture, opportunities to encourage intensive agriculture should be sought. This includes promotion of chemical fertilizers, insecticides, water retention and improved manure, all of which can help to stabilize yields and eliminate the mining of the soil. *Note: Many of our case study sites are in areas south of the 806 mm isohyet, which may have introduced a bias in our preliminary findings. 55 56 6. Eavronmental Protection Because of the dangers of short term mining of the soil, emphasis should be on environmental management, which means stressing the need for long-term (20 years) views with emphasis on sustainable development of resources. 7. Irrigation Generally, large-scale dams should be discouraged and investigations should be made of the prospects for small dams, various types of pumps, and also simple water retention projects such as dikes, tied ridges, and stone bunds. 8. Wet Lands Because of the importance of flooded wet lands as a resource, the development potential of these should be realized. 9. Non-tarm Opportunities Wage labor in the migration area or in the surrounding area, increased fishery activity, and other non-farm employment should be encouraged. 10. Infrastructure One of the most needed interventions is in this area -- in improving communications and in establishing an effective and competitive marketing system. For example, settlers are leaving Manantali, Mali, a resource-rich area, for poorer areas in Senegal because of poor roads and monopsonistic markets). 11. Extension In order to promote intensive agriculture, an effective extension service is essential. 12. Donor Coordination The Bank should continue to take the lead in coordinating activities of multilateral and bilateral donors. 13. Research What is needed are very specific and development studies of what works and what does not work in terms of changes in agricultural practices, technology, and other interventions in site- specific approaches. 14. African Institutions Most West African countries now have a core of skilled professionals at university and research institutes whose skills should be used as much as possible. 15. OCP (Onchocercasis Control Programme) Despite the dramatic success of OCP in controlling the disease and in reducing human misery, it is not the most appropriate vehicle for promotion of socio-economic development in the ancho- free zones, and the Bank should actively inquire into alternative approaches. PART VIII. PARTICIPATORY ISSUES 8.1 The Aga Khan Foundation's Approach to Participation and Its Relevance to Bank Projects An India Drylands Example by Andrew Weir Programme Officer, Rural Development These notes outline AKF's approach to participatory development through the Aga Khan Rural Support Programme (AKRSP) and describe its adaptation -- from its 1982 initiation in the northern areas of Pakistan to the situation in Gujarat, India, where a program was started in 1984. The report then discusses examples of how AKRSP has adapted government programs in the drylands of Gujarat and reflects on the lessons learned from implementing the AKRSP approach in Gujarat. 1. The AKRSF Approach Origins: The approach is rooted in the European rural cooperative movements of the 19th century. The methodology involves the establishment of a partnership between program staff and villagers through collective village management, savings, and improvement of human skills. AKRSP aims to establish effective village organizations which can find and demand services from outside. The program's primary contribution is the working village organization. Explicit in the approach is the need to link these organizations with external resources provided by government, other NGOs, and the private sector. Ultimately AKRSP intends to work itself out of a job -- usually within ten years in a particular area. The approach is still very much at an experimental stage. 2. Comparisons and Contrasts The AKRSP approach is now being applied in the very different environments of the north of Pakistan and the state of Gjarat in India. Pakistan" Indias Total population 802,000 4,431,000 No. of villages 1,410 2,842 No. of village organizations 993 56 Total savings-(USD) 2,700,000 .80,000 1989 program budget (USD) 5,700,000 1,500,000 'Three districts: Gilgit/Chitral, Baltistan. Three districts in Gujarat Junagadh, Surendranagar, Sharuch. 57 58 The main characteristics of the progratm areas that affect the method of program implementation are. Pakistan: physical isolation; geopolitical importance; an institutional and development vacuum; development potential (land, labor, knowledge); socially homogeneous villages with little economic stratification; most families have agricultural land. India: caste and religious differences and serious socio-economic divisions within villages; landless people; large villages; dependence on government with wide ranging subsidies; exploitation and mismanagement; history of failed or only partially successful development projects. Pakistan: The Original Approach AKRSP staff make a diagnostic survey comprised of three dialogues with a village to (i) explain the objectives and methods of AKRSP, invite the villages to form a village organization, initiate a savings program, and identify a productive physical infrastructure (PPI) project to benefit the village as a whole; (ii) determine the physical and financial feasibility of the PPI; (i) discuss the finalized PPI, explain AKRSP's terms of partnership, the form and extent of its assistance and, if the dialogue is successful, pay the first installment of the PPI grant. AKRSP field staff then maintain contact with the villagers to assist their organizational and institutional development and to provide technical services when and as required. Indla: An Adapted Approach AKRSP staff carry out initial dialogues and resource surveys with villagers to identify an infrastructural project. Beneficiaries agree to form a users group and to save wages earned during the construction of the project. Interventions to benefit other groups within the village are then identified. Models are developed jointly to meet the needs of the different groups. AKRSP aims to build a cluster of activities, each managed by a users group. Staff then work with the groups to establish a village-wide umbrella organization. AKRSP had to create a demand for its services, establishing credibility with the villagers as its first imperative. AKRSP is involved in village management in the early stages, because of group rivalries within villages, but gradually withdraws. The program makes use of government funding sources to expand its impact and increase the effectiveness of donor funds. 3. Examples of Implementing the AKRSP Approach In India A key aspect of AKRSP's work in Gujarat is to enable villagers to take advantage of government programs that are intended to help them, but in practice are only partially effective. By introducing participation, flexibility, technical expertise, and some crucial finance, AKRSP is attempting to improve the effectiveness, scope, and potential of these programs. In this section, four examples are used to analyze the process. These involve the development of degraded public and private land; the renovation of irrigation tanks, the construction of biogas plants, and soil and water conservation and water harvesting. When implemented by the government, these interventions are standardized in design and management, involve potential beneficiaries only as laborers, and suffer from technical shortcomings. By working in partnership, AKRSP and the villagers have adapted and developed the approaches to meet site-specific needs. There are indications that the government is willing to become a participant in this process. Nonetheless, obtaining access to government development funds is a time consuming process. 59 4. Lessons from ARRSP's Experience Although still in the experimental stage, experience with the AKRSP approach highlights the following implications for external interventions. The establishment of functioning local organizations takes considerable time and substantial amounts of money and managerial skills. In addition, there is continuing need to support and supervise the organizations as they develop. The development of local organizations means that the integration of services (e.g., agricultural extension, health, education) can become demand-driven and will occur at the village level. This reduces the need for the cumbersome coordination of individual bureaucracies. Establishing the correct incentives for small farmers requires appropriate pricing and macro- economic policies, including agricultural input-output stabilization. The present subsidy policies in India encourage the inefficient use of resources. Incentives to encourage private sector investment in rural development at both small and large scale would increase opportunities for innovation. There is a tremendous need for information. An effective extension system is evidently crucial. Increasing sophistication of farmers could result in demand-led extension. A much more imaginative approach to extension is necessary, probably involving villagers directly. Effective technical packages for agriculture in semi-arid areas are few and far between. The limited information that is available is not being efficiently disseminated. The involvement of villagers in planning and implementation improves the effectiveness of programs by increasing the information available to planners. This encourages site-specific flexibility and developing local responsibility. Despite their increasing involvement in rural development, NGOs still face constraints to more effective interventions in dryland areas. This suggests a lack of facilities which provide technical support, training, and information to NGOs/ The present tax structure provides no incentives to the private sector to contribute financially to NGOs, whose core cost funding is particularly difficult. I 3.2 Issues of Local Participation In Bank Projects by David Richards Director, NGO Programs World Resources institute The trademark of a successful grassroots NGO leader is his or her ability to organize authentic participation on the whole range of issues and views affecting a community, and to find a place at the table where they are considered, challenged and, one way or another, dealt with. I have seen such social entrepreneurs limit the impact of powerful, local vested interests on community development projects by having other interests get a fair hearing, by bringing them to the table. Referring to the "new professionalism," Robert Chambers of the University of Sussex insists that the only way to establish an authentic participation process is to -- participate. To periodically live with the people who are the objects of our policy reconsideration and our assistance program - -but on their terms, which means in their homes, in their time frames, and in their world views. The World Resources Institute does not implement programs at the grassroots, wherever this curious place might be. But the Institutes new partner, the former North American office of the International Institute for Environment and Development, has spent years observing and studying ordinary peoples' participation in natural resource management projects. Or, more specifically, we have studied the little, local organizations in which people lather to make decisions affecting the whole group. Why have we been so interested? Because these local gathering units are where people organize themselves to handle resource management crises. It is where they change policy, sometimes very local policy, and it is where they take advantage of policy changes which favor the notion of sustainable development. It is the basic unit from which people reach out and touch -- the helpful or hinderinS arm of policy and outside knowledge and resources. If human society and the environmental systems that support life are to accommodate each other, it will be through the collective resource management choices of billions of individuals throughout the world. And these people meet and participate in small groups -- "ht is why they are so important. These local membership organizations, or MOs, are the basic building blocks of the non- governmental (NGO) system. Substantially self-reliant, they are member accountable people organizations which exist to serve their members. MOs increase the capacity of the 'poor to be economically productive and politically active. It is within the traditional and modern village or neighborhood level MOs that voluntary energy is organized and released. Voluntary energy directed towards solving resource management problems is the usually unarticulated goal of governments and aid organizations. Its value is its low cost, at least in financial and political terms -- there is potentially a lot of it and, from humble beginnings, it can become a significant force for desired change. 61 62 Voluntary energy is largely inaccessible to development planners, bureaucrats, and technicians because it is not easily controlled or understood. It doesn't easily lend itself to structure, and projects are heavily structured. The individual choice to participate is not based, unfortunately, in logic. And forced participation has no depth and will not last. It is not authentic. Local MOs are where the most important aspect of effective resource management occurs - - what David Korton, an NGO theorist, calls micro policy reform. Micro-policy reforms are the local institutional changes and creations needed to sustain a desired change in individual or group behavior. Sometimes a new association must be formed, or power relationships must change, or gender relationships must be renegotiated. Most complex of all are micro-policy reforms calling for a sharing of power between the national and local levels, and the creation of self-reliant village organizations. Sustaining change is based entirely on participation in defining the problem, defining the options, refining the solution, and sharing the risks and benefits. Local member organizations are served by more formal NGOs of all types, directly and indirectly. They provide services, assist in community organizing, training, and many other activities. More distant from the action, supporting NGOs take on various functions such as grant making, research and information exchange, networking and advocacy to name a few. It is NGOs with which outsider organizations work, and the NGOs in turn interact with the membership organizations. NGOs are participation managers. They know the names and addresses of poor people, and they have staff who will endure considerable hardship living and working and, most important, participating at the local level. What exactly are NGOs? NGO is a code acronym for "solving the right problem, with the right people, with the right resources, at the right time, in the proper way, to help the poor, while promoting pluralism, on a sustainable basis, inexpensively." "In the proper way" means through broad-based participation which legitimizes initiative as no other device can. The happy result is voluntary energy. If voluntary energy is combined with innovation, a powerful force for sustained change favoring prudent resource use is released. The CHIPK6 -- Hug the Tree -- movement in India is a good example. How does the World Resource Institute become engaged in participation? Through NGOs working with local membership organizations in areas of natural resource management of interest to us -- and we have yet to identify one that is not -- we have a policy and project feedback system. We find out what the unintended consequepces of resource policy are. We foretest new policy under consideration to find out how it will play in Pretoria, or Ouagadougu in this case. And we try tb inform policy formulation by encouraging reference to agro-ecosystem history from the experience of elders and traditional scientists. When undertaking forestry policy reviews, the Institute ensures that NGOs are at the "Table." Sometimes we provide the table so we can invite all the guests. We try to reflect local participation at a higher level. Participation doesn't mean agreement. WRPs role is to stick with the process until, almost mysteriously, whatever accommodation is going to be achieved, is. In our Africa resource policy project, we gather information from traditional practitioners and modern scientists on ecosystem history, looking for insight into policy impact on human interaction patterns with the land and water. NGOs are the nodal points. They introduce us to the places and people, and they critically listen to our conclusions. Usually, very critically. MOs and NGOs are at the core of our little project entitled From the Ground Up in Africa or, as we call it with our passion for acronyms, FGU. FGU douments successful cases of effective local management of critical natural resources. The point is to move beyond an anecdotal or development gossip appreciation of what works. The cases are used in orientation seminars for NGO field workers, government planners, and donors. We work with indigenous NGOs who have 63 a bent for barefoot research methodologies. They in turn work with NGOs and village membership organizations to get at the key issues which framed "success" -- leadership, problem definition, technology, and everything else common sense tells us is a key ingredient to sustaining an activity. Our research approach is an adapted form of Rapid Rural Appraisal developed by Gordon Conway for agro-ecosystem analysis. We tinkered with it again and now have Participatory Rural Analysis or PRA. PRA is a systematic, intuitive approach to the social and natural science of resource management at the level where people have to practice it. The first case we did last year was in the village of Katheka in Kenya. Basically, we wanted to know why women have built 12 kilometers of bench terracing with virtually no outside funding or technical assistance. We want to know how such an effort could take root nearby. We want to know which elements of what happened in Katheka could somehow in some measure find their way into government attempts to do "right things" in natural resource management. The quality of participation is a key success variable in everything we have looked at so far. If the quality of participation is so important, how can it complement the efforts of organizations like the World Bank? The first answer rubs one of those touchy national sovereignty spots. Nonetheless, governments have to permit sufficient political space for local membership organizations and their supporting NGOs to flourish. Some call it pluralism and democratization. In any case, the Bank ought to use whatever levers it can to build NGOs into Bank funded project designs. In fact, this is being done. But one nasty result is that people iaving no skills as authentic participation managers are forming NGOs in order to become recipients of Bar -A other ODA funds. The Bank has a responsibility to separate this cynical opportunism from - ependent sector development. Second, Bank programs must support NGOs which play a catalyst role in developing long term local institutions which can sustain the gains made in Bank projects. This is the micro-policy reform mentioned earlier. NGOs must not just fund a whole bunch of little projects, no matter how successful. There has to be a strategic purpose. Burkina Faso is a good example of how well organized NGOs can be. International organizations have made an extraordinary investment in the country to deal with the effects of drought and famine emergency. Today, there is a critical mass of NGOs operating which appear determined to reverse some of the underlying causes of the country's poverty and resource degradation. The NGOs formed an independeDt umbrella organization to coordinated inter-NGO activity, to deal with donors collectively and to ensure that "people" were represented at the table. It is known as SPONG. The government established a similar body known as BSONG which coordinates between government and NGOs. The creation of BSONG is a direct reflection of government's recognition of the significant role of NGOs in national development. Often, there are warts in the relationship, but the basic success of the Government's decision to open political space was risky for NGOs. Yet the success speaks for itself. Fifteen percent of all development assistance to Burkina Faso flows through NGOs, and dozens of little but very important micro-policy reform projects are underway. PART IX. WRAP-UP COMMENTS by Stein Bie NORAGRIC 1. Several workers have predicted a global food crisis early in the 21st century. This is a good reason for suggesting that the drylands take their share of the required production. 2. In any case, greater self-sufficiency is needed, particularly in the more remote drylands where infrastructure is poor and costly to provide. 3. Drylands are more resilient than we had expected. Dryland degradation can be and has been reversed. 4. Dryland peoples are more resilient than expected. They have developed advanced (but to many of us poorly known) adaptation mechanisms. These have enabled them to survive droughts and temporary degradation. Compared to previous disasters, losses during the recent Sahelian droughts were probably low. 5. The drylands of Africa may have entered into a more favorable climatic period. Statistically, this will probably last for some time -- possibly 10-15 years. Droughts, however, will most likely return. 6. We realize that productivity of the drylands will be low. However, investments will have to consider not just direct returns but also the very high costs of doing nothing. 7. In the short term, productivity in the drylands is difficult to measure. Investments in change must focus on the process of change rather than on the products and thus be long-term oriented. 8. We know for certain that yields in drylands are much more variable than in humid areas. Therefore, survival strategies consider much longer time spans than one season or one year. A reasonable strategy would improve the ability of the dryland peoples to survive and profit from wide fluctuations in production. 9. There are a number of interesting technologies for stopping land degradation and improving land productivity. But their sustainability is often poorly understood by the donor community. Failures to achieve sustainability may reflect incomplete understanding of the ambitions of the beneficiaries -- e.g. not knowing that in many dryland areas women are responsible for crop production and that men are unlikely to accept responsibility for maintaining structures pertaining to cropping. 10. In general, traditional adaptation strategies seem to be underrated. It may be wise to reinforce traditional practices rather than replace them with untried alternatives. 65 66 11. Within range management there have been no major technological breakthroughs in recent years. However, rangeland may still be sufficiently productive to warrant attention. Certain actions may give particularly good returns: - veterinary services - forming pastoral associations - sorting out land tenure/access issues. 12. For crop technology, a number of promising techniques are underway: - water harvesting, particularly catching surface runoff. In some instances the exploitation of ground water resources may be valuable. - phosphate: there is much uncertainty associated with the phosphate status of dryland soils. While there is some prospect that leguminous trees and plants may fix valuable quantities of nitrogen, phosphate may have to be applied to avoid P from becoming a limiting factor. More investigation is required. 13. Off-farm income may be important for utilization of drylands, both rangeland and cropland. One strategy could be to increase the sustainability of production systems by adding or stimulating non-farm activities. 14. Past interventions have often failed to prevent land degradation. In the future, investments should focus more on degradation in addition to returns. 15. It is our experience that large projects are rarely successful and often environmentally detrimental. It is probable that programs with many small undertakings are more promising. NGOs have considerable experience in smaller projects and could be used. 16. In my opinion, the World Bank bas considerable economic expertise on drylands but has little technical and environmental competence. The WB should consolidate its scattered expertise. 17. Increased use of local expertise in lesser developed countries would seem advisable. 18. I had some ideas on World Bank attitudes prior to coming to this meeting: a. The World Bank is too negative on drylands. I have found a much more varied and positive outlook during the last two days. b. The World Bank underrates the value of and wisdom in traditional and successful adaptation. I still believe this is true. c. The World Bank overfocuses on imported technology, mostly non-biological. Although this may still be true, the strategies of vetiver grass banks and the use of antropogon, together with shelter belts and other tree-growing techniques, indicate that the WB is on the move. d. The World Bank overlooks the role of women in the economic ard production systems of the drylands. I believe this to be still true and that it adds to the non-sustainability of some intervpntions. 19. In conclusion, I believe the seminar was also useful and constructive for the external participants. Annex 1: WORKSHOP - AGENDA Professional Deveopnent Workshop on Dryland Management1 Washington, D.C. Workshop Chairman: John Hayward, AGRPS Wednesday, May 10 8:45 Registration 09:00-0915 Opening Remarks: Kenneth Piddington, Director, ENV Session I - Introduction 09:15-09:45 The Extent of the "Desertification" Problem and How to Measure It Prof. Stein Bie, University of Tromsoe/NORAGRIC, Norway 09:45-10:15 How Farmers and Pastoralists Adjust to Land Degradation. How Can Donors Help? Prof. Michael Mortimore, University of Birmingham, U.K. 10:15-10:30 Coffee 10:30-11:00 Discussion on Session 1 Presentations Session 2 - Review of Experiences 11:00-11:30 What Has Been the Experience with Dryland Interventions? Juan Seve and Bruce Ross-Sheriff, Consultants, E/DI, International Resources Group 11:30-1200 Discussion on Session 2 Presentation 12:00-13:00 Lunch Session 3 - Rangelands Issues 1100-13:30 Experiences with Pastoral Associations in West Africa - What Should Be Our Future Strategy? Poul Sihm, AFTAG 'For the purposesof the seminar, we are defining "drylands" as areas with less than 800 mm annual rainfall. Staff will have a further chance to contribute to the Review of Dryland Experiences during a special review and brain-storming session on the draft 4eport in the next few weeks. 13:30-14:00 What are the Most Promising Technologies or Research Directions for Either Increasing Productivity or Improving Sustainability in Rangelands? What Should the Bank be Doing? Panel Discussion: Steven Berwick, Consultant, IRG John Glenan, EMTAG, Michael Walshe, AFTAG 14:00-14:30 Discussion on Session 3 Presentations Session 4 - Water and Soil Fertility Issues 14:30-1$:00 Strategiesfor Initiating Sustainable Moisture Conservation Programs in Semi-Arid West Africa Chris Relj, Free University, Amsterdam 15:00-15:30 Dryland Management: Water and Soil Fertility Issues Raymond Meyer, USAID 15:30-15:45 Coffee 15:45-16:15 Discussion on Session 4 Presentations Session S - Policy Issues for Improving Land Management 16:15-16:45 Why and How do Land Users Make Land Management Decisions? How Can Policies Encourage Land Users to Make Them Effectively? Prof. Piers laikie, University of East Anglia 16:45-17:00 Discussion on Session 5 Presentation 17:00-17:30 Vetiver Grass Slides and Discussion John Greenfield, ASTAG 17:00-19:30 Reception jhura- May 11 - Session 6 - Off-Farm Earnings and Migration Issues4 09:00-0930 77e% Importance of Non-farmn Income Sources and Some Food Policy Implications Chris Delgado, IFPRI aJohn Glenn will present the impressions of pants at a recent Holistic Resource Management Seminar (given by Allan Savory). oIe will uch both on a key underly ng technology (intensive animal impact) and0the broad HRMt th" /process We package thestther because we see off-fa earnings as occupationa migrotion. 0930-10:0 Temporary and Permanent Migration in West Africa - Some Early Findings from the Onch6cerciasis - Freed Areas Study: What Support Should the Bank Give to Migration? David Brokensha, Institute for Development Anthropology 10:00-10:15 Coffee 10.15-10:30 Discussion on Session 6 Presentations Session 7 - Participatory Issues, 10:30-11:00 The Aga Khan Foundation's Approach to Participation and Its Relevance to Bank Projects Andrew Weir, Aga Khan Foundation 11:00-11:20 Issues in Local participation and Their Relevance to Bank Projects David Richards, Director, NGO Support Program, CIDE/WRI 11.20-11:45 Discussion on Session 7 Presentations 11:45-12:40 Lunch (We will attempt to seat participants at lunch by Working Group for an initial brain-storming session) Sessiow 8 - Group Discussions 1240-14:10 Working Group Discussions on the 5 topics (see Attachment) - 14:10-14:25 Coffee Sessio 9- Plenary and Wrap-Up 14'Z5.-15:15 Reporting of Findings (10 minutes each group) 15:15-16:40 Plenary Discussion on *What Are the Priorities for Bank Action?" 16:40-17:00 *rap-Up by Prof. Stein Bie Group Discussions Group A - Rangelands Group Chairpersor To be decided QuestionS: 1. What technologies do we have to offer that will improve production or reduce degradation in rangeland areas? What research is needed?, 2. What shotld be the Bank's future approach with Pastoral Associations? Group B - Water and Fertility Technology Group Chairperson: Daniel Hillel, ENVDR Questions: 1. What should the Bank b6, doing to support research or pilot projects on vegetative moisture conservation techniques (e.g., vetiver, A guyanus, etc.) in dryland areas? 2. What should the Bank be doing to support soil fertility research - or pilot projects in dryland areas? Group C - The Policy Group Chairperso* David Steeds, AFSAG Questions: 1. Should the Bank be considering subsidies (e.g., fertilizer subsidies, credit subsidies, crop insurance, etc.) to reduce land degradation in dryland cropping areas? 2. What should the Bank be saying to borrowers about land tenure in rangeland areas? Group D - The Off-Farm Income and Migration Group Chairperson: Peter Hazell, AGRAP Questions 1. What can the Bank do in its project or policy work to increase off-farm income opportunities for dryland populations? 2. Should the Bank eA;pand support for temporary or permanent migration? What problems should,we watch out for? Group E - The Participatory Process Group Chairperson: Michael Cernea, AGRPS Questions: 1. How can Bank projects, while still working through governments, improve the process of participatory development? 2. What sort of training should the Bank support that would help the participatory process? II Annex 2. List of Participants AF2AG Jacob Kåmpen J-10059 34199 James Coates J-10037 34192 AF240 Allen Thomas J-9231 35579 AFSAG Elsie Garfield J-6284 35007 Jeffrey Lewis H-1089 37333 Siri Cerych J-6269 35009 David Steeds J-6271 35013 Daniel Moreau J-6261 34998 AF5PH Nicole Glineur J-7266 35043 AFTAG Alain Seznec J-9026 34128 K. McNamara J-9056 34122 Michael Walshe J-9079 34081 Poul Sihm J-9033 -34115 AFTEN Francois Falloux J-8257 35562 AGRPS Richard Meyers N-8083 30358 J. Hayward N-8069 30347 m. Cernea N-8059 30353 Peter Hazell N-8023 30435 AS2AG Engel B. joeresa E-8003 72365 AS3AG Rameash -Deshpandi ý. A-7033 74769 AS4AG Ben Van De Poll D-10113 76679 AS400 Tom Blinkhorn D-10003 76921 ASSAG Dirk Leeuwrik A-9025 74786 ASTAG Prabhakar Tamboli. F-3031 73519 John Lindt F-3027 72144 John Gr4enfield F-3031 76235 EDIAR Richard *anning W'5043 48135 ENAG Robert Hunt H-5077 32267- Muhawmad Naseem H-5243 32274 Michael Saddington H.Ik075 32262 EKTAG Claude Blanchi H-7111 -32464 John Glen H-7131 32462 EN Richard Ackerman' H-9039 32606 Stephen Lintner H-9037 23508 ENVPR J.N. Trdlldalen S-3046 3330t - John English S-3063 31024 ENVDR J. Listortt S-5037 35102 D. Hillel S-5038 35104-- ENVOS H. Hassan S-5073 31376 IENiE WULlem Floor S-2139 33251 1A1AG Dirkvan der Sluija 1-7191 30056 LA1CO Robert Schneider 1-7051 31931 LEGAF T. Mpoy-Kamulayi E-6085 76259 - "j - --. 3 Special Invitees and External Speakers J. Gritzner World Resources Institute S. Berwick Consultant Henrick Marcussen Principal Technical Advisor, UNSO, New York C. Hutchinson Resources for the Future P. Reining American Assn. of Ad. of Science Stein Bie NORAGRIC, Norway Michael Mortimore University of Birmingham, U.K. Juan Seve International Resources Group Burce Ross-Sheriff International Resources Group Chris Reij Free University, Amsterdam Raymond Meyer USAID John Gaudet USAID - Africa Bureau Michael McGahuey USAID - Africa Bureau Piers Blaikie University of East Anglia Chris Delgado IFPRI David Brokensha Institute for Development Anthropology Andrew Weir Aga Khan Foundation David Richards NGO Support Program, CIDE/WRI