Report No. 18133 Mauritania Impact EvaluationReport Second LivestockProject (Credit 1658-MAU) Public Disclosure Authorized
June30, 1998
OperationsEvaluation Department Public Disclosure Authorized
J IV Public Disclosure Authorized Public Disclosure Authorized Currency Equivalents Currency unit = Ouguiyas (um)
1987 US$1.00 73.4 um 1988 US$1.00 = 75.2 um 1989 US$1.00 = 83.0 um 1990 US$1.00 80.7 um 1991 US$1.00 82.1 um 1992 US$1.00 87.5 um 1993 US$1.00 = 119.9 um 1994 US$1.00 123.6um 1995 US$1.00 127.5um
Abbreviations and Acronyms
BIE Bureau des Intrants pour l'Elevage (Office of Livestock Inputs) DNE Direction Nationale de I'Elevage (Livestock Department) FY fiscal year GNAASP Groupe National des Associations Agro-Sylvo-Pastoral (National Group of Agro-Sylvo-Pastoral Associations GNAP Groupe National des Association Pastoral (National Group of Pastoral Associations) GOM Government of Mauritania ICR Implementation Completion Report LIL Learning and Innovation Loan MDRE Ministere du Developpement Rural et de l'Environnement (Ministry of Rural Development and Environment) PA pastoral association PCA pastoral cooperative association PCR Program of Consolidation and Renewal PREF Program for Economic and Financial Adjustment SOMECOB Societe Mauritanienne de la Commercialisation du Betail (Mauritanian Livestock Marketing Company)
Director-General, Operations Evaluation Mr. Robert Picciotto Director, Operations Evaluation Department Ms. Elizabeth McAllister Manager, Sector and Thematic Evaluations Group Mr. Roger Slade Task Manager Mr. John English The World Bank Washington, D.C. 20433 U.S.A.
Office of the Director-General Operatons Evaluation June 30, 1998
MEMORANDUMTO THE EXECUTIVEDIRECTORS AND THE PRESIDENT
SUBJECT: Impact EvaluationReport on Mauritania: SecondLivestock Project (Credit1658-MAU)
The attachedImpact EvaluationReport (IER) presentsthe results of an OperationsEvaluation Department(OED) studyof a livestockproject in Mauritania. The IDA credit was approvedin FY86 and closed in FY95. The evaluationwas based on a review of Bank files, includingproject completion and appraisalreports, field visits to project sites, and interviewswith projectbeneficiaries and governmentand Bank staff. The evaluationteam used a varietyof participatorytechniques to obtain the views of the principal stakeholders,direct beneficiaries(principally pastoralists), agency staff, governmentstaff, and others related to the livestocksector.
External supportto the livestocksector in semi-aridregions has increasinglyemphasized the developmentof pastoralists' groups as a means of facilitatinglinks betweenpastoralists and support services,especially in animal health, and also to help improvemanagement of rangelandand water resources on which pastoral livestockproduction is based. Little assessmenthas been done of the developmentand performanceof pastoralistgroups of this type. The SecondLivestock Project in Mauritaniastaked its successon such a scheme. The evaluationfocuses on the performanceof the project in developingpastoral cooperativeassociations (PCAs), and reviewstheir development,current position, and the issues that are arising in their continuedevolution two years after project closing.
At present the 39 PCAs formedunder the project are groupedin a national organization,the National Groupingof Agro-Sylvo-PastoralAssociations. The PCAshave an estimatedmembership of 20,000, or more than one-thirdof the pastoralisthouseholds in the country. In principlethis is a considerableachievement, but membershipis often nominal,and the PCAsare in a highly variablebut generallyprecarious state. Almosttwo-thirds of the PCAsinvestigated by the evaluation(8 out of 13) consist entirely of Maurish tribes, while the othersare ethnicallymixed. The vast majorityof PCA leadershipis male, generallyaged between 45 and 56, but women are much more active at the level of the member cooperatives. The leaders of the individualPCAs investigated are generallyfigures of some importance,having both urban andrural interests. Membersof the committeesare predominantlylocal. The active membershipis small,the PCA assembliesseldom meet, and only the smallermanagement committeesmeet with any regularity. Most maintaina smallrevolving fund for the purchase of veterinary supplies,and have undertakensmall productive miniprojects. The PCAs were not able to achieve effective control over their grazing areas and performancein resource conservationand measuresto improve managementwere limited.
While the project succeededin establishingthe PCAs, it was inconsistentover time in its approachto the organizations(moving from experimentingwith a few pilot associationsto expansion and back to consolidation). This inconsistencycreated confusionand conflictbetween the variousroles of the associationsand kept them from developinga clear status in the country. This problem was increasedby the large size and scale of the PCAs (typicallythey cover 2,000-2,500kms2 with a population of about 5,000), which made it difficult for them to overcome internal heterogeneity of interests. Furthermore, the project continued to emphasize range management even though the PCA members were more concerned with water and animal health, and despite evidence that the efforts to establish a freer system of distribution of veterinary products were appreciated by the herders and proved the most popular and durable of all the initiatives. However, this primary concern of the project with natural resource management was not linked to a strategy based on a detailed understanding and analysis of the complex dynamics of contemporary production systems or an appreciation of the major environmental constraints in a rapidly evolving ecological situation. A particular source of frustration to PCA membership was that, at the outset, they understood that the project would place priority on rehabilitating or drilling wells for stock water points. However, this only materialized in a minor way at the end of the project.
Following completion of the Second Livestock Project, support to the PCAs was continued with a view to their playing a role in the successor Rainfed Natural Resource Management Project (PGRNP), approved in FY97. This project also has a primarily natural resource management focus, with greater emphasis at the local level on village development, and involving the full range of land users. To date the PCAs have not been formally involved in this process, and they believe they are being sidelined in favor of more village-based entities, to the potential disadvantage of the pastoralists. It is important for the Bank to recognize that it cannot help to initiate community groupings and then ignore them in subsequent operations that affect them, if it and they are to retain credibility.
In summary, the PCAs have not developed a clear role and consequent visibility, either locally or nationally. They risk becoming sidelined and withering into insignificance. Given the marginality of pastoral livestock producers and their need to integrate into the larger economy, this is increasing frustration and could lead to tensions that will undercut other efforts, including those of the PGRNP. If this is to be avoided, it is essential that the PGRNP recognize and, as far as possible mesh with, the structures developed under the Second Livestock Project. The PGRNP should: (a) work in a fully participatory manner with the PCAs and other local groupings in further developing its strategy for natural resource management and development, and (b) give high priority to improvement of stock water points and, to the degree feasible, complete the program identified under the Second Livestock Project, otherwise confusion and disillusionment are likely to result, to the detriment of the PGRNP's ultimate performance.
The findings of the study indicate a number of more general lessons for efforts to support pastoralists. First, an emphasis on resource management requires intensive support for local initiatives, at least for some time, but it also requires recognition of the limited array of new technologies available for extensive production and natural resource management systems. Priority should be given to actions that will help pastoralists better integrate into the larger economy and society, rather than to attempts to devise and introduce improved technologies. Second, consideration should be given to actions that will foster the growth of a broader pattern of local leadership in the communities, seeking neither to simply reinforce the position of the traditional leadership nor to promote an unrealistic degree of "popular participation," but to provide an enabling environment for new initiatives and the emergence of new cooperative groupings. Third, when strong local structures for development-oriented activities do not exist, a sequenced and process approach to project operations is required, especially when the local socio-economic and resource management systems are not well known. Contents
Preface ...... iii Executive Summary ...... 1
1. Introduction ...... 3
2. Project Design and Implementation ...... 5 Background ...... 5 Implementation and Outcome ...... 6 The ICR Assessment ...... 6
3. Development and Current Status of the Pastoral Associations...... 8 The Development of Pastoral Associations...... 8 The Current Status of the Pastoral Associations ...... 10 Activities of the Associations ...... 13
4. Impact of the Pastoral Associations on the Livestock Sector...... 17 The Impact of the Presence of the Project...... 17 The Nature of the PCAs ...... 17 PCAs' Inability to Establish Legal Rights over Resources ...... 18 PCAs' Dependence on External Support ...... 19 The Limited Nature of Services to Members ...... 20
5. Conclusions and Implications for the Future...... 22 Performance of the Second Livestock Project in Developing Local Institutions ...... 22 Linkage with Rainfed Natural Resource Management Project...... 23 Broader Implications for Pastoral Interventions...... 25
Annex 1: Project Data, Objectives, Components, and Outputs ...... 29 Annex 2 Table 1: Data on Establishment and Operation of Pastoral Associations...... 31 Annex 2 Table 2: Pastoral Associations Revolving Funds ...... 33 Annex 3: Comments from the Borrower ...... 35 References ...... 39 Map IBRD 29574
Boxes
1.1 Methodology ...... 4 iii
Preface
The Mauritania Second Livestock Project, supported by Credit 1658-MAU for US$7.6 million equivalent was approved in FY86. The project did not become effective until June 1987, and an extension of three years from the planned closing date was agreed at the 1990 mid-term review. The project was restructured in 1993 and a revised credit agreement was approved by the Board in FY95. The project closed on Jtne 30, 1995. Cofinancing was provided by the African Development Bank (US$6.3 million equivalent) and the OPEC Fund (US$2.0 million).
This report is based on a field study in the project area carried out between October and December 1997 by Prof. David Seddon of the School of Development Studies, University of East Anglia, U.K., with the assistance of two Mauritanian researchers, Dr. Abdel Wahoud Ould Cheikh and M. Isselmou Abdelkader. The study included individual and group interviews with beneficiaries and interviews with local and national government officials and other persons involved in livestock production. The report also draws on project documents and interviews with Bank staff in Washington and Nouakchott. The kind cooperation and valuable assistance provided by all these individuals and agencies are gratefully acknowledged.
Following standard OED procedures, copies of the draft report were sent to the relevant government officials and agencies for their review and comment. Comments from the Ministry of Rural Development and the Environment in Mauritania have been received, taken into consideration, and are reproduced in full as Annex 3. 1
Executive Summary
1. Externalsupport to the livestocksector in semi-aridregions has increasinglyemphasized the developmentof pastoralists' groupsas a means of facilitatinglinks betweenpastoralists and support services,especially in animalhealth, and also to help improve managementof rangeland and water resources on which pastoral livestockproduction is based. Little assessmenthas been done of the developmentand performanceof pastoralistgroups of this type. The Second LivestockProject in Mauritaniastaked its successon such a scheme. The evaluationfocuses on the performanceof the project in developingpastoral cooperativeassociations (PCAs), and reviewstheir development,current position,and the issues that are arising in their continued evolutiontwo years after project closing. It was based on a review of Bank files, including project completionand appraisalreports, field visits to project sites, and interviewswith project beneficiariesand governmentand Bank staff. The evaluationteam used a variety of participatorytechniques to obtain the views of the principalstakeholders, direct beneficiaries (principallypastoralists), agency staff, governmentstaff, and othersrelated to the livestock sector.
2. At present the 39 PCAs formedunder the project are groupedin a national organization, the National Groupingof Agro-Sylvo-PastoralAssociations. The PCAs have an estimated membershipof 20,000, or more than one-thirdof the pastoralisthouseholds in the country. In principle this is a considerableachievement, but membershipis often nominal, and the PCAs are in a highly variablebut generallyprecarious state. Almosttwo-thirds of the PCAs investigated by the evaluation (8 out of 13) consist entirelyof Maurishtribes, while the others are ethnically mixed. The vast majority of PCA leadershipis male, generallyaged between 45 and 56, but women are much more active at the level of the membercooperatives. The leaders of the individualPCAs investigatedare generallyfigures of some importance,having both urban and rural interests. Members of the committeesare predominantlylocal. The active membershipis small, the PCA assembliesseldom meet, and only the smallermanagement committees meet with any regularity. Most maintain a small revolvingfund for the purchaseof veterinary supplies,and have undertakensmall productive miniprojects. The PCAswere not able to achieve effective control over their grazing areas and performancein resource conservationand measures to improve managementwere limited.
3. While the project succeededin establishingthe PCAs,it was inconsistentover time in its approach to the organizations(moving from experimentingwith a few pilot associationsto expansionand back to consolidation). This inconsistencycreated confusionand conflict between the various roles of the associationsand kept them from developinga clear status in the country. This problem was increasedby the large size and scale of the PCAs (typicallythey cover 2,000-2,500kms with a populationof about 5,000), whichmade it difficult for them to overcomeinternal heterogeneity of interests. Furthermore,the projectcontinued to emphasize range managementeven though the PCA mernberswere more concernedwith water and animal health, and despite evidencethat the efforts to establish a freer systemof distributionof veterinaryproducts were appreciatedby the herders and proved the most popular and durable of all the initiatives. However,this primary concern of the project with natural resource managementwas not linked to a strategybased on a detailedunderstanding and analysisof the complexdynamics of contemporaryproduction systems or an appreciationof the major environmentalconstraints in a rapidly evolvingecological situation. A particular source of 2 frustration to PCA membership was that, at the outset, they understood that the project would place priority on rehabilitating or drilling wells for stock water points. However, this only materialized in a minor way at the end of the project.
4. Following completion of the Second Livestock Project, support to the PCAs was continued with a view to their playing a role in the successor Rainfed Natural Resource Management Project (PGRNP), approved in FY97. This project also has a primarily natural resource management focus, with greater emphasis at the local level on village development, and involving the full range of land users. To date the PCAs have not been formally involved in this process, and they believe they are being sidelined in favor of more village-based entities, to the potential disadvantage of the pastoralists. It is important for the Bank to recognize that it cannot help to initiate community groupings and then ignore them in subsequent operations that affect them, if it and they are to retain credibility.
5. In summary, the PCAs have not developed a clear role and consequent visibility, either locally or nationally. They risk becoming sidelined and withering into insignificance. Given the marginality of pastoral livestock producers and their need to integrate into the larger economy, this is increasing frustration and could lead to tensions that will undercut other efforts, including those of the PGRNP. If this is to be avoided, it is essential that the PGRNP recognize and, as far as possible mesh with, the structures developed under the Second Livestock Project. The PGRNP should: (a) work in a fully participatory manner with the PCAs and other local groupings in further developing its strategy for natural resource management and development, and (b) give high priority to improvement of stock water points and, to the degree feasible, complete the program identified under the Second Livestock Project, otherwise confusion and disillusionment are likely to result, to the detriment of the PGRNP's ultimate performance.
6. The findings of the study indicate a number of more general lessons for efforts to support pastoralists. First, an emphasis on resource management requires intensive support for local initiatives, at least for some time, but it also requires recognition of the limited array of new technologies available for extensive production and natural resource management systems. Priority should be given to actions that will help pastoralists better integrate into the larger economy and society, rather than to attempts to devise and introduce improved technologies. Second, consideration should be given to actions that will foster the growth of a broader pattern of local leadership in the communities, seeking neither to simply reinforce the position of the traditional leadership nor to promote an unrealistic degree of "popular participation," but to provide an enabling environment for new initiatives and the emergence of new cooperative groupings. Third, when strong local structures for development-oriented activities do not exist, a sequenced and process approach to project operations is required, especially when the local socio-economic and resource management systems are not well known. 3
1. Introduction
1.1 In the early 1980sWorld Bank experiencein the Sahel pointed to the value of producers' associations(pastoral associations) as vehicles for sustainablelivestock and pastoral development.At the same time, stoppingthe degradationof grazing lands,particularly by preventingoverstocking on open-accessrangeland, became a centraltheme in the design of livestockprojects for semiaridareas. This arguedfor some means of clarifyingland rights and controllinggrazing managementto prevent degradation.Some experts also feared that the increaseduse of wells to water livestockmight encourageexcessive stocking rates. These issues gained added importancewith the 1982-83 drought,which caused severe livestocklosses and dislocationof pastoral groups.
1.2 Over the past decade,numerous livestock development and natural resource managementprojects in north and west Africa (includingprojects fundedby, the Bank)have been pursuing the same goal: buildingamong pastoralists a local institutionalcapacity for the managementof naturalresources. A group audit of three projects in the CentralAfrican Republic,Niger, and Senegalin 1993reviewed projects implementedduring the 1980s.I It noted the increasingrole of herders' associationsbut did not review their experiencein any detail. Assessingthe degree of project success in this process of capacitybuilding is thereforecentral to assessingthe sustainabilityof such projects. Giventhe increasingprominence of pastoral associations(or similar institutions)in livestockand naturalresource management projects, identifyingfactors in the long-termeconomic and organizationalviability of such associationsis a priority for assessingthe performanceof such projects.The SecondLivestock Project offers an opportunityto make this identification;that is the principalobjective of this evaluation.2
1.3 This impact study concentrateson sustainabilityand institutionalimpact especiallywith regard to pastoral associations,looking at the degree to which the institutionsbeing fosteredhave evolved and seem likely to evolve in the future. The evaluationhas soughtto understandnot only what the project achievedbut also how the project evolved and why it did so in the way that it did. The lack of ongoingproject monitoringand evaluationhindered this pursuit and obliged the studyteam to rely heavily on documentaryevidence from Bank files and other externalreviews duringthe project's lifetime,as well as fieldworkundertaken after the project's closure.The methods used in the evaluationare outlinedin Box 1.1.
1. PerformanceAudit Report,No. 12791,Operations Evaluation Department, World Bank, February25, 1994. 2. The detailedfindings of the investigationare containedin the full report by Prof. DavidSeddon, Dr. AbdelWahoud Ould Cheikh,and M. IsselmouAbdelkader, which is availableseparately. The currentreport was preparedby John Englishand David Seddon. 4
Box 1.1: Methodology
This evaluationis basedon a detailedreview of documentaryevidence, from files and reports in Washingtonand other materialscollected in Mauritaniaand elsewhere,and from fieldworkconducted in three stagesby the team leader and two Mauritanianconsultants from Septemberto December1997. The methodologyadopted for the fieldworkconsisted of three phases:
(i) An initialreview and designphase involved fieldwork by the full team at severalsites. During this phase, qualitative,participatory rural appraisaltechniques (open-ended interviews with key informants, focus group discussions,site visits, transects)were used to developa detailedappreciation of variability acrossthe project area and a design for the secondfieldwork stage.
(ii) Duringthis longerphase national consultants, working to an agreedbrief, carriedout detailed investigationsover a periodof two months.One teammember concentrated on the broaderpolitical- juridical contextand developmentsat the nationallevel, interviewingkey individualsin governmentand the private sector,as well as the officers of GNAP and reviewingofficial documents.
The other team memberundertook local-level fieldwork. This includedinvestigation of a sample of the pastoral associationsestablished by the project(located across the entireproject area), as well as semistructuredinterviews with a range of key informants(including relevant local officials)and group and focus discussions.The size of the project area, approximately150,000 km 2, and its location,up to 500 km fromNouakchott (see Map No. IBRD29574), togetherwith the limitedcommunications in the region, made beneficiariesdifficult to contactand reduced the scopeof the interviewingthat the study team could undertake.The investigationsregarding these associationsincluded (a) interviewswith association presidentsand, where possible,members of the bureaux,and (b) interviewswith ordinarylivestock herders.While the former were usually structuredand took place indoors,the latter were less structured and more open-ended,and took place at waterpoints, in marketplaces,or in grazing areas at random across the entire area coveredby the field study.This approach,while clearlyunable to provideinformation amenableto statisticalanalysis and obviouslylimited, nevertheless clearly revealed the gap betweenthe ordinaryherder, supposedlyrepresented by the pastoral associations,and the representativesof the associations.
(iii) The third stage involvedcross-checking between the three membersof the team and drafting a broad agreed frameworkfor the production,by the team leader,of the main report. 5
2. Project Design and Implementation
Background
2.1 The SecondLivestock Project followedon fromthe First LivestockProject (Credit 273- MAU), which was approvedin November 1971 and closed in December1978. The first project concentratedon improvinganimal health and outputby vaccinationand by the constructionand rehabilitationof wells, with supplementaryactivities relating to natural resource management (such as the constructionof firebreaks).The droughtof 1972-73, and the associatedmovement of the bulk of the livestockpopulation out of the project area, adverselyaffected the project and in 1975 it was revised and effectivelyreduced in scope,but it achievedthese revised targets. The vaccinationprogram was scaled back, the firebreakprogram was no longer feasibleafter the drought, and only 137 wells out of the 300 initiallyplanned were reconstructedor repaired.3 While pastoralistswere keen to have operationalwells, maintenanceof those repairedunder the project was a problem as individualgroups were not allotted rights and responsibilitiesfor specificwells, and governmentdid not have the financialresources to do the job. The audit agreed that on balance the project shouldbe considereda successbut noted that, because of the adverseconditions, the impact of the vaccinationprogram couldnot be assessed.
2.2 The conception,design, and preparationof the SecondLivestock Project extendedover an exceptionallylong period. Althoughthe project was conceivedin the late 1970s,preparation was extremely slow: appraisalwas in October 1983,negotiations in October 1985, and effectivenessin June 1987.
2.3 Throughoutthis period there was considerabledebate within the Bank, betweenthe Bank and other potential financiers,and between the Bank and the Governmentof Mauritania(GOM) about the precise orientationand focus, and means of implementation,of the proposed project. During the period another major droughtoccurred and Mauritaniaexperienced considerable political instabilityand frequent changesin governmentpolicy. While this prolongedpreparation theoreticallyallowed time for detailed and effectivepreparation it was marked in practiceby considerabledelays, hesitations, and reconsiderations.
2.4 Debate betweenthe variousparties revolved aroundthe relative importanceof pastoral associations(whether this effort shouldbe just a pilot or somethinglarger, and the size, structure, legal status, and likely role of the associations),the extent of assistanceto the public livestock servicesand improvementsin infrastructure(e.g., wells and abattoirs);and the need to undertake further studiesto resolve some of the uncertaintyover the structure and performanceof the livestock sector in the country.A number of issues were never settled and created implementationproblems.
2.5 As approved,the objectivesof the project were to slow the deteriorationof grazing areas and to boost livestockproductivity by making groups of pastoralistsresponsible for allocated
3. These were wells or boreholespreviously constructed by government. 6 grazing areas, providingbetter servicesand productioninputs to pastoralists,and increasingthe knowledgeof production systemsessential for formulatinga developmentpolicy and planning future operations.
Implementationand Outcome
2.6 The componentsand outcomesof the project are summarizedin Annex 1. Implementationwas slow. Severalinstitutional changes were required as conditionsof effectiveness,and cross conditionalitywith the other two cofinanciershad to be resolved. This was not completeduntil more than a year after signing.By the mid-termreview in 1990, the project had achievedhardly any of its objectives.Implementation had been doggedby rapid turnover of personnel,particularly in the post of project director,and by procurementproblems that were exacerbatedby the differingrequirements of the cofinanciers.However, there was optimism,particularly on the Bank side, surroundingthe prospectsfor the developmentof pastoral cooperativeassociations. Fourteen PCAs had been establishedand more than 100 had been identified as possible.To make up for lost time and to ensurethat the effort had an adequate opportunityto succeed,the Bank agreed, unusually,to an extension of the planned closing date by three years, to 1995.Project objectivesremained unchanged, however. At the same time, the Governmentof Norway made an additionalgrant to GOM,principally to support the establishmentof PCAs, to be administeredin parallel with the project.
2.7 The perceived successof the project in establishingthe first PCAsencouraged the view that a rapid expansionshould now be pursued. A report on the generalizationof the PCAs producedin May 1990 anticipatedexpansion over the period 1991-95 to producea total of more than 60 PCAsby the new closing date in 1995. Substantialprogress was made in this direction over the next two years. By 1992, 39 PCAshad been establishedand most of the physical works for the livestockservices had been completed,but little progress had been made on the remainingcomponents. A Bank supervisionmission discussedsteps to streamlinethe project by terminatingthe lagging components,but eventsovertook the discussions.A major reorganization of the Ministryof Rural DevelopmentEnvironment (MDRE) eliminated the Livestock Departmnent(DNE-the managingdepartment for the project) and decentralizedits activities.To regularizethe situation,and given the continuedpoor performanceof the project,the Bank fieldeda restructuringmission in 1993.4This missionagreed with the governmenta reformulatedproject that had only two components(the support of PCAs and constructionof an abattoir)and a revised structure.The Board approvedthe new credit agreementin 1995.
The ICR Assessment
2.8 The overallperformance of the project was mixed, and the ICR rated the outcomeas unsatisfactory.The ICR rated sustainabilityof the projectas uncertain,its achievementof institutionaldevelopment objectives as partial, and Bank and borrower performanceas deficient. It was most upbeat about the performanceof the pastoral associationsand concludedthat "althoughthe project has been considereda problem project for the entire duration of its implementation,it was ultimatelysuccessful in establishingPCAs which are properly focused on the target groups of beneficiaries. The 22 PCAsthat were restructuredand appear sustainable
4. The ministerialrestructuring of MDREwas sufficientlyextensive that the Bankdecided that a revised credit agreementwas required. 7 could well serve as nuclei for developmentactivities and mightbe spontaneouslyreplicated in other areas, spurredby beneficiarydemand" (p. iv). It argued that "concerningthe core activity, i.e. the establishmentof the PCAs, those that underwentrestructuring have demonstrated and adequate level of organization,shared objectivesand active beneficiaryparticipation. Together with their umbrella organization(GNAP) they appear sustainableand could serve as a basis for further rural developmentactivities. However,new PCAsare needed in order to achieve the long term project objectiveof nationwidemaintenance and improvementof grazing areas under legally ensured control" (p. 9).
2.9 Regardingthe sustainabilityof the pastoral associations,the ICR concludedthat "it appears likely that those that have so far been restructuredare likely to continuewithout further outside financial assistance. Pastoralistsare convincedof the advantagesof membership,i.e. access to veterinaryinputs and feed, lower water fees, supplementaryfunding of miniprojects, rapid access to extension advice,and representationof their interests at national level through GNAP. The financesof the PCAs,based on membershipfees (UM 300 (US$2.35)per annum), water fees and input sales, as well as incomefrom mini-projects,are still weak. Membership numbersare still rising, however,and shouldenhance the PCAs' financialviability" (pp. 7-8). It suggestedthat "the mini-projectsappear sustainable,especially those dealingwith milk production, as they have been replicatedby individualfarmers, without outside assistance. For other operations,such as pasture developmentand hay harvesting,it is too early to assess their sustainability,although there are serious doubts abouthay harvesting,since the market price for hay does not includethe depreciationof equipmentrequired for mechanicalharvesting" (p. 8).
2.10 As these quotes indicate,the ICR is generallypositive about the outcomeof the project's efforts regarding the PCAs. However,evidence to support many of these assertionsis lackingin the ICR. As will be seen, the field investigationscarried out for this evaluationprovide limited support for the ICR conclusion,which appearsto have been based on an excessivedegree of optimism. 8
3. Development and Current Status of the Pastoral Associations
The Developmentof PastoralAssociations 5
3.1 As noted, project implementationinitially emphasized the establishmentof PCAs. By early 1989, 18 months after the projectbecame effective,7 of the planned 15 experimental associationshad been established,and the other 8 were planned for 1989.6Each compriseda number of groupsof about 20 to 30 families.Each group was representedby a delegate to a generalassembly of the association.This was to meet annuallyto approvethe annual contributionof each memberto the revolving fund, which constitutedthe basis for activitiesand investmentsby the association.Each PCA covered an area of about 2,000 to 2,500 km2,with a populationof about 5,000 persons.
3.2 There was a major push immediatelyafter the mid-termreview (1990-92) to expandthe number of PCAs and to establisha nationalgrouping of some kind. A draft constitutionfor a NationalGrouping of CooperativePastoral Associationswas drawn up, and a national congress to inauguratethe national associationwas proposed.This took time to arrange,but in July 1992 representativesof the 39 pastoral associationsexisting then met and elected a NationalBureau (president,secretary-general, and committeemembers). The NationalGrouping of Pastoral Associations(GNAP) came into being formallyas a properly constitutedinstitution to represent the interests of its members.
3.3 By this time the experimentalphase was completedand, with the additionalNorwegian funds being available,the Bank proposedthat the generalizationof PCAsbe promoted. A review had indicatedthat about two-thirdsof the associationshad a satisfactorylevel of activity.At the same time, however,some cloudswere appearing.Representatives of the associationshad indicatedto the Bank their consternationat the failure of the project or the governmentto establish clear rights for the associationsregarding pasture and water points.7 Withoutsuch legal powers the capacity of associationsto managethe natural resourceswithin their domain remainedextremely limited.A 1990 governmentcircular sent to local authoritiesindicated that pastoral associationscould be granted rights under existingdecrees, but this power was rarely used.8 In additionthe early enthusiasmfor the associationconcept was linked to the expectation
5. The content of this chapteris primarilybased on a surveyof a sampleof 13 pastoralassociations covering all seven majorregions, selectedto providea broaderbasis for assessingthe situationin the projectarea. Of theseassociations, sevenwere establishedin 1988,one in 1989,three in 1990,one in 1991,and one in 1992.Their distributionis indicatedin Map No. IBRD 29574. 6. By the end of 1989, 11 associationshad been formed,but this progresswas interruptedby the inter-ethnicviolence of 1989-90, which spilledover into Senegal.The tensionsextended into someof the associations,particularly in the SenegalValley, and undoubtedlyaffected further development. 7. The credit agreementprovided that the PCAs shouldbe "grantedlegally enforceable grazing and stockwater rights, includingthe right to deny accessto grazingareas reservedfor improvements"(Section 3.05). This was to be as providedfor under the 1983 Law and LandOwnership, which vestedland in the state. 8. Such powers can be held by associationswith "publicinterest" status, which are supervisedby the Ministryof the Interior. The PCAsdid not have this status. 9 that additionalresources would be made availableto them, in particular for developmentof water points. As time passed and this activity did not get off the ground, enthusiasmwaned. 9
3.4 The creation of the GNAP coincidedwith a major change in the political environmentin the country.In 1991, after more than a decade of militaryrule, a new constitutionestablished the frameworkfor a multipartysystem. Presidential elections were held in January 1992, and legislativeelections followed in March. This change greatly increasedthe possibilityfor the developmentof independentinterest groupings.In addition,a new law in January 1993 provided a stronger legal frameworkfor the constitutionof cooperatives,but gave pre-cooperatives,such as the PCAs only two years to adapt to the new legal requirements.The Bank also recognized that many of the associationsestablished between 1988 and 1992 had failed to maintainthe level of activity anticipated,in part because of a lack of supervisionby the project and in part because of a lack of both human and financialresources.
3.5 By 1994pressure for the restructuringof GNAPwas considerablefor both institutional and structuralreasons. The new law affectednot only its constituentassociations, but its status as well. It was also felt that the low level of activity among the PCAs,linked to the top-down structure of the associations,might be progressivelyimproved if they could be seen as local unionsof smallercooperatives with more various activitiesand interests.The PCAs,therefore, were evolving into unionsof cooperativeswith a mandateto coordinatethe local cooperatives, while the GNAPbecame a Federationof Unions.At the same time, the changedlegal statutes allowed for individualor group membership.
3.6 These changes,together with the project's continuedsupport for the PCAs, encouraged the associationsand the unionsto continueto maintaina degree of vitality.PCA officers reportedrenewed interest after the 1993-95 effortsto revitalizethe associations,despite the failure to provide real trainingor support, or to rehabilitateor reconstructthe much-needed pastoral wells, and the failureto put into effect the legal provisionsfor the attributionof rights over water and pasture so that they might be able to managetheir own resources.
3.7 After the IDA credit closed in 1995, some supportfor the associationscontinued through the cofinanciersand financiers;this supporteda congressin March 1997.Among other things, the congress agreed to change the name of the umbrella organizationto the NationalGrouping of Agro-Sylvo-PastoralAssociations (GNAASP). The new title reflects a recognition,by the leadershipat least, of the importancelending agenciesand the governmenthad come to place on integratednatural resource management.
3.8 By this time, 11 of the associationshad been transformedinto unions and the other 28 were expectedto followsoon. Thus, a post-projectevolution of the structuresestablished by the SecondLivestock Project is under way. The extentto which these structurescan be sustained without further external supportis debatable.
9. The field surveyreview (see Box 1.1) foundthat the project wellshad not been constructed.Initially wells were includedas one of the itemsto be financedlargely from the ProductionImprovement Fund (PIF).They loomedlarge as a potentialbenefit to pastoralists,and potentialsites for new wellsor rehabilitationwere drawn up. However,since the PIF never materialized,there was no programon this front until an attemptwas made in 1993 to undertakewell drillingdirectly under contract. 10
The Current Statusof the PastoralAssociations
7TheirCoverage
3.11 The system of delegatesmakes it difficultto determinehow many real members a PCA has. The size of the associationsvaries considerably.Evidence suggests that, whateverthe formal membershipof the PCA, only a very smallminority is closely involvedin association activities.Furthermore, the evidence suggeststhat only about 2 percent of those involved in livestockproduction had any real contactwith the PCAs.The 39 PCAslisted as members of the GNAPcould be said to cover or be present in 1,100to 1,420villages or groups (see Annex 2, Table 1). An estimateproduced by the appraisalreport for the ongoingRainfed Natural ResourcesManagement Project (PGRNP)concluded that "the PCAs are present in 1,000 villages and include660,000 people or 34 percentof the rural population"(para. 1.22).This is clearly a gross overestimationinvolvement in the PCAs.
3.12 The number of delegatesat the 1995national conferencewas 1,189.If each delegate represents 30 families,the PCAs would representsome 36,000 families (or 216,000people, assuming6 personsper family).The PGRMP appraisalreport refers to 1,470 delegates,which might indicate an even larger number of persons officiallyincluded within the PCA structure- around265,000. But this figure (equivalentto most of the remainingpastoralists in Mauritania) also appears substantiallyoverestimated. The more detailed figure of 120 cooperatives representedin the GNAP, with a reported membershipof 17,157might be more realistic.
Ethnic Composition
3.13 The PCAswere initially constitutedat meetings (organizedby the proposed delegatesto the managementcommittee or bureau) of village groupswithin the defined territoryof the association.These groups usually correspondedto fractionsor subfractionsof tribes. The tribal structuresthrough whichpastoralists organized their economicand social life remain of considerableimportance even today. As a mode of economicand social organization,and of natural resource management,the tribe and subtribecontinue to have significance.
3.14 The majorityof the PCAs investigatedconsisted entirely of Maurish tribes and tribal fractions(8 out of 13). Three comprisedboth Maures and Peuls,while one comprisedPeuls and Soninkeand no Maures, and another,while dominatedby Maures,included some Wolof. The main ethnic groups-apart from the Wolof-remain highly stratifiedwith former "slave" groups still severely disadvantagedand wealthy Beidane(white Maures) still dominantin most rural areas.
3.15 The bias of the sample towardareas in whichMaures predominateand the virtual exclusion of village-typeassociations explains in part this pictureof Maurish dominance.It is significant,however, that the Maurish dominancepervades the PCAs-even at the national level, where Maures dominatethe executivecommittee and secretariatof the GNAASP.Out of 35 pastoral associationsbelonging to the GNAASPthe vast majority(31) have Maurish presidents; two have Peul presidents(a third died recently), and one has a Soninkepresident.
3.16 In trying to assess the significanceof the ethnicand tribal affiliationof the bureaux and representatives,it is temptingto stressthe dominanceof the Maures and the tendency for one 11 tribal group to dominatean association.On the other hand, the variation can be seen as an indication of the potential for PCAs to bring togetherdifferent groups within one local institution to serve common interests.There can be no doubtthat the PCAsreflect the political and social structuresof rural Mauritaniansociety, but this society is changing.It couldbe argued that the PCAsprovide yet another form of civil society capableof encouragingcollective action of variouskinds, which cuts across or transcends,at least in a few cases, the ethnicdivisions that have often proved so sharp in the recentpast. Based on evidenceboth from earlier reviews of the situationin the PCAs and the evaluationteam's investigations,however, clear political divisions exist within most PCAs,often as a result of significantcleavage between the leaders of tribes and tribal factions. In several cases these divisionsinhibit the effectivefunctioning of the PCA.
Age and Gender Composition
3.17 Over the past two decades,Mauritanian rural economyand societyhave becomemore diversifiedand individualshave been obligedto take on new responsibilitiesand adopt new roles. Young males increasinglyhave tended to emigratein search of work abroad or in the towns;women now outnumbermen in the rural areas and are householdheads in 20 to 30 percent of cases.Women's participationin agriculturalproduction has increasedand the gender division of labor has been substantiallyaltered.
3.18 The vast majorityof bureau members are male. Where womenare includedit is generallybecause of their involvementin a local cooperative,usually a women's cooperative. Out of a total of 80 bureau membersin seven of the associationsconsidered in more detail, 7 (9 percent)were women, all of whom representedlocal cooperatives.In a few cases (althoughnot in the associationsinvestigated) women may be members of the bureau because of their personalityor skills.Until very recently one of the PCAs had a womanpresident, and in a few other cases there are womenmembers of bureauxwho are not involved as representativesof cooperatives.The overwhelmingmale dominanceof the associationsis clear, but the most active groups,both within the associations(now unionsof cooperatives)and more generallyin rural society, are the numerouscooperatives, many of whichare women's cooperatives.
3.19 The recent restructuringof the associationsto create unionsof cooperativeshas tended to open up the membershipof the bureauxto some degree. One bureau has four female members (one of whom is black) and one other black memberto represent the agriculturistsin the area, another (one of the most active and progressiveof all the associations)has five Harratin and five Peuls.
3.20 The averageage of those presidentsfor whom informationis availablewas 51, but age varied from 35 to 70. The averageage of members of the bureauxwas generallysomewhat lower but varied considerablyfrom one associationto another,ranging from 45 to 56. Generally,the range of ages within the groups was much lower than in the membershipas a whole. Those who were youngerwere usually teachersor public employees.
Pastoral Associations and the Local Notability
3.21 Analysisof the occupationsof those holding positionsof responsibilityin the associationsshows that those people are usuallypersons of some importance.Further, they have both urban and rural interests and in sectorsother than livestock.Out of 36 presidents for whom 12
such detailsare available,9 (25 percent)had a residencein Nouakchottas well as locally,while those living locally in almostevery case had their residencein town. Sixteen(44 percent)had other occupationsbesides that livestockowner; twelve (33 percent)were both livestockowners and merchants.Two were members of the NationalAssembly and two were mayors (one, in the associationthat includedone of the two pilot pastoral perimeters,was both).
3.22 Relativelyfew members of bureauxhave residencesoutside the immediatelocality (90 percent lived locally).Twenty-two declared themselves to be only livestockowners. Fourteen were either public employees(usually teachers) or retiredpublic employees.Fourteen combined livestockproduction with commerce.Ten combinedlivestock production with agriculture.Two were farners and one was a livestocktrader.
The Nature ofRepresentation
3.23 The principle of one delegatefor each group of some 30 families has been retained, as has been the principle of an annual generalassembly of delegates(and one every two years to reelect the members of the bureau). In none of the 13 pastoral associationsinvestigated, however,had there been a generalassembly or meetingwhere the majorityof delegateswas present since the foundingmeeting. The restructuredUnions of Mederdra,Male, Barkeol-Est, and Barkeol-Ouest,however, had held generalassemblies to reelect the members of their bureaux,although the membershiptended to remain virtuallythe same after the elections.
3.24 Meetingsof the members of the bureaux(whole membership ranged from 7 to 16) were also extremelyrare. Only the managementcommittee-officially consistingof the president, secretary-general,treasurer, and those officersresponsible for legal matters, monitoring,training, purchaseand distributionof medicinesand inputs, and accountingbut rarely includingall of these-managed to meet reasonablyfrequently in most cases.Most of these meetingswere ad hoc. Usually the permanentcommittee (the president,secretary-general, and treasurer),or even just the president,makes the decisions.
3.25 In a few cases, an associationhas a small office that serves as its headquartersand is availablefor meetings.More commonly,meetings convene at the home of the president or in the offices of the Ministryof Rural Development.With a few exceptions,meetings of the bureaux are not subject to formal procedures.
3.26 From this it is possibleto concludethat the associationsare not genuinelyparticipatory structuresand largelyreflect the prioritiesand interests of a limitednumber of notables-the members of the local elite. This is not to say that the actions of the representativesdo not in some sense serve the commoninterest, even if they mostly benefitthose who are most involved. Certainlythe establishmentof veterinarypharmacies and several other local projects attests to the enterpriseof some associations(particularly their leaders),but the extent to which a wide range of local livestockowners benefit from these activitiesis difficultto assess.Most of the evidence suggeststhat the number who have benefitedis relativelysmall, even among associationmembers, while the impact on the local livestocksector of the activitiesundertaken by the associationsis also likely to have been limited. 13
Activitiesof the Associations
Revolving Funds
3.27 The item that was alwaysmentioned first, when the subjectof activitiesand impact was raised in interviews,was the revolvingfund. Giventhat this required a significantresource commitmentby the associationto reach the sum requiredby the project (500,000 um, about US$6,000in the early 1990s)to provide the full matchingfunds (1 million um, US$12,000),the issue of how the initial investmentwas raised and who benefitedis very important.In one case the president claimed to have raised the amountrequired himself. Apart from two cases10 all the associationsinvestigated had benefitedfrom a revolving fund of some kind for the purchase of veterinarymedicines and drugs and for animal feed (see Annex 2, Table 2).
3.28 Virtuallyall the associationsused the revolving fund for the purchaseand distributionof veterinarymedicines. Two used the fund for animal feed. Nine used it for both purposes.The majority of those associationsusing the fund for animal feed were able to amass a local contributionof 500,000um or more. The majorityof those using it for medicinesamassed less than 500,000um (see Annex 2, Table 2). This stronglysuggests that priority is given to veterinarymedicines. Given the lack of transparency,supervision, and control, and the highly personalizedway in which these operationswere managed,it is not surprisingthat instances of fraud and embezzlementassociated with the revolvingfund have been reported."
3.29 Several associationscontinue to operatesmall pharmacies.Nowhere, though, was it possibleto find an associationwith a permanentstock of animal feed. The stocks acquiredby the associationsare sold, often on credit, and are not renewedunless required. Reasonableaccess to veterinarydrugs and medicines,for example,undoubtedly improves the health of the local livestockand thus tends to increaseproductivity. But the specificcontribution of the project, even in this area-the most commonand most successfulactivity undertaken by the PCAs,is difficult to quantify,given the relativelylow usage of the stocks availablethrough the PCAs and the fact that similarproducts are availablethrough livestock services and the private sector.
Miniprojects
3.30 All but two of the associationsinvestigated had initiatedminiprojects of some kind with funds from the project. These were intendedto be financedfrom the ProductionImprovement Fund'2 (PIF) but this did not materializeuntil the Norwegianfunding was made availablelate in the project's life. The projects undertaken usually involvedmilk production and fattening, essentially for the local market.Here, as in the case of the pharmaciesand stocks of animal feed, the lack of managementtraining and experiencecombined with the lack of transparencyhas tended to limit severelythe effectivenessof these enterprises.
10. One of which had not raised the requiredsum in time to receiveany counterpartfunds, and anotherwhere internal conflicthad blockedthe raisingof funds. 11. The uneven nature of financialrecords means that it is not possibleto obtain consistentor currentdata on the financialstatus of most of these organizations. 12. The PIF was to have been financedby profitsgenerated by a parastatal,the LivestockInput Service(BIE) which importedand sold veterinarysupplies and animalfeed. However,this serviceof fundsnever materialized. 14
3.31 With the diversificationand integrationencouraged by the restructuringof 1993,new kinds of enterprise(besides those related to milk and meat production)such as a poultry project and projects for the marketingof livestock,regenerating Acacia Senegal, and exploitingsalt pans, have been proposedby the PCAs.These proposals have, however,failed to attract funding.
3.32 Despite some considerationof these kinds of enterprisesat the time of restructuring,the SecondLivestock Project did not see fit to support such a diversificationand remained focused on milk productionand livestockfattening. Even if, perhapscorrectly, it was felt that the PCAs were primarily associationsof livestockproducers that shouldconcentrate on livestock-related activities,there were otherpossibilities. One PCA, for example,has constructeda vaccination pen. But the projectnever soughtto involvethe PCAsdirectly in livestockinfrastructural developmentin ways that might have generateda greater sense of involvementand ownership. Nor was there ever any effort to link the programof well rehabilitationand reconstructionand maintenancedirectly under the auspicesof the pastoral associations.
Pastoral Wells
3.33 The issue of the drillingof boreholesand the rehabilitationand constructionof pastoral wells illustrates one of the major problems of this type of project.The SARincluded the "maintenanceand improvementof wells" (para 3.31) as one of the interventionsthat could be fundedby grants from the ProductionImprovement Fund. 13 However,although this was a top priority for pastoralists,it was not given high priority by the project when funds finally materialized.In the associationsinvestigated only four boreholeshad been drilled.No well rehabilitationwas undertakenat all, althoughnumerous sites had been identified in collaboration with the PCAs.Provision of water was a matter of great interestand a motivationfor the establishmentof PCAs. This failure of the project to fulfill any of the expectationsraised by the early surveysof wells to be rehabilitatedor reconstructedwas deeply felt by all the associations investigated.The reasons for the lack of funds may have been quite reasonablein bureaucratic terms, but the details and uncertaintiesof potential fundingare likely to be lost on the potential beneficiariesof a highly valued program.The impact on the grassrootsperception of the project was extremelynegative.
NaturalResource Management
3.34 In the absence of a program to encouragethe PCAsto develop users' groups and maintenancegroups aroundrehabilitated and reconstructedwells, the major issue of the Second LivestockProject has been (at least at the level of rhetoric) "forage and forestry"-encouraging the cutting, stocking,and sale of hay; preventingbrush fires and inhibitingthe cutting of trees; and controllingaccess to pasturesto prevent overgrazing.
3.35 But the first of these shouldbe considereda miniprojectbecause it relates to better use of local resourcesto increaseanimal production and incomes.As for the protectionof nature, the project did little to support activitiesregarding two of its major concerns.Early in their inception the associationsexpressed strong intent to act to reduce bush fires and prevent tree cutting,but today all agree that resourcesare needed if there is to be effective action,14and they tend to look
13.The beneficiarieswere to cover30 percentof the cost in cash or kind. 14. To emnployand deploypersonnel and to make firebreaks. 15 to the local authoritiesto provide what is needed. In one case the associationhad nominatedone person in each village to be responsiblefor the preventionof brush fires and to prevent tree cutting-although how those personsare expectedto achieveconcrete results is not clear.
3.36 The project certainlypressed the Mauritanianauthorities to provide the legalbacking that would enable PCAsto control access to water points and to pasturesto prevent overgrazing and land degradationand to stop unnecessarytree cutting.The response (whichwas the August 1990 circular to all local authorities [para 3.3]) appearsnot have been effective.In any case, the wholerelationship between the PCAs as cooperativesor unionsof cooperatives(and as interest groups or professionalgroups) and as entities with a territorialjurisdiction or remit remains problematic.
3.37 Few of the associationsinvestigated had taken part in the effort mountedby the project to encourageand supportthe cutting of hay for stockingand for sale by the PCAs. Only one had fenced in a pastoral perimeter with the aim of producingand stockinghay for the dry season.
Consciousness Raising
3.38 Without other concreteactivities and for lack of resources,most associationsemphasize their value in providinga frameworkfor consciousnessraising and disseminatinginformation on such key issues as improvedresource management, more cost-effectivelivestock production, and better marketing of livestockproducts.
3.39 There can be little doubt that the associationsare able to some extent to act, at the local, regional, and national levels, as a lobby that keeps the problemsof livestockproducers and of the livestock sector alive in the planning and policymaking of governmentagents and agenciesat all levels, as well as in the minds of NGOs and lending agencies.In all regions the Regional Delegationof the MDRE is aware of the existenceof the associationsand the pressure exertedby their representatives;in most cases an effort is made to keep them at least informed of current plans and programs.Many representativesargue that the associationsalso serve to provide ordinary livestockproducers with relevantinformation from "outside"and "above."
Training
3.40 While this may be the case, the more formal process wherebyinformation is passed on, and consciousnessraised, which involvestraining and education,has been limited in scope and value. The trainingof encadreurs(recruited from LivestockServices) produced a limitednumber of trainedpersonnel who the project initiallyassigned to the associations.But they were always few and over the years their superiorstended to move them on. About half the associations investigatedhad benefited from having an encadreur(usually an extension agent or veterinary nurse) for some time.
3.41 Trainingat the level of the associationsthemselves has been limitedbut of some value. Sessionslasted anywherefrom 15 to 45 days and focusedprimarily on animalhealth aimingto creating a cadre of local veterinary auxiliaries.Among the PCAs investigated,93 persons had received such training,but these belongedto only six PCAs.
3.42 The project had providedno managementtraining of any kind, whetherfor the running of a pharmacy,the identificationand design of a miniproject,or even the managementof a PCA. 16
The project provided no technicaltraining either-although at one point the production systems componentencouraged local livestockproducers and associationsto collect informationon their livestockproduction systems.
Representation of Herders 'Interests
3.43 The PCA was originallyconceived as an innovativeinstitutional form to protect and advancethe interests of livestockproducers, providing a frameworkfor the organizationof a variety of activitiesthat it was hoped would improvetheir lives and livelihoodsand at the same time maintain and enhancethe local resourcebase. To some extent they are able to do this, but the field surveys found that the activitiesundertaken by local PCAs dependheavily on the enthusiasmand resourcesof the association'srepresentatives, particularly their presidents.
3.44 The PCAs createdby the project, some mightargue, have becomevery much the creaturesof their presidents,even if these may have the perceived interestsof their members at heart. Certainlythe active involvementof ordinarymembers is lacking,and many livestock herders are not even members of an association.Of the 41 herdersinterviewed for this study, only 5 had heard of the PCAs or had any contactwith them. Roughlyhalf of those interviewed were not owners of the livestockthey were herding(and were hired for the task) and were relatively indifferentto increasingproductivity or marketingthe animalsin their charge.'5 The remainder,who were generallysmaller livestock owners, tended to considersuch organizationsa matter for townsfolkor "importantpeople."
Pastoral Associations and Emergent Livestock Production Systems
3.45 As mentionedearlier, the proportionof the rural populationthat is pure pastoralists (nomads)has declined significantlyover the past 25 years, from an estimated65 percent in 1965 to below 20 percent (some sourcessay 10 to 12 percent)by 1995. This does not mean, however, that livestockowners no longer have a considerablevested interest in the managementof their herds and in the managementof the natural resourceson which the animals depend.It does mean that livestockproduction increasingly has moved from a predominantlysubsistence-based sector to one more closely linked to the dynamicsof capital accumulationand to the market.
3.46 The emergingsocio-economic (class) divisionswithin the livestocksector raise importantquestions about the capacity of pastoral associationsto operatein the interests of all their members. The move to establishthem as unionsof cooperativesdoes, however,hold out somehope that within these generalstructures more tightly knit and smaller groups with strong commoninterests (for example,of smalllivestock owners) might be able to emerge.
15. The herdingof livestockincreasingly is undertakenby hired herdersor throughcontracts between livestock owners involvedin other economicactivities and thoseremaining directly involved in livestockand naturalresource management.The separationof ownershipand managementhas majorimplications for assessingthe impactof external interventionsin this area. 17
4. Impact of the PastoralAssociations on the Livestock Sector
The Impact of the Presenceof the Project
4.1 Althoughthe outcomeof the SecondLivestock Project did not meet expectations,its existence had consequencesfor developmentof the most importantelement of Mauritania's agriculturalsector. There can be no doubtthat (i) the process of projectpreparation itself had an influence (althoughdifficult to measure) on governmentpolicy with respect to the livestock sector and that (ii) the difficultnegotiations around the terms of reference and procedures required for project implementationthemselves contributed to some significantdevelopments in legislationand administration.
4.2 Governmentofficials and leaders in the livestocksector indicatedthat the constant pressure from the Bank, throughthe discussionsaround the shapeand contentof the project, for privatization,decentralization, and accordingmore responsibilityfor developmentactivities to those directly concerned,undoubtedly helped change the policy environment.16Above all, that the Bank, togetherwith other donors, was committedto puttingsubstantial resources into the livestocksector with the aim of increasingproduction and, at the same time, improvingnatural resource managementand the protection of the environment,gave this sector a high priority.
4.3 The livestocksector continuedthrough the project period to play a major part in the national economy, althoughits contributionwas perhapsless visible than that of agriculture. Factors contributingto reduced visibility may have been the dramatic"losses" in successive droughts, the mobility of livestock,the significanceof unofficialcross-border movements, and sales of livestockand animalproducts. There has been, however,a continuingtendency, by both governmentand external lending agencies,to privilege the agriculturalsector, particularlythe irrigatedsubsector. The extensionof the projectto 1995, and further two-yearsupport provided by other funders, has servedas a valuablecounterbalance to this tendency.The creation of the PCAs and of the national agro-pastoralgroup (GNAP)has also supportedthis articulationof a major interest group within the national economy.
The Nature of the PCAs
4.4 As of the middle of 1997 the PCAswere somewhatamorphous organizations that had been banded together into the GNAP.At the field level, their operationswere rather spasmodic, dependinggreatly upon the effortsof their leading figures.The extent of the benefits receivedby the "ordinary"members is not clear. They may receive some improvedaccess to drugs for their herds and, whereminiprojects are operational,be able to benefit frormthese. Perhapsthe most telling features of the organizationsare their low visibility,even in their own areas of operation, and their limitedimpact on the livestocksector. Given the emphasison this aspect of the project, why are the organizationsstill strugglingto get a footholdin the Mauritanianlivestock industry,
16. Particularly during the period of the Program for Economic and Financial Adjustment (PREF), 1985-88, and of the Program of Consolidation and Renewal (PCR) 1989-91. 18 and is there a future for them? The foregoinganalysis of the PCAs suggeststhat a number of factors may have been involved.
4.5 A major constraint,it might be argued,was the character of the PCAs themselves. Althoughfor many these increasinglybecame the central and crucial componentin the project- as the livestockinfrastructure development failed to take place as planned and the Office of LivestockInputs (BIE) failedto operateeffectively, and as the whole emphasismoved away from strengtheningthe governmentLivestock Services toward supportingprivate and cooperativesector activitiesand institutions-they never became effective grassroots organizationswith significantmass participation.Nor did those representingthe mass of the livestockproducers receive the kind of support required (in training,financial and managerial support, and other kinds of assistancefrom the project)for them to becomethe managersof cooperativeenterprises effectively serving the interestsof ordinary livestockproducers- although in some associationsthe leadershiptended to see the associationsas vehicles for providing a serviceto the local livestockproducers, their members.
4.6 Some vague notion of economiesof scale seemsto have led to the identificationand establishmentinitially of very large associationsextending over substantialterritories (including a variety of heterogeneoussocial groupsand vested interests).But this proved at variancewith the idea of collective,or at least cooperative,activity by all members, throughtheir delegates and their executivecommittee. Although there is ample evidencethat many PCAs undertooka variety of activities (miniprojects)during the project's lifetime,often using the revolving fund to do so, most of the activitiesundertaken involved only a few local notables and leaders in the local community.
4.7 Not only did most of the PCAs identifiedhave large memberships,they were also associatedwith a relativelylarge defined territory.The attempt to combinesocially defined groups with populationsinhabiting a given territoryinevitably created confusionand difficulty. This is particularlythe case in those regions closer to the SenegalRiver where agriculturists, agro-pastoralists,and pastoralists all exist in various combinationsand forms of cohabitation.
PCAs' Inabilityto EstablishLegal Rights over Resources
4.8 Effortsby the projectto create a new legal basis for allocatingrights over territoryand naturalresource managementwithin a given territory to these project-createdassociations provedlargely fruitless. Governmentpolicy duringthe early and mid-1980s ran directlycounter to this idea of fragmentedlocal rights, based on a combinationof customaryrights and an assertionby newly constitutedlocal entities of rights within a territorydefined by the project as theirs. Only in August 1990 (midtermin the project)did the government,under heavy pressure from the Bank, send a circularto all local authoritiestelling them to permit associationsto lease what had been declared state resourcesin 1983. The governmentcircular, however,was rarely applied.Even by 1992,the GNAPremained concerned about the lack of an effective move to assign rights over natural resourcesto pastoral associations.
4.9 In any case, often followingthe generalsouthward shift of populationafter the successivedroughts of the 1970sand 1980sthe livestockproducers belonging to the project PCAs sharedthe same territorywith other livestockproducers (who were not members of the association)as well as with agriculturists.Under those circumstances,attributing to these associationssole or even prime rights over key resourcesin the rural economywould have 19
proved extremelycontentious and difficultto apply.An alternativemight have been to work more closely with the local councilsand administrationsto developa flexible frameworkfor attributingrights over, and responsibilitiesfor, naturalresources by combinationof negotiation, consensus,and arbitration.This is a criticalissue that the PGRNPwill have to deal with extremely carefully.It appears that the latter projectwill take the rural communeas the basic entity within whichto work and try to construct village groupingswithin this broad administrativeframework. This might have been appropriatefor the SecondLivestock Project from the outset. There is certainly evidencethat the initial identificationof PCAs considered such administrativeboundaries.
4.10 In the absence of collectiverights overresources within a given territory, the PCAs became, in effect, a new kind of local corporateinstitution. 17 Thisnow representeda variety of specific interests (usuallybut not exclusivelyassociated with livestockproduction) and was constitutedas a cooperative:that is, a union of cooperatives.As such they are able to undertakea range of activities(usually fairly limited)broadly linked with the interestsof livestock producers,at the local level, and as livestockproducers, to have a voice, throughthe National Federationor associationof unions(GNAP) at the nationallevel. In the longer term this may prove a good thing. For the initial groundingof these sociallydelimited "traditional" rights to water and pasture overlap a set of more recent claimsto land and other resources.In some instancesthe problem involvesthe existenceof differentsocial groups within the defined territory (not all of whom are members).In other cases the problem occurs when an association wishes to undertakeactivities outside its defined territory,as when one associationwishes to establish a milk processingplant in a nearby town that is outside its perceivedjurisdiction.
PCAs' Dependenceon ExternalSupport
4.11 The emergenceof PCAs as the basis for a distinctivelocal, regional,and national lobby on behalf of livestockproducers is a major achievementof the project. The GNAP,however, which is essentiallya product of the SecondLivestock Project, has extremelylimited resources and is barely able to function.The idea that it could function,not only as a national association for livestockproducers but also as an institutioncapable of purchasingand distributinglivestock inputs to its memberassociations and individualsacross the country, seemsfanciful without more substantialresources. The project seemsnever to have made the necessarycommitment to support effectivelythe institutionsit had called forth. The Norwegiangrant from 1993 onward helped, but the overalllevel of fundingin support of the PCAsremained extremelylimited throughout-about 10 percent of total expenditures.
4.12 While at least in the initial stages considerablecare (and time) was taken to identify, explain, and provide some support for the new institutions,once it was decidedto expand the number of associationsrapidly, the time taken was reducedand PCAswere establishedfar too rapidly to be effective.Direct support was limitedto the revolving funds and to a program that trained a smallnumber of livestockagents to be deployedin support of the associationsas trainers (encadreurs)of the local associationrepresentatives (and members).This program experiencedconsiderable difficulties and few encadreurswere trained. Sometrainers were never
17. This was particularlythe case afterthe introductionof the law of January1993 amplifyingand alteringthat of 1967on the status of cooperatives. 18. This was once proposedby the project. 20 assigned to the PCAs and others were moved soon after being assigned (for they remained essentially government employees). Few associations benefited from systematic support of this kind and none received more than a perfunctory training in business or organization management or any other relevant skills. Efforts after 1993 to restructure and revitalize the local associations revealed what regular visits and encouragement by the project could accomplish. The cessation of project funding in 1995 meant that many associations, while established on paper, were not effective as development institutions (even within the narrow limits established and encouraged by the project) and were certainly not sustainable.
The LimitedNature of Servicesto Members
4.13 The only area in which the project appearsto have generatedsomething like a lasting and sustainableactivity is the purchase and sale of veterinarydrugs and medicines.Despite the weaknesses of the BIE, the lack of government quality control,19 and competition from both Livestock Services and the private sector, the majority of associations had been involved in establishing small veterinary pharmacies20and selling veterinary products to local livestock producers. Often this activity had proved capable of maintaining the revolving fund over several years, and the practice continues to this day.
4.14 Without continuing support from government agents or pressure from an active membership, however, the pastoral associations' involvement with improving animal health remains limited. Often the president of the association runs the pharmacy. He keeps a limited stock and sells to a small number of livestock producers. Here the impact is also limited. A number of locals, trained as health auxiliaries through the project, continue to provide a service, but the extent of this also appears to be limited. The PCAs evidently value the project-supported vaccinations by the Livestock Services, but the associations are not significantly involved in the provision of this service.
4.15 Other contributions of the project in the domain of animal health, such as the construction of vaccination parks and buildings for veterinary centers, had some value. But these were conceived as separate components and were not linked into a systematic institutional structure (either that of the PCAs or that of the Livestock Services) in a way that maximized their usefulness and ensured their maintenance.
4.16 The desire to increase the number of associations established compromised the consolidation of the earlier associations to some extent, but it is remarkable how much those associationsthat appearedto have started well also appear to have maintainedtheir vitality. However,insufficient attention (and insufficientresources) was devotedto evaluatingand increasingthe effectivenessof institution-buildingactivities for the PCAsand to assessingwhat determinedpastoral associationeffectiveness. This limitedthe capacity of the project to evaluate the effectiveness of the PCAs as a basis for sustainable development. Perhaps critically-given the centrality of the pastoral associations to the project-insufficient preparation and support for these new creationsthroughout the project largely limitedtheir capacityto promote a set of
19. A substantialpercentage of drugsand medicinespurchased by the associationswere perishedand couldnot be sold. 20. The project providedtraining for localsto managethese and to prescribemedicines as health auxiliaries. 21 associatedactivities within the livestocksector, and limitedtheir impact on the lives and livelihoodsof their members. 22
5. Conclusionsand Implicationsfor the Future
5.1 Over the past 15 years externalsupport to the livestocksector in semi-aridnorth and west Africa has increasinglyemphasized the developmentof producers' groups,both to help improvemanagement of rangelandand water resourceson which pastoral productionis based, and to facilitate links betweenpastoralists and support services.The SecondLivestock Project in Mauritaniawas one of the first Bank-supportedefforts in this area. This impact evaluationis the first attempt to assess the effectivenessof the producers' groups the project developed.As such it is as a case study of the issues faced in supportingdevelopment largely through local organizations,particularly in marginaleconomic conditions such as those found in semi-arid regions of developingcountries. The previoustwo chapters have indicatedthe nature of the difficultiesfaced in attemptingto establisha new type of local organizationwithin a traditional societal framework.This chapter discussesthe implicationsof the findings,both specificallyfor the succeedingIDA-funded project in Mauritania,the RainfedNatural ResourceManagement Project (PGRNP),and more broadly.
Performanceof the SecondLivestock Project in DevelopingLocal Institutions
5.2 Althoughthe SecondLivestock Project failed to reach most of its original stated objectives,there now exists, albeit in a relativelyprecarious state, a national groupingof pastoralistsand other livestockproducers, and a number of local structures.The 39 associations (unions) formedunder the projecthave an estimated20,000 or more members,or about one-third or more of the pastoralists(or perhaps 15percent of the livestock-owninghouseholds). This is a considerableorganizational achievement in a short time. These could, with appropriatesupport from outside, providethe institutionalframework for a networkof livestockproducers' cooperativeunions across the countryand the basis for an effectivelocal, regional,and national lobbyingeffort, supportingthe interests of those livingdirectly from pastoral and livestock activities. There is, however,a sense of frustrationwithin the PCAsthat they were not able to accomplishmore of a concretenature at the local level, in particularlyreflecting the constraints on the scope of miniprojectsfunded by the project, and the continuingproblem of the legal rights of groups to control natural resources.
5.3 A series of problemsin the designand implementationof the SecondLivestock Project impingedon its efforts to developpastoral associations:
* Sufficientresources and support for the new institutionswere not forthcoming. * The rhetoric of participationand "bottom up" institutionaldevelopment was not matchedby the overall structureand organizationof the project in implementation. * The approachof the project towardsthe PCAs was inconsistentover time (moving from.theidea of an experimentalapproach with a limitednumber of "pilot" associations,to the idea of expansion,and then back to consolidation). * The large size and scale of the PCAs identifiedand establishedmade it difficultfor them to overcomeinternal heterogeneityand to act as genuine collectivitiesor cooperatives,as hoped. Also, the confusionof their role as socio-economic associationsand their role as territorialassociations inhibited their effective development. 23
* The project continuedto place primary emphasison range managementeven though the prime concern of the PCA memberswas with water and animal health, and despite evidencethat the effortsto establisha freer systemof distributionof veterinaryproducts were appreciatedby the herders and proved the most popular and durable of all the initiatives. * The primary concernof the project with natural resourcemanagement was not linked to a strategy based on a detailed understandingand analysis of the complex dynamicsof contemporaryproduction systems and an appreciationof the major environmentalconstraints in a rapidly evolving ecologicalsituation.
5.4 In assessingthe impact of the project's efforts in this area, one issue is what would have happened in the absence of the project initiative,that is, the "counterfactual."This is difficult to assess because the life of the project was characterizedby considerablepolitical, social, and economic change in Mauritania,which openedup the arena for politicaland related social activity. This mighthave led to an effort by pastoraliststo form somenational groupingbut, given the observedtendency for the PCAsto languishwithout support,this would appear unlikely.Without the efforts of the SecondLivestock Project, therefore,there would now not be any frameworkupon which efforts to mobilize pastoralistscould be based.
Linkage with Rainfed Natural Resource Management Project
5.5 As a result of the problemsnoted above,the groupsformed under the SecondLivestock Project are in a situationof flux and uncertainty. The fundingprovided after the closure of IDA funding for the SecondLivestock Project was implicitly,if not explicitly,to maintain the PCAs and the national GNAP/GNAASPstructure until it could be incorporatedinto the implementationof the PGRNP. However,now that the projecthas been approvedand implementationis starting, it is not clear that this has happened. The PGRNPAppraisal Report refers to the PCAs,but notes that they, and other rural associations,have not been "overly effective" and that "village investmentscarried out by now-completedprojects have shown a poor level of sustainability"(para. 1.22). However,the appraisalreport is not explicit about what role the 39 PCAsand their local sub-unitswill have under the new project.
5.6 The major focus of the PGRNPis at the village level. US$16million, or 68 percent of the project's total estimatedcost of US$23.5million is to be spent on village-levelinvestments. Three major types are envisaged:
* land and water conservationactivities, including water harvesting techniques; preservationof soil fertility,erosion prevention, and control of sand dune encroachment;and biomassproduction on rangelandsand fuelwoodproduction; * water supply and productiveinvestments, including civil works, mechanical equipmentand small agriculturalequipment; and * micro-enterprisematching grants for activitiesalong the lines of the earlier miniprojects,such as cooperativeor individuallivestock fattening and grain storage.
Supportfor human and animal medicine supply,as fundedby the existingPCA revolving funds, is also envisaged. 24
5.7 It is expected that US$4 million will be spent on institutional support and capacity building in support of this program. This will include community organization, establishing the legal framework "required to ensure reasonably secure access to common resources," (para. 2.15) and technical support.
5.8 This program clearly overlaps with activities promoted by the Second Livestock Project. However, the impact evaluation team found little evidence of PCA engagement in the PGRNP's preparation phase over the past two years. At the local level, project activities are to be implemented by the village management committee (CVGRN). These committees are to prepare (with project support) natural resource management action plans. Given that the focus is on a defined area, this inevitably will emphasize measures aimed at increasing the capacity for cropping (such as soil and water conservation), rather than supporting livestock production. The participatory approach appears not to have been adopted so far -at least regarding the PCAs and their involvement.
5.9 The evaluation team was impressed that pastoralists are clearly worried that the PGRNP village-based model appears likely to squeeze them out and limit their access rights to key resources. It is crucial that this not be the case. At least one group of pastoralists interviewed emphasized the need (given the history of social conflict in areas now increasingly under pressure from the concentration of human and animal populations) to develop institutional forms for cooperation and collaboration between all those living in a given area-herders and farmers, transhumant and sedentary livestock producers, and people of different ethnic and tribal origins. They felt strongly that ignoring or excluding the pastoralists and their associations because they did not fit easily into the village territory framework, would risk increasing tensions and reducing the possibilities of genuinely cooperative natural resource management.
5.10 The same issue was raised in the earlier group audit of three Sahelian livestock projects,22 in discussing the interaction between pastoral and agricultural production systems. "But one consequence of livestock integration within a farming system is a reduction in the scope for interaction between farming systems. Those who are able to benefit from integration will not be the same population as those no longer able to benefit from interaction, unless new forms of interaction can be developed. At the present time it is not clear that what increasingly looks like a zero-sum game can be avoided, in which the decline of the extensive, transhumant pastoral system is assured" (para. 3.21).
5.11 If the PGRNP is to reduce rather than increase intracommunity tensions, the project must address from the outset the issue of its relationship with GNAASP and its local units, as well as their links with the local authorities. Recent experience in efforts to foster community-based management of natural resources has pointed to the importance of these links and the value of creating structures including both community and agency bodies if efforts are to be scaled up beyond the very local level.23 The GNAASP could help provide an informal communication
21. Althougha few experimentalstudies using participatoryrural appraisaltechniques have been carriedout. 22. OED (1994)Performance Audit report on: Republicof Niger-Livestock Project(Cr. 885-NIR); CentralAfrican Republic-Livestock DevelopmentProject (Cr. 894-CA);and Republicof Senegal-Eastern SenegalRural DevelopmentProject (Cr. 1406-SE).
23. John Farringtonand CharlotteBoyd, 1997 "Scalingup the ParticipatoryManagement of CommonPool Resources,"Development Policy ReviewVol. 15,pp. 371-391. 25 channelbetween the project and participatingvillages, when they are not activelyundertaking a miniproject,as part of just this type of framework.Criteria for selectionof participatingvillages mightalso be framed to ensurethat, if there is a local unit of GNAASP,its involvementcan be guaranteed,thereby minimizinglocal frictions.
5.12 The recent Bankreport on Investingin Pastoralism24discusses the difficultiesof developingherder organizationsand notes that differenttypes of organizationsare best suited to different tasks. As noted above, one of the major problemswith the Second LivestockProject was that it attemptedto undertaketoo many differenttypes of activities.The project emphasized two potentiallyconflicting objectives of producerorganizations, that they would be both (i) cooperativeorganizations (within which smallercooperative organizations could form and be active) and (ii) territorialinstitutions that shouldhave jurisdiction(perhaps even exclusive rights) over water and pasture within "their" territory.
5.13 This two-prongedapproach confusedtwo things be best kept apart. Particularlyin the southernpart of Mauritaniathe concentrationof populationthat has taken place in response to the droughts (and to someextent the developmentof infrastructure)has made exclusivewater and grazing rights problematic.Leasing by PCAsof specificresources or "areas" from the new local authorities(such as the rural communes)might be a way of dealingwith the "control and managementof naturalresources" under the new circumstances.If so, the idea of village-based entities (with clearlydefined territory)built into the PGRNPmight be appropriateif (and only if) flexible mechanismsfor regulatingaccess to resourceswithin these territoriesby pastoralists, other livestockowners, agriculturalists, and others are established. It is not clear that the foundationfor such an arrangementhas yet been laid.
BroaderImplications for PastoralInterventions
5.14 In discussionof the initialresults of this evaluationa number of genericquestions have arisen related to the implicationsof the findingsfor pastoral and wider rural development activitiessupported by the Bank. These revolve around (a) the emphasis on the introductionof improvedrange managementtechnologies, (b) the lack of traditionalparticipatory institutions with which to work, (c) the sequencingof interventions,and (d) the optimal size of institution that shouldbe supported.
24. David John Pratt, Francois Le Gall and Comelis de Haan. 1997. Investing in Pastoralism, Sustainable Natural Resource Use in Arid Africa and the Middle East. World Bank Technical Paper No. 365. Chapter 7. 26
The lack of clear, proven range managementtechnologies
5.15 The justification for projects such as the SecondLivestock Project has frequentlyrested on the expectedbenefits from the introductionof improvedtechnologies. However,recent work25has concludedthat most existing,traditional systems in marginal environmentssuch as the MauritanianSahel are alreadymaking very efficientuse of the availableresources. The overwhelmingconstraint on attemptsto make a livingin these areas is that they are of inherently low productivity,primarily because of high temporaland spatialvariability in rainfall. While they have supportedpastoral and related societiesfor centuries,they resist modernization,except where reliable water resourceshave enabledthe developmentof irrigatedagriculture. Consequently,these semi-aridregions have becomeeconomically, socially, and politically marginalized. Furthermore,as developmentproceeds, the traditionalsystems are usually impingedupon by changesin the broader society and economy,and they become more difficult to sustaineven at their earlier levels of productivity. The major problem is how to maintain a livelihoodwhile copingwith high levels of climatic and economicuncertainty.
5.16 As elaborated in Living with Uncertainty(and more briefly in Investing in Pastoralism), the priority in assistance to the affectedpastoral groups shouldbe on measuresto enable them to integrate themselvesmore effectivelyinto the wider society and economy,to enable them to modify their livelihood systemsto fit the new realities. This includes actions such as improvementof trading systemsto increaseaccess to markets and increasethe ability to vary herd size in response to (or anticipationof) adverseclimatic conditions,and modificationof legal and administrativeframeworks to enablepastoral groups to clarify rights to key resources and to encourageuse of the latter in ways that best reconcilecompeting demands.
5.17 Althoughit was not put into this context,the SecondLivestock Project did take some steps in this direction, most obviouslyby creatingthe GNAP,which mighthave the potential to act as a lobbying group and bring the issues of most concernto the attention of policy makers (or interpreters)at the national and local levels. Unfortunately,the means it had to respond directly to local needs and priorities(the revolving funds and miniprojects)were very limitedand circumscribed. And, at the broader level, it is not clear that the PCAs developedthe ability to be effective as a lobbyingforce, since they were not able, until late in the life of the project, to induce a response in the provisionof wells, nor were they able to have themselvesformally includedas stakeholdersin the PGRNP.Therefore, it did not increasethe ability of pastoralist members of PCAsto cope with the uncertaintiesthey face.
The lack of traditionalparticipatory structures
5.18 The beneficiarystructures developed under the project were not participatoryin any real sense. Pastoral societiesin Mauritaniaare stronglyhierarchical. Therefore, any structures developedon the basis of the response from the existingsociety will similarlybe hierarchical, rather than fully participatory. There is a dilemmahere, as it is clear that a genuinecommitment to a full participatoryapproach impliesa very strong external involvementfrom the outset, includingsupport for monitoringand evaluationand, whereneeded, resourcesfor training. However,this type of approachmight have been seen as threateningby the existingstructures
25. See, for example,Roy Behnkeand Ian Scoones(1992), "Range Ecology at Disequilibrium,"World Bank EnvironmentalWorking Paper No 45. and Ian Scoones(ed.) (1994),"Living with Uncertainty,"London ITP, which discussthe managementof semi-aridlands and meansof improvingthe welfareand future prospectsof pastoralists. 27
and, thus, generatedan adversereaction. Perhaps a projectsuch as the SecondLivestock Project, with limitedTesources for support, shouldnot have claimedthat it could be participatory. It engaged the local leadershipand made effortsto distinguish"real" pastoralists from others,but found even this difficultbecause it did not have a deep knowledgeof local productionssystems (who owned what, who did what, etc.). What it has doneis to develop some local (districtand regional) bases for livestockproducers to federate, ally, and lobby at local and national levels.
5.19 However,the initial steps will themselvesbegin to cause some local social changes, includingthe developmentof alternativeleadership.26 This calls for the design of process, rather than blueprint projects that can adapt as implementationunfolds to accommodatesuch changes. This was clearly a weakness in the SecondLivestock Project as, instead of being responsiveto the different demandsof the individualPCAs, it decidedupon a standardizedpackage (the revolving fund for financingveterinary supplies and other inputs). The only area where innovationcould be manifestwas in miniprojects,but these also tended to be very similarand, when PCAs suggesteddifferent activities (such as poultry and exploitingsalt pans), the project tended not to support them. Smallventures of this type are likely to be the breeding ground for new societal leadership,but fosteringthem would require a substantialleap in the availabilityof resourcesto develop the appropriatecapacity in the implementingagency. In some circumstances(as in the Philippinesand Brazil) it may be feasibleto use NGOs to support such change,but this is not the case in Mauritania.
The optimal size of an association
5.20 The experiencewith the pastoral associationsillustrates the problem of matchingthe form of organizationsupported by a project to the type of activity it is envisaged it will undertake. In this case large groupswere identifiedto take advantageof economiesof scale, but this resulted in heterogeneousinterests within the groups. This approach also reflected the lack of a proven technical"package," which could form the basis of the project intervention,and which would imply a particular institutionalstructure. Experiencesuggests that in many pastoral areas the existing,limited resource base is effectivelyused, and that improvementsin incomes and welfare must be soughtthrough means other than physical intensificationof existing productiontechnologies. There is also a "chicken-and-egg"problem in that, if the local organizationsto be supportedare to be "bottom-up"and participatoryin form, then their priorities for action,and the appropriateinstitutional structure, will not fully emerge until implementationbegins. The SecondLivestock Project never resolvedthis dilemma. It attemptedto substitutethe leadershipof large PCAs for project inputs that could have provided direct support to smallerlocal groups, which wouldhave operatedat a scale more suited to efforts to address natural resourcemanagement issues. The PGRNPis primarilyemphasizing local (communal)groupings, and devoting 17 percent of projectresources to their support. It is relying on public agenciesand administrativestructures as the basis of this support, however, which may also compromisethe "participative"nature of the basic units.
26. See for example:Mary Tiffen, MichaelMortimore and FrancisGichuki, (1994) MorePeople, Less Erosion: EnvironmentalRecovery in Kenya,London, John Wileyand Sons.Chapter 9 discusseschanges in social structuresin MachakosDistrict, Kenya, related to changesin landresource rnanagement practices and in agriculturaldevelopment. 28
The needfor a sequencedor comprehensiveapproach
5.21 The experienceof the SecondLivestock Project indicatesthe need for a sequenced approach,especially where the local socio-economicand naturalresource managementsystems are not well known. The pastoral subsectorwould seem to be a prime candidatefor the use of the new Learningand InnovationLoans, which may be used to prepare the groundworkfor future activities,in particular to supportthe manpower-intensiveprocess of developinginternal networks and processesto necessaryto enablethe implementationof subsequentactivities. In particular, the comparativeadvantage of the Bank would appear to be in elevating this process to ensurehigh-level visibility, and thereby develop support for the proposed action. Clearlythe Bank's efforts to assist the livestocksector in Mauritaniahave not been very successfulin that direction. A relativelylow profile for effortsto developnew forms of local leadershipmay have been appropriatewhen the SecondLivestock Project was appraised-when governmentwas more autocraticand centralizing. However,indications from this and other projects indicate that it is better to start with a combinationof (a) investigation,and (b) specific actions that bring evidentresults in areas where there is a clear felt need (such as animalhealth or water resources).27This could be broadenedby allowingthose beneficiarygroups that successfully implementtheir first operationto have a widerrange of choicein the second.
27. This is essentiallythe approachsuggested by the FrenchCooperation Agency during the preparationphase of the SecondLivestock Project. 29
Annex 1
SECOND LIVESTOCK PROJECT
Credit no: 1658-MAU Approved:February 4, 1986 Total cost: US$18.1million Effective:June 30, 1987 WorldBank credit:US$7.6 million Closed:December 31, 1995 Cofinancing: Disbursementrecord: AfDB - US$6.3million US$2.3million canceled OPEC - US$2.0million Norway - US$2.2million Objectives: * Slow deterioration of grazing areas and boost livestock productivity by making groups of pastoralists responsible for allocated grazing areas. - Provide pastoralists with access to better extension services and production inputs. - Improveunderstanding of existingproduction systems to assistin planning future developments. Components: D Establishmentof pastoral associations(PCAs). - Strengtheningof livestockservices. - Establishmentof specialfunds to partially financemini projectsto be managedby PCAs,and to financerecruitment costs of field services. * Supplyof inputs for sale to livestockproducers. - Constructionof an abattoirin Nouakchott. - Project monitoringand the studyof livestockproduction systems. Outputs: * 39 PCAsestablished. * Physical facilitiesconstructed for livestockservices and equipmentand trainingprovided. * Special fundsnever generated. * Input and supplyof productioninputs through public sector unsatisfactory;importation now privatized. * Abattoir constructedbut not yet operational. * Studiescarried out.
31
Annex 2 Table 1
DATA ON ESTABLISHMENTAND OPERATIONOF PASTORALASSOCIATIONS (1)
Date of lVo.of No. of No. of No. of Region PCA Names Creation Localities Members Delegates Co-ops CurrentSituation
H. Charghi Bassiknou-Ouest 21.01.1991 9 441 18 3 Reactivated 10.12.94 Adel Bagrou 15.08.1989 21 746 23 2 " 15.12.94 AmoujEst 1990 907 6 " 22.12.94 Walata 09.10.1991 20 643 30 2 " 31.12.94 Nouaoudar 20.10.1991 10 291 18 23 " 06.01.95 DjigueniEst 28.06.1989 12 841 27 5 " 12.02.95 AoueinatZbil 23.10.1992 45 785 25 " 28.02.95 TimbedraEst 26.11.1989 65 441 65 I TimbedraQuest 27.09.1991 50 n.a. 66 n.a. Planned BassiknouEst 25.09.1991 10 n.a. 50 n.a. Planned
H. Gharbi KobeniEst 02.01.1989 65 948 32 5 Aioun Ouest 21.10.1991 60 581 24 5 Reactivated 22.05.95 TintaneOuest 29.08.1988 71 498 22 7 10.03.95 Aioun Est 25.06.1991 65 468 17 4 12.08.95 TintaneEst 27.09.1988 40 1278 45 3 " 16.08.95 330 24.08.95 29.09.94 Assaba Boumdeid 21.10.1991 24 254 20 3 30.10.95 KiffaNord 10.11.1991 25 10 Planned Kiffa Est 16.10.1991 49 542 59 n.a. " BarkeolEst 1989 " Barkeol Ouest 1990 55 n.a. 65 n.a.
Gorgol Tifonde Cive 31.12.1991 10 277 25 2 Lexeiba 01.01.1992 20 n.a. 25 1 Reactivated 16.06.94 Monguel 01.02.1990 65 451 16 4 02.10.94 Waly Diantang 05.06.1992 14 n.a. 20 n.a. Planned Source:National Groupof PastoralAssociations. (1) Status as of 1996. 32
Annex2 Table 1
Date of No. of No. of No. of No. of Region PCS Names Creation Localities Members Delegates Co-ops CurrentSituation
Brakna Boghe Ouest 28.08.1991 36 531 20 3 Reactivated 18.11.95 Mal 02.02.1992 20 662 25 3 " 24.10.95 Mederdra 30.03.1988 19 176 21 1 Reactivated 09.10.94 DjidrelMohguen 04.06.1992 10 31 14 4 " 08.07.95 Tekane 08.06.1992 39 42 16 5 " 12.07.95 Rosso-Nord 28.02.1998 45 1 10 3 " 15.07.95 Keur-Macene 02.02.1998 60 818 65 1 " 20.12.95
El Medah 28.07.1991 19 55 30 1 Planned
Tidjikja 08.01.1992 14 21 Gjpidoua 16.01.1992 16 24 Niemlane 21.01.1992 14 20
Ould Yenge 17.08.1991 30 171 13 2 Reactivated 08.06.94 Ajar Soninke 07.06.1992 20 612 21 1 " 12.06.94
Akjoujt 05.03.1993 50 200 65 Planned iOwilayas 39ACP 1418 17.157Ad. 1189 120Cop 33
Annex 2 Table2
PASTORALASSOCIATIONS REVOLVING FUNDS
Contribution by Pastoral Contribution by Years of Region Pastoral Association Associations (CFA mil) Project (CFA mil) Contribution Use of Fund
Trarza Rosso. Nord. 675.000 1.000.000 1990 VeterinaryMedicines and AnimalFeed Mederdra 812.000 1.000.000 1990 Jdes El Mohguiene 160.000 320.000 1992 Keur Macene 300.000 600.000 1992 VeterinaryMedicines Brauna Boghe-Ouest 448.302 896.604 1991 VeterinaryMedicines Gorgol Monguel 500.000 1.000.000 1990 VeterinaryMedicines and AnimalFeed TonfoundeCive 400.000 800.000 1982 Wali Dianta 223.000 446.000 1992 VeterinaryMedicines Lexciba 200.000 400.000 92-93 .. Assaba Barkcol- Est 500.000 1.000.000 1982 VeterinaryMedicines and AnimalFeed Barkiol-Ouest 330.000 660.000 90-93 VeterinaryMedicines Kiffa-Est 500.000 1.000.000 1992 Kiffa-Nord 500.000 1.000.000 1992 Boumdeid 200.000 400.000 92-93 VeterinaryMedicines and AnimalFeed HODH Timbedra-Est 500.000 1.000.000 1990 VeterinaryMedicines Timbedra-Ouest 500.000 1.000.000 1990 BassiknouOuest 500.000 1.000.000 1991 Mcgvc 220.000 440.000 1992 EL Amourj 350.000 700.000 1991 Oualata 200.000 400.000 1992 VeterinaryMedicines and AnimalFeed Chargui Ugiguieni 500.000 1.000.000 1991 VeterinaryMedicines AoueuatZbil 200.000 400.000 92-93
Guidimakha Ould Yeuge 370.000 740.000 1991-94 VeterinaryMedicines H. El Gharbi Tintanc-Ouest 500.000 1.000.000 1990 VeterinaryMedicines Tintane-Est 250.000 700.000 92-94 Aioun-Est 500.000 1.000.000 91-93 Aioun-Questx 150.000 300.000 91 Kobenni 100.000 200.000 92 Tagant Tidgikja 80.000 160.000 1992 VeterinaryMedicines Ninlane 500.000 1.000.000 1992 VeterinaryMedicines and AnimalFeed Ghoudia 500.000 1.000.000 1992 Animal Feed Adrar El Meddah(Amejeft) 500.000 1.000.000 1992 Animal Feed Inchiri Aujoujt 500.000 1.000.000 1993 VeterinaryMedicines
35
Annex 3 Comments from the Borrower
[English translation of the letter from the Borrower]
THE ISLAMIC REPUBLICOF MAURITANIA Honor-Fraternity-Justice
Ministryof Rural Developmentand the Enviironment
The SecretaryGeneral
Tel.: 515 00 Nouakchott, July 16, 1998 Fax: 574 75
No.: M.D.R.E/M 00000670
To: THE RESIDENTREPRESENTATIVE, WORLD BANK
Re: Comments on the Evaluation Report, Second Livestock Project
Ref.: Your letter of June 28, 1998 concerning the Report on Credit 1658-MAU
I have received the above Report, and congratulate the World Bank on the document.
Project performance is certainly mixed, but the Ministry is completely satisfied with the start that has been made in structuring rural activities by establishing 39 Agro-Sylvo-Pastoral Associations and the National Agro-Pastoral Grouping (GNAP), which has given clear evidence of its dynamism in promoting increased consultation between the Government and operators in rural areas.
In order to build upon these achievements (and as detailed in the attached note), the Ministry intends to increase the involvement of these groupings in the efforts to achieve the integrated management of natural resources, particularly through the introduction of the PGRNP (the Rainfed Natural Resource Management Project).
I trust that you will take account of these intentions on the part of the Ministry, since they are bound to have a positive impact on rural development.
Sincerely,
Is! Hayda Amadou Kane 36
Annex 3
THE ISLAMIC REPUBLIC OF MAURITANIA Honor-Fraternity-Justice
Ministry of Rural Development and the Environment
Directorate of Agro-Pastoral Resources
Note
Re: Participation of GNAP (the National Grouping of Agro-Sylvo-Pastoral Associations) in the Integrated Management of Natural Resources
The following comments relate to the rural-development objectives of the Second Livestock Project, and refer to the overall conclusions of the World Bank's Evaluation Report (see Annex 1, page 56):
1. The pastoral (and general rural) environment was successfully restructured (with the establishment of 39 Pastoral Associations).
2. The special funds were never generated.
3. Support for development (and therefore the success of miniprojects for restructuring the sector) has been marginal.
4. It was not possible to transfer responsibility for natural resource management to local level, particularly because constraints relating to land rights could not be eliminated.
Although it may appear disappointing that the establishment of the Pastoral Associations was the only item in the list of objectives to be achieved (the list being, perhaps, too ambitious), it is nonetheless a considerable achievement in the Mauritanian context, and one that the Government intends to maintain and strengthen, if possible.
Unfortunately, another aspect must be noted, in addition to this positive achievement: the failure to involve the Pastoral Associations in the process of preparing the PGRNP, an outcome that became all the more likely in 1994 when they became Agro-Sylvo-Pastoral Associations. This omission represents not only a considerable loss in itself, but is also very likely to lead to an aggravation of local conflicts, as noted in the Report (see pages 45-48). The pastoralists, who play a major role in rural areas, feel justifiably excluded from the process, which is certainly more concerned with agricultural issues, but which cannot disregard local constraints. 37
Annex 3
This latter issue jeopardizes the hope we expressed previously, and the Government fully intends to remedy the situation. To this end, all necessary measures will be adopted to involve the existing local organizations in the form of integrated management proposed under the PGRNP, and thus, to the extent possible, any unnecessary overlaps with competing local groupings will be avoided, so that the Pastoral Associations -- once established and operational -- will constitute the essential channel for local partnerships.
At the same time, the authorities will continue to involve GNAP in their ongoing efforts to structure stockraising activities, particularly with regard to the establishment of interdisciplinary organizations, which will exercise a very positive influence by increasing the professionalism of private- sector operators and transferring responsibility for activities to them.
39
References
Behnke, R. H. and I. Scoones. 1992. Rethinking Range Ecology: Implications for Rangeland Management in Africa" Environment Working Paper No. 53. World Bank. Washington, D. C.
Behnke, R. H., I. Scoones, and C. Kerven (eds). 1993. Range Ecology at Disequilibrium. London: Overseas Development Institute/International Institute for Environment and Development/Commonwealth Secretariat.
Cousins, Ben. 1992. Managing Communal Rangeland in Zimbabwe: Experiences and Lessons. Case Study prepared for Workshop on New Directions in African Range Management Policy, Matopos, Bulawayo, Zimbabwe. de Haan, Comelis. 1994. An Overview of the World Bank's Involvement in Pastoral Development. Pastoral Development Network Paper No. 36b. London: Overseas Development Institute.
Gilles, Jere and Cornelis de Haan. Recent Trends in World Bank Pastoral Development Projects: A Review of 13 Bank Projects in Light of the 'New Pastoral Ecology "' Pastoral Development Network Paper No. 36b. London: Overseas Development Institute.
English, John, Mary Tiffen, and Michael Mortimore. 1994. Land Resource Management in Machakos District, Kenya, 1930-1990. World Bank Environment Paper Number 5. Washington, D. C.
Maliki, Agelo. 1992. Pastoralists at the Crossroads: Survival and Development Issues in African Pastoralism. UNICEF/UNSO. Nairobi.
Overseas Development Group. 1998. Mauritania Second Livestock Project, Impact Evaluation Norwich, United Kingdom.
Pratt, David John, Francois Le Gall and Cornelis de Haan. 1997. Investing in Pastoralism, Sustainable Natural Resource Use in Arid Africa and the Middle East. World Bank Technical Paper No. 365. Washington, D. C.
Scoones, I. (ed). 1994. Living with Uncertainty. London: Intermediate Technology Publications.
Tiffen, Mary, Michael Mortimore and Francis Gichuki. 1994. More People, Less Erosion. Environmental Recovery in Kenya. London: John Wiley and Sons.
World Bank. 1994. Performance Audit Report: Republic of the Niger Livestock Project (Credit 885-NIR), Central African Republic Livestock Development Project (Credit 894-CA) and Republic of Senegal Eastern Senegal Rural Development Project (Credit 1406-SE). Operations Evaluation Department. MA SEFCTION BRD 2957
1R' 16' 14 . '_ 1 81 6 ------...... MAURITANIA BoO -h~ n ,SECONDLIVESTOCK PROJECT NooHdh,booI \ Z ) Ju9--- DOuodoneOFLOCATION PASTORAL ASSOCIATIONS AND INF RASTRUCTURE
lo Gouero :-{1 i gX ' . _ PAVEDROAD DAKHLET / I PASTORALASSOCIATIONS --- PAVEDROAD UNDERCONSTRUCTION NOUADHIBOU A R " ADRA R SAMPLEDFOR IMPACT EVALUATION LATERITE/EARTH ROADS / ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ISOHTETSIT19A9-19741 IN MM ! J - - - \ -- VETERINARYINSTALLATIONS __ ISOHYETS(IN968 19741 IN MM HEADQUARTERS --- RAILROADS 20/ /' INCHIRI REGIONALOFFICES SELECTEDTOWNS ~/ ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~VETERINARYSTATIONS REGIONAL CAPITALS . CRUSHESFOR VACCINATION NATIONAL CAPITAL i PROPOSEDVETERINARY STATIONS REGIONALAND DEPARTMENTALBOUNDARIES
N Id9oG'--- { A EXPERIMENTALPROJECT AREA -- INTERNATIONALBOUNDARIES
U 25 50 75 10o 125 150 175 KILOMETERS
vf X ~~~~~~~~ ~ ~ cpriiki~ ~~~~~~~~~~~TamssoTc f\
NOUAKCHOll .,!i Id , / , B RA ~~~KNA MonSen *, / ------
ARZA,, I to Id '~~~~~~o Mot-}Rjo- de -- -- '0ea\-
ir-\.^Spanish -' 'o0 ALGERIAnBAK 1 \
- C NGER TnI Soe,. ol ,d,~,dooIEl THE-jTKs,,-~~~~ orIR,,r,opdoro~~~~~,op,p ,p,o1Th,W,,I~~~~~5BorMaiNLtp ,,dor,orr IM I, oIsroI~~~~~~~~~~~~~~~~~~~~~~~~~lory,e~~~~~~~~~to9dt GAMRIA-' -~~~~~~~~~~~,,. -~~~~~~ - Snooko f ~~~~~~~~~~~~Noeey 5 Rknooo "-R~PASO BURKIA < pIonneo1cbo~'rdonA GUINEA-GUINEA / ~~~ f Onog~do~go,,r)1 '/ I6-.ke RISSAU '~~~'-''K~~SH 'RENIN) 14' 14- 10~~ ANA \ " I~~~~~~-h 10 5~~\nlAi. CoookrpD,. OTE~~~~~~~~~~~~~~~~~ / ~~~~~' G 0 10 5' 00 I, el AtT.0,,~~~~~~~~~~~~~~~~~~~~~~~~~~~~V iL
G SpoEush r N,,NE S-i.rERIAALG r,~ (- /14 ,K AdSH\ 1
4 DookryO. I 'gTE D I