20801 SOCIALASSESSMENT of the Community-BasedRural Public Disclosure Authorized InfrastructureProject (CBRIP)

Hanoi, March2000

Preparedby Oxfam Hong Kong consultants Oxfam Hong Kong Consultants for the Ministry of Planningand Investment and the World Bank Lead consultantand author: Natasha Pairaudeau Public Disclosure Authorized Sub-teamleader: Dau QuocAnh

Researchcollaboration and contributions: Phan Lac Tuyen NguyenThi Nghia Tran Quy Suu Nguyen Van Linh Ha Thi PhuongTien Le Thi Phi Van Public Disclosure Authorized Public Disclosure Authorized Acronyms

CBRIP Community-Based Rural Infrastructure Project CEMMA Committee for Ethnic Minorities and Mountainous Areas CF Community Facilitator CIDA Canadian International Development Agency CMC Commune Management Committee DPI Department of Planning and Investment GOV Government of Vietnam GP Good Practice (World Bank Policy Guidelines) ha hectare HCMC Ho Chi Minh City HERP Hunger Eradication and Poverty Reduction Programme IPDP Indigenous People's Development Plan MOLISA Ministry of Labour, Invalids and Social Affairs MPI Ministry of Planning and Investment NCFAW National Committee for the Advancement of Women NGO Non Governmental Organisation PMT Project Management Team PMU Project Management Unit PRA Participatory Rural Appraisal SA Social Assessment UBND Uy Ban Nhan Dan (People's Committee) UNESCO United Nations Education, Science and Cultural Organisation UNICEF United Nations Children's Fund VASI Vietnam Agricultural Sciences Institute VLSS Vietnam Living Standards Survey VMC Village Management/Maintenance Committee VND Vietnamese Dong (unit of currency) WB World Bank WU Women's Union

VietnameseTerms ap remote village; form used in southern Vietnam ban ethnic minority or mountainous village; term used in north central and northern mountainous regions ban nghiem thu inspection board can bo giao thong va xay dung transportation and construction cadre; commune-level cadre assigned to oversee infrastructure work chu ho household head Dinh Canh Dinh Cu literally, "fixed cultivation, fixed residence": government (re)settlement and development programme for upland ethnic minorities: referred to in this document as the Fixed Settlement Programme doi moi renovation du an project (usually refers to an externally-funded project) gia lang village elder, traditional leader ho goc literally 'root households': used in (spoken) Vietnamese to refer to indigenous communities in areas of high in-migration ho khau household registration book; required in principle across the country for each household and its members Kiem Lam Forest Department staff Kinh term used for the (Vietnamese) ethnic majority lac hau backward liet sy martyr /op hoc classroom-refers to temporary classrooms for early grades built in remote villages in some communes mau giao mau he literally,"matriarchy"; used to refer to matrilocaland matrilinealcustoms of some ethnic minoritygroups me tin superstition muong former Thai and Muongadministrative unit nguoi tin nghiem respectedperson nha rong traditionalcommunity meeting house nha tre day care centre Phong Giao Thongva Xay Dung Co Ban (district level) InfrastructureBureau phong tuc tradition ray swiddenor uplandfield tho cam handicrafts thon village thung unit of measurement trinh do thap low level of education truonghoc school xa commune

Equivalents

1 USD = 14,000VND (approx.March 2000) 1 sao = 100 m2 (in centraland south central regions)

Notes:The terms "CBRIP"or "the project" refer to the Community-BasedRural InfrastructureProject. "Subproject"refers to any projectoption chosen by communesfrom the list of eligibleworks. Communeswithin the projectarea are called "project communes"to distinguishthem from "poor communes"which may be any communeon the national list of 1715 poor communes.Terms in Vietnameseare written in italics.

ii

ExecutiveSummary

Part One: Introduction

This Social Assessmentwas commissionedin order to incorporatesocial developmentand capacity buildinggoals into the design of the Community-Based Rural InfrastructureDevelopment Project, or CBRIP.The CBRIP is under preparationby the GOV Ministry of Planningand Investmentfor funding by the World Bank,and is being designed in supportof GovernmentProgrammes 133 and 135. Its aims are to enhancethe capacityfor decentralisedand participatorymanagement of small-scaleinfrastructure works in these regions, provide essentialpublic infrastructureto 532 of the poorest communesin 12 provincesin north-centraland south-centralVietnam, and create opportunitiesfor paid off-farmemployment. The key elementof projectdesign is the direct involvementof communeauthorities and local peoplein all stages of the projectcycle. Communeswill be allowedto prioritise and choose for themselvesthe type of small-scaleinfrastructure they want from a list of eligibleworks. Budgetswill be transferreddirectly to communeswho will then be responsiblefor financial accountingand projectmanagement. This is in line with governmentefforts to encouragegreater decentralisation to local levels. Emphasis will be on the maximumuse of local labourand materialsand on informingand being accountableto local people.

The aims of the Social Assessmentare: 1. to ensure that the project benefitsthe poorest and most vulnerablein the project area, includingethnic minoritiesand women; and 2. to strengthenthe participationof beneficiaries,including the poorest,women and ethnic minorities,and other vulnerablegroups, in all stages of the projectcycle.

The Social Assessmentis based on findingsfrom four main exercises:a field study in 11 communesin the project area, a questionnairecovering all 532 communesin the projectarea, a selective survey and two stakeholderworkshops, held in Nha Trang.

Part Two: Social and EconomicSituation in the ProjectCommunes

Economicand social conditionsvary acrossthe project area, from the changing livelihoodsof ethnic minorityfarmers in mountainousareas; to in-migrant communeswhere indigenousgroups are marginalisedby pressuresto sell land to in-migrants;to coastal and lowlandcommunes which experiencepoor agricultural opportunitiesand harsh climaticconditions.

The 30 distinct ethnic minoritygroups within the projectarea are among the poorest peoplewithin the projectarea. Similaritiesbetween some groups may built upon for planning purposes,despite policiesand approacheswhich to date have not often made use of the local knowledgeof these groups. Elementsto work into a strategywith these groups includeethnic differenceswithin the communes,low Vietnameselanguage and literacyskills, and prevalentspiritual beliefs. Some indigenousforms of social organisation-the role of traditional leaders (gia lang) and of women in some matrilinealand matrilocalminority groups-remain strong.

iii Aside from those peoplewho are economicallypoor within the project area, and the constraintsfaced by ethnic minorities,a numberof other sociallymarginalised groups exist, includingsome types of women-headedhouseholds, the disabled, elderly, and children.

Part Three: Key Playersin InfrastructureWorks

Three types of infrastructureworks exist in the project communes:formal governmentprojects built with externalfunding and assistance,works financed and implementedby local peoplebut motivatedby government,and works financed without govemmentfunding and built entirely by local people. District administrators are the key players in the first type, althoughthere is a will to hand greater responsibilityover to communelevel in many districts.The second type are constructedlargely by local people,motivated by govemment'sguarantee to provide social services.The third type are constructedentirely by local peopleusing local forms of socialorganisation. An opportunityto build upon this type of local self- reliance existsalongside the need to improvelocal capacity to implement govemmentprojects.

Part Four: ParticipationIssues

The distributionof benefits of infrastructureprojects may be uneven across communesand within communes.It is influencedby road accessto the commune itself, the distanceof remotevillages from shared communeinfrastructure, the designand layout of infrastructure(such as water supply systems)within villages, insufficientinvestment, and, most importantly,the existence of appropriateand affordablesocial services.Some types of infrastructuremay be particularlybeneficial to the most vulnerablepeople or groups in the projectarea. The infrastructure priorities of local peoplecome from the practicalneeds withintheir communities,and people rarely separatethe interestsof individualswithin their householdsfrom the interestsof the householdas a whole.

Obstaclesand opportunitiesexist for beneficiaries'greater involvementin all stages of the projectcycle under the CBRIP.Consultation of beneficiariesabout governmentinfrastructure is presentlylimited, and knowledgeof governmentpolicy and programmesis poor. Hesitationto air grievancespublicly is exacerbatedby a lack of knowledgeof proceduresfor doing so; similarly,there is little knowledge among local peopleabout the budgetsfor infrastructureworks built in their areas. Local design and constructionskills are strong in many project communes;these could be incorporatedto better fit infrastructureto local preferences,to increasethe local labour used on such projectsto lowercosts and to increasecapacity for maintenancethrough use of local materials.Currently, those governmentprojects that pay for labourpay very low rates, penalisingpoor householdswhich tend to contributetheir labour rather than cash to such arrangements.Administrative delays in project implementationcan act against local participationby holding up work in times when local people are free from agriculturallabour, or forcing rushed constructionthrough before the rainy seasons.Some experienceexists at commune level in hiring contractors,but communitieswill need to be better aware of bidding and contractingprocedures under governmentprogrammes to ensure they are

iv receivingfair treatment.Despite new regulationsallowing for the existence of managementboards at commune level, these as yet remain limited.Where they do exist, non-officialsrarely play a recognisedrole in bodies set up to manage and supervise projects.

Part Five: ProposedStrategy

The strategy laid out in the final section of the report for incorporatingthe goals of the Social Assessmentaims to ensure that vulnerablegroups will benefitfrom the project and to build upon the recognisedstrengths of the full range of stakeholders currently involved in implementinginfrastructure works.

Communeclassifications are proposed,to provide specific types of supportto ensure local people's involvementin the project, such as the type of monitoring required to ensure the interests of all groups in the communesare addressed.Type A communesare those with in-migrants;Type B communeshave no road access;all other communesare classifiedas Type C.

Ground rules are recommendedfor ensuring communityinvolvement in every stage of the project cycle, includinga conflict resolutionmechanism. Changes to CBRIP design are suggested.Adjustments and additionsto the list of eligible works aim to make the list more appropriateto local needs and priorities, and, by placing choice regarding district-levelworks at commune level, to include local consultationin these type of works. Beginningwith a manageablenumber of communes (and CommunityFacilitators) is recommended,with other project communes phased in subsequentyears.

A recommendedstrategy for project preparationbegins with the prompt selection of good Community Facilitators(CFs), selected for their previousexperience and proven skills. Training requirementsfor these facilitators,Commune Management Committees and district techniciansin areas related to promoting greater participationare outlined.

Policy changes are recommendedin existing laws in support of the goals of the Social Assessment.,and measuresto interlink infrastructuredevelopment with quality social services,and increased access to social services by marginalised groups, are recommended.

A strategy for informationdissemination to communes about the CBRIP is based on disseminatinginformation through multiple channels, and emphasising understandingby the target audiences.The informationwould be channelledthrough CFs and mass organisations,through printed factsheets, informationvideos and provincial printed media. Subprojectinformation at communeand village levels, it is suggested, should be publicly posted in communecentres and villages and regularly updated.

The strategy for implementationincludes recommendationsfor organising subproject selection, appraisal and approval, selectionof ManagementCommittees at commune and village level, financing mechanismsfor subprojects,project design,

v use and maintenance,and additionalmeasures for conflictresolution in the case of inaction or impasses.

A monitoringand evaluationmechanism is proposedthat takes information directlyfrom village and communeslevels (from the public postings)up to central level, as well as indicatorsthat monitorthe distributionof projectbenefits and degree of local communityparticipation. A feedback mechanismis recommendedfor adaptingthe projectdesign in responseto informationfrom local levels on project progress.

Tasks for the economicteam are suggestedwhich aim to contributeto more equitableproject benefitsand better maintenance.

Finally, risks in implementationof the CBRIP are outlined,and mitigation measuressuggested.

vi Acknowledgements The SocialAssessment Team would like to thank: Mr. Pham Hai, Director,Department of Local Economyand Territory,and staff of the CBRIPProject ManagementUnit at the Ministryof Planningand Investment;and Chris Gibbs,Tosca Van Vijfeijken, Social Developmentand Participation,and Le NguyetMinh at the World Bank;for their supportin carryingout this Social Assessment;also YoshikoIshihara, Rural Sociologist,and Josef Ernstberger, NaturalResources Economist, of the Foodand AgricultureOrganisation of the United Nations. At MPI we are particularlygrateful to staff of the ProjectManagement Unit for their work inputtingthe questionnairedata, and to Mr. NguyenVan Quy for processingthe data. We are gratefulto the provincialPeople's Committees and Departmentsof Planningand Investment in ThanhHoa, NgheAn, ThuaThien Hue, QuangNam, Binh Dinh, Phu Yen, Khanh Hoa, Lam Dong, BinhThuan and Binh PhuocProvinces, for their assistancein organisingthe field trips to communes in their areas. We are especiallygrateful to those DPI officerswho accompaniedus and assistedus in the field study. Their commentsand views on developmenttrends withintheir respectiveprovinces helpedto contributea wider perspectiveto the commune-levelstudies. Authoritiesin the followingdistricts and communesassisted greatly by preparingfor the field studies in their areas, and contributingtheir ideasand perspectives:Thach Thuong Commune,Thach Thanh District(Thanh Hoa); Xuan Phu Commune,Quan Hoa District (Thanh Hoa);Xa Luong Commune, Tuong Duong District(Nghe An); Dak NhauCommune, Bu Dang District(Binh Phuoc);Tan Hoa Commune,Dong Phu District (Binh Phuoc);Phan Hoa Commune,Bac BinhDistrict (BinhThuan); Son Hoa District (Phu Yen); Da Sar Commune,Lac DuongDistrict (LamDong); Loc Binh Commune, Phu Loc District(Thua Thien Hue); Son Thai Commune,Khanh Vinh District(Khanh Hoa); Vinh Son Commune,Vinh Thanh District(Binh Dinh),Tra NamCommune, Tra My District(Quang Nam). We thank provincial,district and communeauthorities in Tra Xinh Commune,Tra Bong District, QuangNgai Province,and Ta BhingCommune, Nam Giang District,Quang Ngai Provincefor making preparationsfor field researchteams to visit. We apologisefor being unableto visit due to the severe floodingof November1999 whichprevented access to these communes. We are very fortunateto have had input via questionnairefrom nearly all communesin the project area.We thank those responsiblein each communefor takingthe time to completeand return the questionnaire. We are gratefulto Mr. Ngo Huy Liem for his hardwork in facilitatingthe stakeholderworkshops, and we also thank those participantsof the stakeholderworkshops for their contributions. At OxfamHong Kong, ProjectOfficer Van Minh Chau, ProgrammeAssistant Nguyen Thi Yen, and AdvocacyOfficer Vu HongAnh each accompaniedthe fieldteams on one communevisit. Their experienceof participatoryplanning and community-basedapproaches allowed them to provide helpfulinput to the field methodologyand analysis.Ms. Tran Thi Minh Thu providedcrucial administrativebackup to the projectthroughout the field studies:the field trips and stakeholder workshopscould not have gotten off the groundwithout her organisationalskills and hardwork. We thank Ms. Nguyen Thanh Ha, Oxfam HongKong AdministrativeAssistant, for taking over administrativetasks once Thu left, and DonTuan Phuong,Administrative Volunteer, for his cheery acceptanceof the task of translation.A debt of gratitudeis owed to SusannahHopkins Leisher, Representativeof Oxfam Hong Kong in Vietnam,for taking on this project.Without her supportand insights,this projectwould not have been possible. Finally,the Social AssessmentTeam would like to thank all of those peoplein the communesvisited who took their valuabletime to share their views with us. This report has benefitedfrom the contributionsand insightsof many people.Any errors of fact or interpretationhowever, are entirelymy own. Natasha Pairaudeau Hanoi, March 19, 2000

vii Note on Oxfam Hong Kong'sInvolvement in the Social Assessment

Oxfam Hong Kong has been working in Vietnamsince 1988 to help improve the livelihoods of poor peoplein remote, mountainousareas, with a focus on women, ethnic minorities and the environment.In all Oxfam Hong Kong'swork, we strive to implementour agency principlesof participation,gender equity and empowermentthrough partnershipwith govemmentand non-governmentorganisations at the grassrootsand higher levels. Sectorally,Oxfam Hong Kong focuses on food security(irrigation, credit and savings, handicrafts,agricultural and veterinaryextension and training, tree planting, and other areas), health (water supply,women's health,traditional health, and other areas), capacity buildingfor partnersto strengthentheir abilities in long-term programmemanagement and replication,and advocacyon landmines,Mekong river basin development,participatory approaches,and other issues.

When we first learnedof this rural small-scaleinfrastructure project, Oxfam Hong Kong believedthat the community-basedapproach proposed by the Ministry of Planning and Investmentand the World Bank--anunusual approach for a large-budgetinfrastructure project--hadthe potential to ensure participationof, and extensionof project benefitsto, the marginalised.The project'slarge scale--over1 million people in over 500 of the poorest communesin the country--meantthat the participatoryapproach could have an impact on a significantsegment of Vietnam's poor and marginalisedpopulation, much greater than Oxfam Hong Kong could have through our usual approachof working directlywith partners at the grassrootslevel. Becauseof our belief in the principle of participation,and our interest in advocatinga participatoryapproach at the highest levels of Government,we agreedto take on the task of the Social Assessment.

Oxfam Hong Kongwas reimbursedfor all external costs associatedwith the Social Assessment,but all Oxfam Hong Kong staff time was given without charge, as the Social Assessmentsupports our strategic aim of ensuring that the poor have a "right to be heard". SusannahHopkins Leisher, Representative

20 March 2000

viii xi

99. -sa8!1joud IeBol 61It bS...U .aW. AodwI pPeo wOJJsassOl le!lualOdpue silIauag 8 1 t £. sdnoi3 alqejauInA01 199u98 oq4!oadS lo ainmonilseijuI L I 1 £9. sl4au89 anJnlDfise.Julpue slueJ6!i(m99 Itb . ssae-1gaua8 Aerj sploLasnOH.JJOOd 4 !LM u!sAeM g It' Z. --. sauaq auo6ajoj pue aLUSW9sAUI IUaI!,4nsuj t' l4t' 9. a6eII!A941 U!iqlM splo14snOH J0O SS803V £-IVb 09.-... anuJseiui aunwwoo PajeBS01 ss83o3 ,s 6BII!A 8aLowU Z-I bt 6tb.. ainjonlsejjul saou8nful saunwwoo o0ssGooV peoU MOH L I t 6b7...... 4sj49ua9SoqM :9jnjonjsejjuj aleoS-IjewS L1> 6t . s3nSSI NOIlLVdIl3IlVd :fnolQ IHVd 9S... .q d0.d B.o.l A A..ddela ullng pue paoueu!- sJoM uI z-Z 1 -- Aoilod1U9WUJLAO9 Aq paeAe!l OL sNioM pMouBuid-Alleoul ul Z£ Z7....ou.s!ssV IB....x3 LII...... s1O..oJd U l'Z'£ Z7...... sjOM aJfnlnlseJSilullo sadAl1uqjq14!a !£ 8L u! sjaAeldAq>I I... s)ioM leol pue ,s138foJd, 1£ LtV..N9WdOI3A3a 3un13nlm1SVMNINI SN3AV1d A3>1:33)IH lVd 9£.. sdnojE alqejeulnAPue lOOd £ Z 9£. suoippei8leUol!;eIq LZ-Z-Z .S4...... siapeal elpej 9'ZZZ 9c ...... sja!lalg--- leool 9 zZzZ vC~ ~ ~ ...... - ...... A ap-l pue 96en6uel b Z-Z-Z cE...... UoieJisiUIWPV 811 Ul 4uiM sailpouIN L IUq3 o0uol luasO1JdaUS ZdZZ Z. saunwwoo la8OJda41 u!q!!M so!weuBAj O!UI13 Z Z ZZ .£..lu.uwdolGAa f40!ouivq oiUq11301 4oeoJddV 1USWUSAOO [ I£. s.Jploq8aeieSse s8!UOu!'IO!U413 Z Z Z ---...-...Z ea V foafoida41 u! sdnojs AlUou!IN!u1413 I Z-Z 8Z ...... sanssl 4!JOuIVYOut413 Z'Z -..- ... .z sea Iielseoopue pueBMolU! SG!1Iunwwo £ I Z

Z .saunLuwoo pueldnu! 4U!i>paL4siqBIs3; VZz CZ...... spuejdnf 4101 S1U J6i q Z IvZ ...... seBiV pueldn U! A!Beds!OO!WLUOO3-O!30S £ LV VZ ZO...... LU J6..d SwaIU8SPax!d 811z Z I I 61. suJaUed1uawal1as pue 6UiwJeBA.ewolsn0 I,lIZ 6 -... seajV pueldn U! SPOOq4!18A/!lILZ 81.SUO!.!PUO. IB..OSPpuB 3!WU0Us3 Z 81.. S4NnwwIOo1io3rOtd NI NOIJvnflJSOIWONO:3-OIOOS :OM1 .lVd 91.B.. eepleo!isilels IL £L Z. --.. .. M9!8Auo IAj!4deJ6oa0 I£ I Z. . .. -.. ea.l3..OJdJ°.M.!U.A0. £ L 1. sdoqsNioMJaPIlo4a1elS9-t'ZI L-.. A8aJnS aA!l1loIS 9St'Z-L

9.-. . . Apnls plaS £Z L 9. -. . - -uBwssassV. . leiooSaLl lo sluauodwoo Zlb Z I 9.Slo.1 U eU!1Pf4lO1J8l bZ~ I S ...... I, " ...... 9jnjonjjS pue auijjnO:pod9aS 1 z L t. .SA!..lsqo Ppuesw!V £ Z-L ...... dlJ8O 849Builuawaldwl 0l sa6ualleB43ZZ L £...... lu8wssassV Ie!oS s! 1eB4MI z I £.. -uawssassv le!OOS Z I Z --..-.... sajnfels U6Is8CdINI8O £4V Z...... -.-.-.. .. .D.[OJd Z' do.. 'II

-...... J.ldnolseJjullejnS d paseg-/4!unwwoa814j 11 L.------'NOIlUOnO lNI s,ueluo:) Jo ojqej 4.2 Local Involvement in Small-Scale Infrastructure ...... 56 4.2.1 Consultation...... 56 4.2.2 Information Dissemination ...... 58 4.2.3 Planningand Design...... 60 4.2.4 Implementation ...... 61 4.2.5 Management ...... 63 4.2.6 Maintenance ...... 64 4.2.7 Conflicts and Grievances ...... 65 4.2.8 Financial Oversight ...... 66 PART FIVE: PROPOSED STRATEGY...... 69 5.1 Commune Classification ...... 69 5.2 Proposed Ground Rules for CBRIP Subprojects ...... 70 5.2.1 Ground Rules to Ensure Community Involvement ...... 70 5.3 Recommended Changes to CBRIP Design ...... 73 5.3.1 Changes to the List of Eligible W orks ...... 73 5.3.1.1 Revised List of Eligible Works ...... 73 5.3.1.2 Inter-commune Works ...... 77 5.3.1.3 Further Remarks ...... 77 5.3.2 Phasing ...... 78 5.4 Project Preparation ...... 79 5.4.1 Selection of Comm unity Facilitators ...... 79 5.4.2 Training ...... 80 5.4.2.1 Support Required...... 80 5.4.2.2 Study Visits ...... 81 5.4.2.3 Training Providers ...... 81 5.4.3 Policy/Advocacy ...... 82 5.4.3.1 Policy Changes...... 82 5.4.3.2 Advocacy Issues ...... 83 5.5 Roles and Responsibilities ...... 84 5.6 The Project Cycle ...... 85 5.6.1 Information Dissemination ...... 85 5.6.1.1 Dissemination of Project Information ...... 85 5.6.1.2 Dissemination of Subproject Information within the Project Communes ...... 85 5.6.2 Subproject Selection ...... 86 5.6.3 Subproject Appraisal/Approval ...... 86 5.6.4 Selection of M anagement Comm ittees ...... 87 5.6.5 Financing of Subprojects ...... 87 5.6.6 Project Design ...... 88 5.6.7 Use/Maintenance ...... 88 5.6.8 Conflict Resolution Mechanism ...... 89 5.6.9 Monitoring and Evaluation M echanism ...... 89 5.6.9.1 Reporting ...... 90 5.6.9.2 Monitoring ...... 90 5.6.9.3 Indicators ...... 91 5.6.9.4 Evaluation ...... 92 5.6.10 Audits ...... 92 5.6.11 Review and Adaptation ...... 92 5.7 For the Economic Team ...... 92 5.8 Project Risks...... 92 Appendices

1. Termsof Reference...... 96 2. Listof ProjectCommunes ...... 106 3. DeskReviews of EthnicMinorities in the ProjectArea ...... 116 4. SummaryAnalyses from Field Study ...... 191 5. Questionnaire...... 215 6. SelectiveSurvey ...... 222 7. CommunesWhich Returned the SelectiveSurvey Form ...... 224 8. StakeholderWorkshop Agenda ...... 225 9. PersonsInvited to andPersons who Attended the StakeholderWorkshops ...... 229 10. Reasonsfor LowAttendance at StakeholderWorkshop ...... 233 11. ProjectCommunes by Classification...... 234 12. SecondarySources Consulted ...... 235

List of Tables

Table 1: The 'Four Pillars" and the CBRIP Social Assessment...... 5 Table 2: CommunesVisited in the Field Study ...... 8. Table 3: Repliesto SelectiveSurvey Question: 'Where Did You Get this Information?"...... J1 Table 4: Locationof ProjectCommunes ...... 15 Table 5: Populationin Project Area...... 15 Table 6: Villageswithin the Project Communes...... 15 Table 7: Numberof Householdswithin the Project Communes...... 15 Table 8: HouseholdSize within the Project Communes...... 15 Table 9: Ethnic Breakdownin the ProjectCommunes ...... 15 Table 10: Distancefrom Districtto Commune...... 1.6 Table 11: Ethnic Minority Groupsin the ProjectArea ...... 28 Table 12: Small Local Ethnic Groups in the Project Area...... 29 Table 13: Representationof Kinh Cadres within the ProjectArea ...... 33 Table 14: Ethnic Groups and LanguageGroups in the Project Area...... 34 Table 15: Matrilocaland MatrilinearEthnic Groups within the Project Area...... 37 Table 16: W age Labour Rates by Genderwithin Project Communes...... 39 Table 17: Women's Representationin Commune-LevelInstitutions ...... 44 Table 18: CommuneClassification ...... 7.0 Table 19: Ground Rules for CBRIP Subproject Implementation...... 72 Table 20: RecommendedChanges to List of EligibleWorks ...... 74 Table 21: SuggestedProject Phasing...... 78 Table 22: Recommendationsfor Hiring Community Facilitators:...... 79 Table 23: SuggestedPolicy Changesin Support of the CBRIP...... 82 Table 24: Project Risks...... 93

xi

1 INTRODUCTION

1.1 The Community-BasedRural InfrastructureProject

The Community-BasedRural InfrastructureProject (CBRIP)is under preparationby the Government(GOV) Ministry of Planning and Investment(MPI) for funding by the World Bank. Its design is being developed in support of GovernmentProgrammes 133 and 135,1 and in responseto a number of challengesto implementing satisfactory and sustainablesmall-scale infrastructurein poor rural areas of Vietnam. Subject to World Bank approval, project implementationis expected to begin in January 2001.

The CBRIP is to be implementedin the context of limited resources,during a period of transition from state-dependenceand central control to greater self-relianceand decentralisationof responsibilityto lower administrativelevels. Within this context, the project aims to address a number of recogniseddifficulties in extending physical and social infrastructureto poor rural areas of Vietnam:

1. The poor cannot afford essential physicaland social infrastructure.A governmentprogramme of fiscal decentralisationrequires that rural infrastructurebe funded by district and communeauthorities through funds raised locally. The ability of poorer communesto fund this infrastructureout of local contributionsis very weak, resulting in poor access to physical and social infrastructurein these communes.In addition, income gaps betweenareas mean that the rural poor pay significantly more, relatively,for their infrastructureand services.

2. The ability of commune-levelauthorities to plan and implementgovernment developmentactivities is weak. Governmenthas been promoting decentralisation through a series of laws, decrees and regulations.2 On paper, this has given greater power, autonomyand responsibilityto districts and communes;moreover, the February 1998 Politburo Directiveon Grassroots Democracyproposes that people at commune level should be more actively involved in planning and decision-making.In practice, though, poor flows of informationto local levels regarding the new laws, and low capacity on the part of communeauthorities to carry them out, mean that they are yet to be fully realised.

3. The poorest groups have difficultyarticulating their needs and priorities. Top-down approacheshave given the poor little say in the past on the programmes intended to help them. This is particularly relevant for ethnic minority groups, who

Both programmeswere launchedin 1998, andaim to alleviatepoverty through targeted assistance to poor communes. Programmeactivities of the NationalTarget Programme on Hunger Eradicationand PovertyReduction (HEPR Programmeor 133)and the Socio-EconomicDevelopment Programme for EspeciallyDifficult Mountainous and RemoteCommunes (Programme135) includeinfrastructure development, the developmentof sedentaryagriculture, improvement of socialservices and capacitybuilding. The two programmeshave manycommon characteristics, though the second (135)reflects a stronger concernwith integratingethnic minorities into nationaldevelopment. See NguyenThe Dzung,1999, pp. 24-26,for a detailed explanationof differencesbetween the two programmes,or'planning frameworks,'as he calls them. 2 Recent laws anddecrees relevant to the CBRIPinclude: Decree No. 2911998/ND-CPof May 11, 1998, Promulgatingthe Regulationon the Exerciseof Democracyin Communes;The 1998 Budget Law; DecreeNo. 24119991ND-CPof April 16, 1999, Promulgatingthe Regulationon Organisingthe Mobilisation,Management and Use of People'sVoluntary Contributions for the Constructionof Infrastructureof Communesand DistrictTownships; Law on Complaints and Denunciations (No. 09/1998/QH of December2, 1998).

1 are disproportionately poor in simple economic terms, and live in the more remote parts of the country with distinct languages, cultures and livelihood practices. Putting ethnic minority groups at the centre of development efforts, while respecting these differences, is an ongoing challenge to development initiatives directed at them.

1.1.1 Objectives of the CBRIP

Development Objectives of the CBRIP 1. To enhance the capacity for decentralised and participatory management of small-scale infrastructure works in the poorest provinces of two regions (north- central and south-central Vietnam); 2. To provide essential public infrastructure to selected communes in these regions; and 3. To create opportunities for paid off-farm employment.

1.1.2 Project Scope

The CBRIP targets poor communes where infrastructure is inadequate and where the greater use of public funds is justified by the very limited capacity of these communes to raise funds through local taxation for infrastructure works. This includes 532 communes in 12 provinces in north-central and south-central Vietnam. These communes have been identified by the Ministry of Labour, Invalids and Social Affairs (MOLISA) or the Committee for Ethnic Minorities and Mountainous Areas (CEMMA) (or both) as among the poorest in the country.3 CBRIP documents estimate each commune is able to absorb investments of about US $140,000, in three tranches, for small-scale infrastructure works.4

Some 33 poor communes situated in national parks were excluded from the project,5 as it was assumed that infrastructure investment would conflict with environmental protection measures in these areas. Many of these communes are included under a World Bank-assisted project for forest protection.

The list of 532 communes within the project area is included as Appendix 2.

1.1.3 CBRIP Design Features

The key element of project design is the direct involvement of commune authorities and local people in all stages of the project cycle. Communes will be allowed to prioritise and choose for themselves the type of small-scale infrastructure they want from a list of eligible works (hereafter referred to as 'subprojects'). Grants, it is proposed, will be transferred directly to communes. Communes will get up to three grants, disbursed annually, dependant upon performance. Works will be planned,

3MOLISA definescommunes as 'poor' if morethan 40% of totalhouseholds fall belowthe 'rice povertyline,' equivalent to 15 kg of rice per personper month.CEMMA uses a combinationof criteria includingremoteness, status of infrastructure,social indicators,production systems and poverty to define 'poor'communes. 4Figures from World BankAide Memoire,7 October1999. Communesin bufferzones mayalso be excluded,but to date (March 2000)this decisionhas not beenfinalised.

2 managed, implementedand superviseddirectly from commune level. Emphasisis placed on the use of local labour and materials,to maximise income and employmenteffects in the communesthemselves. The project will provide supports to enhancethe capacity to plan, implement,manage and maintain small-scale infrastructureat the communelevel, as well as to fully involve local people in these processes.The participatorydesign of the project, and the objective of improving infrastructurein remote areas, as well as the World Bank directive on resettlement, would preclude any relocation of villages under the CBRIP.

It is proposedthat at least 80% of funds provided for infrastructureunder the project be spent on commune-levelworks, and up to 20% in districts with three or more project communeson district-levelworks that would benefit at least two communes in the district (if these communesdecide it is in their interest to do so).6

1.2 SocialAssessment

1.2.1 What is Social Assessment?

Social Assessment is a processfor ensuring that World Bank-financeddevelopment operationsare informedby and take into account relevantsocial issues. The practice aims to ensure developmentinitiatives:

3 contributeto poverty alleviation;and at the same time, - enhance inclusion; - increasesocial capital; * build ownership;and * eliminate or at least reduce adversesocial impacts.7

Social Assessmentsare expectedto cover 'four pillars' in support of project operations:

The 'Four Pillars' of Social Assessment 1. Identificationof key social developmentand participationissues; 2. Evaluationof institutionaland social organisationalissues; 3. Definitionof the participationframework; 4. Establishmentof a mechanismfor monitoringand evaluation.

In covering these four areas, Social Assessmentsshould be selective and strategic, focussing only on those variableswith operationalrelevance.

sAt the time of writingit was not yet clear whetherthis would involveonly projectcommunes, or other communesadjacent to projectcommunes. World Bank G.P. 10.05,Social Analysis through Social Assessment. Materialin this sectionis drawn fromthis GP as well as from An Overview of Social Assessment, based on a presentationdeveloped by Sue Jacobs of the Social Policyand ResettlementDivision of the World Bank's EnvironmentDepartment.

3 1.2.2 Challengesto Implementingthe CBRIP

During preparationof the CBRIP, a number of challengesto its successful implementationwere identified:

* weak capacityof commune-levelauthorities to implement government programmes; * limited opportunityfor the large numbers of ethnic minoritieswithin the project area to articulatetheir cultural needs and priorities; * considerablerisk of project benefits being captured by more powerfulinterest groups; and * high degrees of ownershipbeing crucial to the success-and sustainability-of the project.

Furthermore,the CBRIP is a Box1: KeyRequirements of Community-Based community-based,demand- Projects driven project-but on a very 1. Well-chosenand well-trainedproject facilitators; scale.As such, the key 2. Sufficientguidelines within which local peoplecan largescale. s such, e ey implementthe subprojectsin their areas; requirementsof community- 3. Clear understandingby all playersof the conceptof based projects (see box) are community-baseddevelopment and the assurancethat they (or at least some of them) understandand acceptthe concept; more importantthan ever-and 4. Abilitywithin the communityto generateresources to pose an even greater challenge maintainthe infrastructure; to implementationas project 5. Measuresto identify,leam from and correct mistakes managementmanagementstrivesstrives tto promote6.promote earlyMonitoring in implementation;of participation and at all levels. a genuine process of community involvement.

1.2.3 Aims and Objectives

This Social Assessmentwas designedto add value to the CBRIP by addressingthe challengesoutlined above.As facilitating increasedethnic minority participationin the project process and project benefits is a key aim of this Social Assessment,it was agreed that an independentIndigenous People's DevelopmentPlan (IPDP) is unnecessaryfor project preparation.

Aims of the Social Assessment * To ensure that the project benefits the poorest and most vulnerable in the project area, includingethnic minorities and women; and * To strengthenthe participationof beneficiaries,including the poorest, women and ethnic minorities,as well as other vulnerable groups, in all stages of the project cycle.

Objectivesof the Social Assessment 1. Collect sufficient informationon stakeholdersin the project area to assess: * the situation of and capacity for participationof beneficiariesin all stages of the project cycle; and * the possibilitiesfor and obstacles to project benefits going to the poorest and most vulnerable.

4 2. Establish indicators,produce a baselinefor each indicator,and develop a monitoringand evaluationmechanism.

3. Make concrete recommendationsto strengthenproject design.

4. Identify any possible negativeimpacts of the projecton different groups of the poorestand most vulnerable (includingdifferent minoritygroups, men and women), and propose mitigationmeasures to protect the most vulnerable as needed.

1.2.4 Report Outline and Structure

This report is in five parts. The introductionis Part One. Part Two exploresthe social and economicsituation of peoplewithin the project communeswho may be involved in or hope to benefitfrom the CBRIP. Part Three describesthe roles of key players in implementinginfrastructure works within communes.In Part Four, participation issues related to the CBRIP are addressed.Finally, Part Five lays out a strategyfor incorporatingthe aims of the Social Assessmentinto the CBRIP.

Each pillar of SocialAssessment is covered in this report, although,in order for argumentsto flow logically,the report has not been structuredstrictly accordingto the four pillars.

Table 1: The "Four Pillars" and the CBRIP Social Assessment "Pillar' Sectionof thisSocial Assessment 1. Identificationof keysocial development... PartTwo: Socio-Economic Situation in ProjectCommunes ...andparticipation issues; PartFour: Participation Issues 2. Evaluationof institutional... PartThree: Key Players in Infrastructure Development ...andsocial organisational issues; PartsTwo, Three and Four. 3. Definitionof the participationframework; PartFive: Proposed Strategy 4. Establishmentof a mechanismfor monitoring PartFive, Section 9 andevaluation

The design and implementationof the present SocialAssessment was guided by World Bank GP 10.05 on Social Assessment,and World Bank OperationalDirective 4.20 on IndigenousPeoples.

The Terms of Referencefor the Social Assessmentare includedas Appendix 1. 1.2.4.1 Methodology

1.2.4.2 Componentsof the SocialAssessment The size of the projectarea (532 communes)posed a challengeto conductinga Social Assessmentthat was both truly representativeof all the communitiesinvolved and of high quality. It was not feasible, within the time available,to conduct an in- depth field study in a large numberof communes.Instead, the Social Assessment

5 was designedto scale up findings from a smaller group of communesthrough six exercises:

1. A literaturereview: This compriseda desk review of informationavailable on ethnic minoritygroups in the projectarea, and a review of other reportsand documentsrelevant to the project.The ethnic minority desk reviewwas compiled from availablesources in Vietnamese,French, and Englishby Mr. Tran Quy Suu of the Departmentof RuralSociology, Hanoi Instituteof Sociology,and Ms. LuongThi Thu Hangof the Instituteof Ethnology,National Centre for Social Sciencesand Humanities.It is found in Appendix3. Other reports and document reviewedare listed under 'SourcesConsulted' at the end of this report. 2. Consultationwith colleagues:Mr. Matt Desmondat the NGO Training Project, and Ms. Mia Hyun, SeniorTechnical Advisor, NationalCommittee for the Advancementof Women,shared their experienceswith training in participatory approaches. 3. A field study: A sample of communeswithin the projectarea was selectedto be representativeof ethnic groups,geography, relative remotenessand relative povertywithin the projectarea. 4. A questionnaire:A questionnairesurvey was sent out to all 532 communes. 5. A selective survey:The findingsfrom the questionnairewere cross-checkedby a survey in 30 projectcommunes. 6. Stakeholderworkshops: Two stakeholderworkshops were held in order to allow local-level representativesto providefeedback on findings and recommendationsfrom the first part of the field study.

The four main exercises (the field study, the questionnaire,the selectivesurvey and the stakeholderworkshops) are describedin more detail in this section.

1.2.4.3 Field Study

1.2.4.3.1Design The sample of communeschosen for study was intendedto include: • the main ethnic groups living in the projectarea; * both ethnically homogeneousand ethnicallyheterogeneous communes; * both coastaland mountainouscommunes; * communes relativelyfar from district centres;and * some of the very poorest projectcommunes.

Limitationsto selectingcommunes according to this criteria included: 1. A lack of data on the main criteria:although the GeneralStatistics Office had just conducted its 1999 census, which contains informationon ethnic makeupwithin communesacross the country,this informationhad not yet been compiledat the time of the field study, and no informationwas as yet available in Hanoi on the remotenessof projectcommunes. 2. Time constraints:five days were allocatedfor each commune,but travel reduced time in the field to anywherefrom 3 to 4 and a half days. This limitedthe ability of the teams to visit more than one very remote commune(for which ten days had to be allocated).

6 Thus, communes were chosen according to the criteria for which data was available, as well as through calls to provinces and districts for further information. Informationsupplied was sometimes incorrect in relation to ethnicity, and there may have been a tendency among informantsto propose communesthat were not among the most remote.

7 Table 2: Communes Visited in the Field Study Team Commune District Province EthnicGroup(s) Distancefrom CommuneCentre to District Dates Visited Centre A and Thach Thach Thanh Muong 3 hoursby car. Dry-weatherroad. November10-13, 1999 B Tuong Thanh Hoa A Xa Luong Tuong Nghe An Kinh, Thai, and (in remotevillages) Kh 5 km (15 minutes)by car. All-seasonroad. November15-18, 1999 Duong Mu and H'Mong B Xuan Phu Quan Hoa Thanh primarilyMuong, with some Thai and 2 hoursby car (dry-weatherroad), then 1 hour November15-18, 1999 Hoa Kinhvillages walking. A Dak Nhau Bu Dang Binh M'Nongand Stiengindigenous groups; 4 hoursby motorbike(when bridges out) or car. November21-24, 1999 Phuoc in-migrants(mainly Kinh, Tay and Nung) Dry-weatherroad. B Tan Hoa Dong Phu Binh Kinh, Tay and Nung in-migrants 1 hour by car. All-seasonroad. November21-24, 1999 Phuoc A Phan Hoa Bac Binh Binh Cham (Bani Islam) 10 minutes'drive on National HighwayI and then November25-28, 1999 Thuan gravel road. B Son Thai Khanh Khanh Xie Trieng,Rac Glai 3 hoursby car (dry-weatherroad), then 2 hours November26-29, 1999 Vinh Hoa walk to communecentre. A & B Teammeeting, Nha Tran Nov. 29-Dec. 1, 1999 A (did not Son Hoa Phu Yen Cham (ChamHroi) 2-3 hoursby car. Dry-weatherroad. December2-5, 1999 reach commune level) A Da Sar Lac Lam Dong Cil (Co Ho subgroup) 2 hoursby car from Dalatcity. Dry-weatherroad. December7-10, 1999 Duong B Vinh Son Vinh Binh Dinh Ba Na 2 hoursby car. Dry-weatherroad. January 9-13, 2000 Thanh B Tra Nam Tra My Quang Xe Dang 4 hourdrive (dry-weatherroad), then I% days' January 14-23,2000 Nam walk. B Loc Binh Phu Loc Thua Coastal Kinh 1 hour by car. All-seasonroad. January 24-28, 2000 Thien Hue .

8 Field researchtools were designedprior to the communevisits. These included a list of guide questionsfor semi-structuredinterviews, a number of suggested ParticipatoryRural Appraisal (PRA) exercises,and standardformats for reporting findings.

1.2.4.3.2Teams The field study team consistedof nine consultants,each hired for their expertise in areas relevantto the Social Assessment.The lead consultant, Natasha Pairaudeau,has training in anthropologyand experiencein conducting similar researchwith ethnic minorities in Vietnam. Mr. Dau Quoc Anh is an agronomistand an experiencedVietnamese social developmentspecialist, who acted as a team leader. Social developmentconsultants joining the team were Mr. Tran Quy Suu, of the Departmentof Rural Sociology, Instituteof Sociology, Ms. Le Thi Phi Van of the Instituteof Agricultural Economicsand Mr. Nguyen Van Linh of the Vietnam AgriculturalSciences Institutes(VASI). Two gender consultants,Ms. Nguyen Thi Nghia of the College for ManagementTraining in Agricultureand Rural Development,Hanoi, and Dr. Ha Thi Phuong Tien of the Centerfor Family and Women Studies, National Centerfor Social and Human Sciences, also joined the social assessmentteam. Dr. Pham Lac Tuyen, Professor of Ethnologyat the UNESCO Humanismand DevelopmentCentre in Ho Chi Minh City, joined the team as an ethnologyconsultant.

The consultantsdivided into two teams for the communestudies. Both teams worked together in the first communeto ensure uniform methodology.Thereafter, each team conductedresearch in its own assigned communes.Natasha Pairaudeau led Team A, and Team B was led by Dau Quoc Anh. Other membersswitched teams at several points to ensure a good flow of informationand exchangeof ideas.

The schedule for the field trips had to be changedseveral times, due to severe floods in some of the provincesin the project area just before, and again during, the field studies.The field study was originally plannedas a six-week trip from mid-Novemberto late December 1999, with a two-day break in Nha Trang for the two teams to exchange informationand ideas, re-designresearch tools, and rest. Because of the floods, however,three separatetrips were conducted from Hanoi. The first trip was to the north-central region on November 10-18, 1999. The second trip, beginning in Binh Thuan Province,was cut short after the Nha Trang break, as heavy rains forced one team to abandon their work and return to Hanoi, and the other team to change their schedule. One team then returned to the south-central region on January 9-20 to complete the field work. 1.2.4.3.3 Fieldwork The studies in each commune consistedof: 1. Group meetings with authoritiesat district and commune levels, including Departmentsof Planning and Investment(DPI), People's Committeesand heads of mass organisations; 2. Group meetings with leaders at village levels, in three villages per commune. Village leaders and heads of village-level mass organisationswere present at these meetings; 3. Semi-structuredinterviews with traditional leaders;

9C 4. Semi-structuredinterviews with a selection of households.Research teams asked three people in each village to identify the poorest householdsin the village, in order to ensure data was picked up on the range of socio-economic conditionswithin the villages. In some villages up to 15 householdswere interviewed;in a few villages, time constraintsmeant that only three or four household interviewscould be conducted.

The researchtools were refined during the course of the fieldwork. The field researchtools were designed as a series of group meetings and semi-structured interviews,within which the use of a numberof PRA exercises was suggested.Very few of the PRA exercises proposed, however,were actually used in the field. Team membersfound that, aside from being more time-consumingthan more conventional research methods,these exercises were not capturing informationof good quality on complex issues. Althoughthe use of PRA exerciseswas limited, an effective participatoryapproach was used in group meetings, by recordinginformation on flip charts so that all participantscould see and challenge or add to the issues being discussed.The approach used in individual interviewswas to conduct interviews in as relaxed a way as possible,in order to elicit people'scandid views of needs, priorities and problems.In the evenings, team membersdiscussed the day's findings, comparingand analysing what they had heard, and identifyingissues to be clarified further. At the end of each visit, a team meetingwas held and summaries drawn up of informationfrom each communedirectly relevantto the aims of the Social Assessment (see Appendix4). 1.2.4.4 Questionnaire The blanket questionnairewas designed in the course of the first ten days of the field trip and sent to all 532 communesin the project area in mid- December.

The questionnairewas designedto gather information: 1. about the potentialfor participationof different villages and ethnic groups in the project cycle (to further develop findings from the fieldwork); 2. about who has benefited (at village level, and from different ethnic groups)from past projects; 3. to produce a classificationof project communes,allowing WB/MPI to plan for different levels of support on the basis of institutionalcapacity; 4. to produce baseline data for some monitoring indicators;and 5. to assess willingnessand ability to pay.

Althoughall questionnaireswere returned, poor levels of responseto some questions has meant that the Social Assessmenthas been unable to achieve objectives 2 and 5. These two objectives required informationabout infrastructure works constructedwithout external assistance,the costs and number of works constructedprivately, and local contributions.Often such informationis only known at village level, which may have caused the poor levels of response.

Baseline data for communesis not included in this report, due to ongoing work at the Project ManagementUnit on organisingboth the Social Assessmentdatabase and an economic database. However,once this is

10 completed,the informationcan be generatedfrom the database at the Project ManagementUnit.

The questionnaireis includedas Appendix 5.

1.2.4.5 Selective Survey The selective survey was conducted by provincial administrative staff of eight provinces, who received an introduction to this work during the stakeholder workshops. Each surveyor was asked to conduct the survey in 3 communes,for a total of 27 communessurveyed. Of the 27 selective survey forms dispatched,22 completedforms were returnedto Oxfam Hong Kong.

The purpose of the selective survey was to check whether respondents had difficulty in answering any of the questions, and what sources of information were used to gather the data, in order to judge the quality of the information gained through the questionnaire survey. For the great majority of questions, however,surveyors repliedthat they had no difficulty in answeringthe question. Repliesto the final question in the survey however revealedthat while very few questions had been answeredusing the most local levels or direct sources of information(village level informationand field trips were given as the most local or direct sources of informationused, but very few respondentsgave these as informationsources), respondentsclaimed to have answereda third of the questions from commune level sources.

Table 3: Repliesto SelectiveSurvey Question: "Where Did You Get this Information?" Answer No.of Responses Percentage From reality 138 37% From communepeople's 121 33% committee From villages 28 8% From district of higherauthorities 33 9% From statistics 47 12% Through meetings/field trip 3 1% Total 370 100%

The Selective Survey form is includedas Appendix 6. The list of communeswho completed the selective survey is found in Appendix 7.

1.2.4.6 StakeholderWorkshops Two stakeholder workshops were held back-to-back on January 5-8, 2000. Nha Trang was chosen as the site for these workshops as it was felt this was the most accessible location for project communes.Mr. Ngo Huy Liem acted as lead facilitator for the workshops. He was assisted by Mr. Dau Quoc Anh, Ms. Nguyen Thi Nghia, and Mr. Nguyen Van Linh, who, as researchteam members,also presentedthe team's findings. Ms. Tran Thi Minh Thu organised logistics and, with the help of Ms. Nguyen Thi Yen, recordedfindings from the workshop. Representativesfrom MPI and the World Bank joined as observers.

Representatives were invited from province, district and commune levels. A total of 18 communes (five of which were communesvisited in the field study) from ten districts and ten provinces were invited. Communeswere asked to send one

1 1 male and one female representative;and districtsand provinces were each asked to send one representative.

Attendance, particularlyof communerepresentatives, was poor for Workshop B. Only 6 out of 18 expected delegatesfrom communelevel attended.While this allowedfor closer discussionthan may have been possible with a larger group, reasonsgiven for the poor attendancereflected some of the difficulties involved in ensuring participationof people from remote areas (see Appendix 10).

The purposeof the stakeholderworkshops was to gather feedbackfrom commune-levelrepresentatives on findingsand recommendationsfrom the first part of the field study. District and provincialauthorities were invited as observersand asked to listen to the views of the communes,rather than to actively contributeto the commune-leveldiscussions. Views were gatheredfrom participants by asking them to write their responsesto specific problemson cards: these cards were then pinnedon poster boards and analysedthrough group discussions.The methodologywas refined in workshop B: two separate sessions were conducted, one for communerepresentatives and one for district and provincial representatives. The district and provincial representativeswere then actively involved,when the two groups came together, in analysing the cards producedin communediscussions.

The agenda for the workshopsis includedas Appendix 8. Appendix 9 provides lists of people invited to the two workshops,and those who attended.

1.3 Overviewof ProjectArea

This section briefly describesthe areas within which project communesare situated and provides some basic quantitativeinformation about communeswithin the project area.

1.3.1 GeographicalOverview

The project area covers twelve provincesin two regions of Vietnam.

Nghe An and Thanh Hoa Provincesmake up the north-centralregion of the project area. Within these two provinces,there is a total of 196 'poor communes',as defined by MOLISAand CEMMA,selected for inclusion in the project.

The second, south-centralregion, consists of a chain of ten provinces followingthe coast south from Thua Thien Hue. The chain of project provinces edges the Central Highlands Provinces,turning inland south of Dak Lak Provinceto meet the Cambodianborder at Binh Phuoc Province (see map overleaf).A total of 336 communesin the south-centralregion are included in the project.

Most communesselected to be includedin the CBRIP are located in districts clustered up against the western borders of all these provincesin upland areas. This reflects the selectioncriteria: communes in the uplands, inhabited by ethnic minority groups, number disproportionatelyamong the poorest in the country.

12 Also includedwithin the project, though, are some communeslocated in the plains and coastalareas 8 which, in MOLISA'sjudgement, also experienceextreme poverty even though they are less remote. Lowlandcommunes are inhabited largely by the Vietnamesemajority (or Kinh people), with the exceptionof the Cham, who have lived for many centuriesalongside Kinh people in lowland areas of south and central Vietnam, but retain their distinctive languageand culture.

8 This includes five communes located in the coastal township of Cua Lo, Nghe An Province.

13 I Map of Vietnam'sProvinces

14 1.3.1.1 Statisticaldata

As the number of projectcommunes has yet to be finalised, and some data is incomplete,the data provided in the following tables is intended as a guide only.

Table 4: Location of Project Communes Zone Number Percentage Highand remote 252 48% Mountainous 220 42% Lowland 25 5% Coastal 28 5% Total 525 100V source:Social Assessment Questionnaire (communes self-reporting)

Table 5: Populationin ProjectArea Persons Totalpopulation in projectcommunes 1,429,881 Averagepopulation per commune 3,429 Largestcommune population 19,964 Smallestcommune population 279 source:Social Assessment Questionnaire

Table 6: Villages within the Project Communes Villages Totalnumber of villagesin projectarea 3,056 Averagenumber of villagesper commune 6 Highestnumber of villagesper commune 25 Lowestnumber of villagesper commune 1 source:Social Assessment Questionnaire

Table 7: Numberof Householdswithin the Project Communes ___ _Households ______Totalnumber of householdsin projectcommunes 291,354 Averagenumber of householdsper commune 709 Largestnumber of householdswithin a commune 39,902 Smallestnumber of householdswithin a commune 70 source:Social Assessment Questionnaire

Table 8: HouseholdSize within the Project Communes HouseholdSize Persons Average 5 Largest 12 Smallest .1 source:Social Assessment Questionnaire

Table 9: Ethnic Breakdownin the Project Communes Populationwithin Project Area Number Percentage Ethnicminority 700,566 49% Kinh 729,315 51% Total: 1,429,881 100% source:Social Assessment Questionnaire

9This total does not match the 532 communesinterviewed, as some communesfailed to answerthis question.

15 Table 10: Distance from District to Commune Dry Rainy Season Season Averagetime requiredto walk from communecentre to district centre 5 hours 7 hours Longesttime requiredto walk from any communecentre to its district 12 days* 8 days* centre Averagetime requiredto drive from communecentre to district centre 2 hours 2 hours (for communeswith road access) Longesttime requiredto drive from any communecentre to its district 4 days 10 days* centre Shortesttime requiredto drive from any communecentre to its district 3 minutes 3 minutes centre *sic.

16 PART TWO: SOCIO-ECONOMICSITUATION IN PROJECTCOMMUNES

Part Two explores the social and economic situation of people within the projectcommunes who may be involved in or hope to benefit from the CBRIP. The first section looks at the economic and social conditionsfound in poor communeswithin the different geographiczones in the project area. The second section identifiesthe ethnic minority groups within the project area, and analyses policy, social and cultural issues relevantto them as stakeholdersin the CBRIP. The last section identifiesthe most marginalisedgroups within the project communes, and describes factors that make them impoverishedor relegatethem to weak positionswithin their communities.

1 7

2 PARTTWO: SOCIO-ECONOMIC SITUATION IN PROJECTCOMMUNES

2.1 Economicand SocialConditions

Communes(xa) are the lowest administrativeunit. They comprisevillages, which, dependingon the area, may be referredto as ban, thon, or ap (see below). Aside from those communeswith migrant and resettledcommunities, most communesare much more recententities (constitutedin the post-revolutionaryperiod) than villages. It is villages, rather than communes,that are genuinecommunities in the sense that they are traditionalsocial units with varying levels of internalcohesion; although social ties exist within communes,communes are firstly administrativeunits. Resettledvillages visited in the field study were whole villagesthat had moved, and thus (althoughsome villagers had stayed behindor returnedto their village of origin) socialties were similar to those in the village of origin. Villages,despite some intermarriagebetween the ethnic groups, tend to be ethnicallyhomogenous. Marriageis more prevalentbetween people of the same ethnicity,and this tends to create ties betweenvillages of like ethnicitywithin the communeand with nearby communes.

Ban is usedto refer to a mountainousethnic minorityvillage (and is a Thai term). Its use is prevalentwithin the project area in NgheAn and Thanh Hoa. Within communes,however, villages have also been numbered,for administrative purposes.At local levels, villages may be referredto by their name, or by their number,in which cases the lowlandterm for villages (thon) or a compositeterm (thon/ban)is used. In some southern provinces,the term ap is sometimesused locally to refer to a village within a commune;strictly speaking,it meansa remote village. Some confusionarise in translationin southern parts of the project area, as xa is at times translatedas 'village', and ap as 'hamlet'. This may be due to the fact that some hamletsin lowlandareas of the south are generallysmaller in size and populationthan those found further north and in more mountainousareas.

Communeswithin the projectarea cover a number of geographiczones, from extremely remote mountainousareas through (relatively)more accessibleupland areas, to coastalfishing areas. Settlementpatterns vary acrossthese regions.In mountainousareas, villages tend to be more scattered,made up, particularlyin higher mountainareas, of separatevillage clusters.Villages in midlandareas tend to be built in valley beds with houses more closely clusteredtogether but still with gardens and paddy fields betweenhouses. In lowlandareas, villages may have houses quite closely gathered together,with fields outlying. In Cham communes, householdsare laid out particularlyclose together.The Cham communevisited had houses arrangedalong street axes as in a small town, with no agriculturalland betweenthem.

As might be expected,there is some variationin livelihoodpractices across the different geographiczones. To some degree in every projectcommune, though, the transition to a market economyor some effect of it is likely to be a day-to-dayreality: in the field interviews,discussions inevitably came to focus upon the impact of very

18 recent changesthat had taken place or been orchestratedin the communes in the last five to ten years. This section describes the range of socio-economicconditions found within projectcommunes, and highlights the socio-economicdisparities that exist betweenhouseholds in each of these areas.

2.1.1 Livelihoodsin UplandAreas

2.1.1.1 CustomaryFarmning and SettlementPattems H'Mong and Dao ethnic groups customarilyinhabit the high mountainareas of the North Central Region and have long been dependenton the cultivationof tubers, rainfedcorn and hill rice on sloping, swidden fields. Traditionally,this is supplementedby the exploitation of natural forest resourcesand rather limited animal husbandry(almost exclusivelyfor householdor village use). Until recently, these were largelysubsistence economies. Despitesome longstandingpatterns of trade betweenthe minority groups themselves,very limited trade existed between the upland groups and the lowland Kinh.

Those ethnic groupswho inhabit the lower mountainareas (or midlands)of the north-centralregion (such as Muong,Thai, Tho and Kh Mu) have long traditionsof wet rice cultivation.They customarilylive in valley beds where paddy cultivation is feasible; this is combined with some uplandcultivation and livestock raising (includingbuffaloes to break land for paddy cultivation). Comparedwith their H'Mong and Dao neighbours,they have a more active history of interactionwith the lowland majority.

However, the connectionbetween ethnicity,elevation, and cropping patterns is not consistentin the north-centralregion. Some H'Mong communities have grown wet rice for many generationswhere local conditionshave permitted; conversely, some midlanderslive in areas where wet rice cultivation is not possible (such as the Muong community in Tan Hoa commune, Quan Hoa District, Thanh Hoa).

Many highland minoritygroups in the south-centralregion of the project area also practice slash and burn cultivation. Terracingin a Xe Customarily,Custmarly,this has been the mainsource of BoxDang2: Traditional Community in Tra Nam food for the household, and is supplemented by Commune,Quang Nam' 0 exploitation of forest products for household A Xe Dangfarmer in this remotecommune use. This system is practised by ethnic groups owns 5 sao (500 m2) of wet rice fields where or living in a variety of geographiczones, from atshe least growsone either summer two ricecrops crop. of riceHer perfields yearand mountainouszones (the Ta Oi and Pa Co in A those of householdsin the village are irrigated Luoi District, Thua Thien Hue) to areas that are by a local system of bamboopipes and plateau and low hills (the M'Nong and Stieng in riceearthen and canalscassava createdcultivation by localare farmers. both Wet Binh Phuoc, the Co Ho, Cil and Ma in Lam longstandingtraditions in the commune.Tra Dong Provinces).However, communitieswith Nam Communecurrently has 77 ha of rice longstandinglongstandingtraditions of wet rice cultivation ~~~~~~~~~~~offieldstheir wherefood requirements local people produce fortheir roughly own half can also be found in the south-centralregion consumption. (see box opposite).

The authorof boxed cases studiesis NatashaPairaudeau, unless otherwise stated.

19 Upland minoritiesin Vietnam are commonlysaid to be 'nomadic' or 'semi- nomadic'.This stems from the fact that swidden farmers cultivate one upland plot for a period of several years, before leaving it to lie fallow and either returningto a previously cultivated plot, or clearing a new plot for cultivation.Nomadic, however, is a misnomer:during the courseof the field research,the only instancesfound of movementsin human settlementsin recent times were in responseto external events. Many highlandcommunities in the south central regionof the projectarea fled their native villages two decades ago due to wartime disruptions;others were forcibly moved during the same period as part of the StrategicHamlets Programme. Still other communitieshave movedfurther downhill in the last decade as part of the government's'fixed cultivation,fixed residence' or dinh canh dinh cu programme (hereafterreferred to as the 'Fixed SettlementProgramme'). Communities may move as a responseto illness (such as malaria) in one location:while this type of move may be locallyjustified for spiritual reasons-to get away from the yang or spirit that causes the difficulties-there may also be practical reasonsfor these moves, in that people are leaving locationsprone to such illnesses.The notion implied by the term 'nomadic', though-that these groups are moving on a seasonal basis, or that whole villages move at the slightest whim, clearing large areas in their wake-is unfounded.

=> The CBRIP is not being implementedin areas inhabitedby 'nomadic'groups in the proper sense of that word. Upland communitiesin the project areas are customarilystable settlements;it is farmers' plots, where slash and bum practices are employed,that shift across limited areas. 2.1.1.2 The Fixed SettlementProgramme Governmentregards the practiceof slash and burn cultivationas unsustainableand environmentallydamaging, and has programmesto encourageethnic minoritiesto change to sedentaryand intensiveforms of farming. The emphasis is on encouragingwet rice cultivation,wherever possible. In the period of economic openingthis has been expandedto includethe promotion of commercialanimal husbandry,the cultivation of cash crops, and forest tree planting.

Efforts to settle ethnic groups have been stepped up in recent years along with a general ban on the exploitationof forest products.Although the fixed settlement programmehas been in place for many years, in local areas people often speak of the campaignto 'settle' in their areas having occurredfrom 1993 onwards, aroundwhich time a firm resolve by governmentto protect forest land was instituted. Efforts to encourage 'settlement'sometimes contradictforest protectionmeasures, however. In one ethnic Ba Na village, for example, local people had cleared trees to prepare plots of land for planting coffee, whereas under the reforestationprogramme they had receivedseedlings from governmentto reforest barren hills. Photo:Coffee growing on recentlycleared land in VinhSon Commune, Binh Dinh Province (photo: DauQuoc Anh)

'Fixed cultivationfixed residence' means actual relocationfor some communities;for others, it meansstate-regulated changes in agricultural production. The term dinh canh dinh cu is used widely by highland peoplewhen discussingtheir livelihoods,but they often mean different things by it. Thus in a

20 H'Mong village in Nghe An Province,the whole village has moved underthe programmeto a location further downthe valley, to a new cultural context in closer proximitywith other ethnic groups, to contact with schools and a clinic, and to new agriculturaltechniques and crops. 'Dinh Canh Dinh Cu' is also associatedin this village with giving up opium cultivation,with subsequentloss of income. In a Cil minority communein Lam Dong Province,households have been encouragedto give up cultivation of their swidden slopes, acceptthe altocationof forest protection plots, and adopt kitchen gardens and coffee cultivation through a subsidised governmentprogramme. It is only in some extremely remote areas, inaccessibleby car (such as Tra Nam Commune in Quang Nam), that relatively little is known about the fixed settlement programme.

> In many upland communesin which the CBRIP will be implemented,changes in livelihoodpattems have been strongly influencedby govemment development priorities.

2.1.1.3 Socio-EconomicDisparity in UplandAreas Households in relatively accessible upland areas across the project area are growing a number of cash crops. Encouragedby varying degrees of government directionand state subsidies,farmers in the upland communesin Nghe An and Thanh Hoa cultivate sugar cane and fruit trees; coffee, cashews,and pepper are cultivated in highland (including plateau)areas of the south-centralregion.

Those with better access to knowledge about the new agricultural techniques, and those who can afford the inputs-fertiliser, pesticides, new varieties-are inevitably among the ranks of the better-off. These householdsare characterised by more labour and capital-both economicand social-than their less successful neighbours.Often, they will have some connectionwith or position within the administration.The tendency to assign forest protectionland to those most able to protect it reinforcesthe position of these successful households(see box below).

21 These more successfulfarmers, Box3: Forest Protection Land Allocation In a Cil who have managed to weather Village the transition from subsistence All the forest protectionland for this villagehas alreadybeen upland farming to commercial allocated.Households with these plots are paid 40,000 VND crops, are cautiouslyoptimistic ha per quarter to protect the trees on the land. Nineteenplots wereallocated to thisvillage: the largest allocabon to any about their future. They are single householdwas 33 ha; the smallestwas 15 ha. Mrs K., keenly aware of the constraints witha familyof nine grownchildren, has a rather large plot. placed on them by poor Whenasked how her family protectsit, she says they have builta smallhut on the land,and one person is postedthere transportationto their local areas, keepingwatch, when the weatheris fine, to make sure no one which raises the price of comesand cuts any wood. Her householdeams 325,000 agriculturalinputs and prevents VNDper month for doing this. Mrs.T., by contrast,was not allocatedany forest land. It wouldbe difficult,it is true,for her them from selling at competitive householdto free up a memberto stand guard over a stand of prices. They are deeply troubled treesfor long periods: she herself,the head of the household, by drops in commodityprices in is disabledby polio and walkswith difficulty. The twoother adultsin the household,her younger sister and her brother-in- some sectors (sugar cane in law, have their handsfull with twoyoung childrenand most of particular)that have been subject the agrcultural tasks-althoughMrs. T. herselfdoes manage to strong state control. to do some work tendingthe coffee plants behind Thisfamily lives in a one-room,one-bed house. When theirasked house. whyshe was notallocated a forestplot, Mrs. T. saidno one However, there is a marked had ever asked her if she wantedone. The People's difference between these Committeehad held a meetingand allocatedthe land very quickly,and 'many peopleare awayfrom the villagefor households and those which severaldays at a Umetending their upland plots (ray), so they are less able to adapt. In many could not be herewhen the meetingwas held." communes,people describedthe better-offas those who had 'fixed their cultivation and residence',and the poorer householdsas those who had 'not yet fixed their cultivation and residence'or were 'not yet able to fix it' (chua dinh canh dinh cu).

Because of shortfalls in the new forms of production,poorer households may retain, or return to, cultivatingstaple crops on hillsides. In communitieswhere hill slopes are very distant from villages, these people may spend long periods away from villages, which in turn distancesthem from new types of extension knowledge, and, significantly,from decision-makingprocesses in the commune. Frequently,they continue to depend on gatheringfirewood, a practice now regardedby those who have improvedtheir lot in recent years as a last resort (as well as sometimesbeing against the law).

In addition, upland householdsincreasingly use wage labour as an extra income-earningstrategy. This frequentlyconsists of agriculturallabour within their own communes,but may also take peoplefarther afield. In better-off householdsit tends to be young people who engage in wage labour, leaving key agricultural labourers (almost always women)to continuewith farming. In very poor households, however, it is often key adult labourerswho engage in wage labour to earn money for immediateexpenses. z Swidden cultivatorsare the most vulnerablefarmers within the project communes. => Farmersdependent on cash crops remain vulnerableto the risk of crop failure and, increasingly,to fluctuationsin commodityprices in the global economy. = Among the most vulnerable householdsin uplandminority communesare those which lack the resourcesto succeed in more intensive and commercialised farming techniques. They may revert to their customary farming practices,

22 possibly marginalisingthemselves further through distancing themselvesfrom decision-makingprocesses within their communities.

2.1.2 Migrants to the Uplands

In some project communes,in-migration over the years has resulted in the presence of individualsor communitiesfrom the lowlandswhose livelihoodsdiffer somewhat from those of the indigenousgroups. Migrationto the uplands began prior to the period of economicrenovation (doi mol) as an organisedgovernment programme, with land offered to (primarily) Kinh migrantsfrom the lowlands,sent to 'new economic zones' to open up and develop the more remote areas. Some of these in- migrants are now well-establishedwithin the communeswhere they live. Since restrictionson travel were lifted in the late 1980s,a wave of spontaneousmigration (of Kinh but also of minoritiesfrom NorthernVietnam) has overtakenthe organised migrations and is currently causing significantsocial changeswithin the communes where it occurs.

2.1.2.1 EstablishedKinh in Upland Communes In upland ethnic minority communes,the key players in commerceat communelevel are often the few Kinh residentsof the communeor district. These are peoplewho may have been sent from lowlandareas up to three decades ago to develop the uplands. Many have developed very basic commercialinterests to supplementtheir agriculturalproduction and are now concentratedin commune centres, where they are the proprietorsof shops selling basic necessitiessuch as oil and salt. Their activities often extend to the sale of agricultural inputs and the purchase of local produce. In addition,they frequentlysell rice or other goods on credit in shortageperiods: thus, although these people act as middlemenin the limited commercialtransactions of upland areas, they are also consideredby some local peopleto play an importantrole in sustainingthem through times of hardship.

Other Kinh residentsof upland communesare teachers, assignedto teach in minority villages. Due to their frequent travel outside their villages, some Kinh teachers involvethemselves in trade to supplementtheir teachers' salaries.

In many placesthese same Kinh in-migrantsplay prominentroles in the administrationof their communes.This issue is discussed in greater detail in the section that follows, on ethnic minority issues.

=> In upland minority communesit is common to find some Kinh residents who have been establishedfor relatively long periods of time within the communes. They tend to hold dominant positions in commerce,education and administration.

23 2.1.2.2 SpontaneousIn-migration Some upland communes within the project area are currently experiencing rapid in-migration. Two communes Box4: RapidInflux visitedduring the field study(Dak Nhau Inthe two communesvisited in BinhPhuoc Province, Commune,Bu DangDistrict, and Tan the completionof the mainroad into the provincein HoaCommune, Dong Phu District,both 1998coincided nicely with impressive rises in theprice Habo ueDgi, of coffee.The combinationhas created a migration In BinhPhuoc Province) where in- boomin thearea. Provincial authorities put the number migrationis ongoing.Migrants-most of of spontaneousmigrants into the province each year at themspontaneous-continue to arrive 30,000persons. Organised migration, they say, is now verylimited. Tan HoaCommune in DongPhu District in both of thesedistricts at a rapidrate reportedreceiving 200 permanent settlers from Cao (see box opposite).Many are from Bangand Lang Son Provinces in 1999,and a further ethnicminority groups from the northern 70-80seasonal migrants in the sameperiod. provinces-primarily Tay and Nung." Other immigrantsencountered were urban Kinh from Ho Chi Minh City. In the two communesvisited where in-migrationwas prevalent,the migrants had only arrived in many cases in the last three to five years, and had yet to reap their harvests;their main intendedsources of income,though, are cash crops (coffee, rubber, cashew, and pepper).

Factors pushing migrants from other rural areas include limited agricultural land and little opportunity to make a better life from agriculture alone in their home areas. Spontaneousmigrants are often motivatedto select certain destinationsbecause of networksof friends and relativeswho have migratedto the chosen area. Within the chosen areas, they tend to favour locationswith better access to road networks,and where infrastructureis better established.

Box 5: Urban to Rural Migrant Box 6: Ethnic Identity and Migration The motorcycletaxi driver explains, as he drivesup Sovaried is the ethnicmakeup of onein-migrant to DakNhau Commune (Bu Dang District), how he communeof Bu DangDistrict that it hasbeen dubbed cameto settlehere. Until three years ago he locallythe 'UnitedNations'. Minority groups have workedas a truckdriver based in HoChi Minh City. flockedhere from all over the northemmountain Attractedby the prospectsof growingcoffee, he provinces,and it is strangefor the visitorto catch boughtland and settled here with his wife and four glimpsesof headscarvesso typicalof the northin the childrenin 1996.Last year the familyreaped their heatof a southemplateau. There are also,however, first harvest.While he is concemedat the dropin indicationsof deepchanges occurring in theway coffeeprices this year (he is expectingto get peoplesee themselves: in somemigrant areas the Tay 12.000VND/kg of coffeethis year against about andNung now consider their ethnic differences to be 20,000VNDlkg in 1997and 1998), he remains negligibleand speak of themselvesinstead as Kinh. optimisticand says life is betterfor himhere than it Interestinglythough, many of the migrants,like their was in the city. indigenousneighbours, are Christians. Within the communes studied, most migrants claimed to have established themselves by purchasing or renting land from indigenous peoples. There were also a very few informants-some of the earliest arrivals-who reported having been allocated land ("land that no one was using," said migrant informants,or "unoccupiedland"). These earlier arrivals may now also be enlargingtheir holdingsthrough purchase.Some migrants may wage labour until they earn enough to establish their own holdings.Seasonal migrationto these areas for wage labour is also common. The ProvincialPeople's Committeeof Binh Phuoc also cited a problem of spontaneousmigrants clearing (and effectively

" There is a historyof Tay and Nung groupssettling in the southfrom 1954, andmany of the more recentTay and Nung immigrantshave establishedthemselves in the new areasthrough contacts with relativesand friendswho have beenthere much longer.

24 claiming)forest land, and expressed great concernwith preventingthis from happeningany further.12 However,the field study did not produceclear evidence of actual conflicts arising over land use betweenestablished and new land users. This was primarilydue to the limited coveragethat the study was able to make of in-migrant communities.13 Within the two communescovered in the study, no respondents spoke of conflicts arising in the course of land transfers,nor was any evidence given of conflicts over land rights caused by earlier governmentallocations of plots to in- migrants under organised programmesof migration.Other commentatorshave found it equally difficult to point to clear evidenceof conflicts arising in the allocation of land to in-migrants,but it is clear that this issue has yet to be adequately explored.'4

However,these movementsare clearly creating uncertaintyfor the indigenous groupsliving in these areas. Local people in these areas tend to be relativelypoor indigenousminority groups, who either Box7: Differentviews on land continueto live a very simplesubsistence acquisitionand in-migration existence, or are experiencingsome of the Somemigrants claimed that indigenous people ex s , oriae exp er some ofth are very willing to sell land, eitherbecause they problems,described above, associatedwith do nothave the labouror the industryto farmit the transitionto 'fixedsettlement'. When themselves,or in somecases, because they spontaneousmigrants arrive, indigenous aretaken by the suddenvisible wealth in the area and are eager to exchangetheir land for minority groups appear to sell or rent land short-termgains, such as motorbikes.Others- situated in areas that are (or become) notablyindigenous people themselves- relatively'central,' through the constructionof claimedindigenous groups are muchmore wary ro of losingtheir land. One district leader in Lac ads or commercial centres. They thus DuongDistrict, from the local Cil ethnicgroup, gradually installthemselves on the parts of statedthat within his district, local people will theirown holdingsthat have,through onlyrent out land to incomingmigrants, but infrastructureimprovements, become by neversell it. neversell ft. I definition more 'marginal'.There was some suggestion that indigenousgroups are moving through preferenceaway from areas as they open up and become more commercial,but other sources suggest some indigenous minorities may sell their entire holdings and move to towns in the highlands in response to the arrival of in-migrantsin their areas.15 Curiously, no indigenous people interviewedin Dak Nhau Commune claimed to have sold (or transferred)any of their land to in-migrants,and none were landless.Both M'Nong and Stieng villages in this communeare situated on the peripheryof the commune, however.

Box 8: Different attitudes towards in-migrants and indigenous ethnic groups In interviewswith cadres and households themselves, clear distinctions were made between the migrant communities-'driven','ambitious' and 'industrious'-and the indigenousgroups in thearea-'poor cousins', 'slowto adapt'and uninterested in or unableto improvetheir household economies.

12 Provincialauthorities in Binh Phuoc say that landallocation in that provincehas not been completedbecause many people- the Tayand Nung, it is claimed-havecleared land categorised as forestland, and are now using it for agriculture.The provincialauthorities cannot forcibly removethese people,whose livelihoodsnow dependon this land. However,nor can the provincechange land categories without a rulingfrom central levels. 13 Of thetwo communes where in-migration was on-going, only one hada mixtureof in-migrantsand indigenous people, and the otherwas madeup entirely of in-migrants. 14 SeeA. Hardy,Red Hills, (Migration to the UplandFrontier), forthcoming. 5 See Hardy,p. 328.

25 => Indigenousethnic groups in areas of high in-migration are without doubt the most vulnerablecommunities within the project area. Already the poorest households and villages within in-migrant communes,with difficulty in adapting to new agriculturaltechniques, they are subject to pressuresto sell land to in-migrants. This may place them at the peripheryof their own communesor may cause them to leave their communes altogether. > Contraryto the proposal made in CBRIPpreparation documents,1 6 proof of completionof land allocationis not a sufflcientcriterion to identify in-migrant communes wherepopulations are clearly and pernanent established.As indigenouspeople are selling land of their own free will, the completion of land allocation does not necessarilyguarantee that the community will remain well- establishedif and when land allocationis completed.17

2.1.3 Communitiesin Lowlandand Coastal Areas

Poor local conditions limit the successof agricultureand fishing,the main sources of income in these communities.Agricultural production is limited by poor, insufficienttracts of land, and inadequateirrigation. Some poor communes near the coastface persistentdifficulties of salinated drinkingwater and irrigation supply. Coastal communitiesreliant on fishing are faced by increasinglylimited stocks. Harsh climactic conditions-the floods that afflicted lowlandand coastal project communesfrom Thua Thien Hue to Khanh Hoa are an extreme but clear illustrationof this-mean that people periodicallysuffer tremendoussetbacks as they try to improvetheir well-being.

Some poor lowlandand coastal communesin the project area are new economiczones. For example, Loc Binh Commune in Thua Thien Hue is a community of peoplewho moved in 1975 from an adjacentcommune due to land constraints.These communitiestend to be built on marginal tracts of land which are highly susceptibleto flooding or other natural calamities.

Box 9: FloodDamage in LocBinh Commune Loc Binh is a poor and small coastal commune of Phu Loc District, which is about 40 km from Hue city to the south-east. Until mid 1999, there was no vehicle access to the commune. People's main mode of transportation was by boat. In the third quarter of 1999, the govemment invested approximately 10 billion VND to construct a road 14 km in length from National Highway No. 1 to Loc Binh. Before the floods in November 1999, 7 km had been completed and the rest had been cleared and levelled and a sewer system put in place. Thankfully, the commune suffered no loss of life in the two floods in late 1999. However, much property was swept away or seriously damaged. The road was damaged in 30 places. Main bridges and sewers were swept away. The road is being cleared again, but people have gone back in the meantime to relying on their boats. (Nguyen Van Linh) Photo: This bridge in Loc Binh Commune,Phu Loc, collapsedin the floodingof November1999. Completedin the same year, it had barely been used. (photo: Dau Quoc Anh)

16 CBRIPpreparation documents propose that the satisfactoryresolution of all landrights issuesbe used as a criteriato prove that communitymembers are capableof puttingforward proposals that reflectthe genuinelong-term development objectives of the groupas a whole.The World Bank would not financeinfrastructure, it is suggested,for newlycreated communities under the government'svoluntary resettlement programme unless these issueshave been satisfactorilyresolved (PCD, October 9, 1998,p.12). 17Moreover, if settlementof land rightsissues is to be used in all projectcommunes to approveCBRIP funds,many communitiesthat are 'permanentlyestablished' would be excludedfrom the CBRIP.Land allocationis incompletein many areas where communitiesare long-and well-establishedand there is no in-migration,due to lack of funding,lack of time, capacityand staff to conductsurveys, and lack of commitmentto or interest in carryingout allocation.

26 The better-offhouseholds report some improvementin their well-beingin recent years. Those with the Handwoventho cam brocade plays an important role in resourceshave been able to traditions;sashes, shawls and ceremonialgarments are madeof o 1s S multicolouredcloth, which is used at importantoccasions and in Box 11: Single Mother In Phan Hoa religiousceremonies. In the past, this clothwas madefor local Us ahe is 36 years old-with4 children.The and neverfor commercialpurposes. Cham Bani weaversof Bac ngest is a baby girl. A thirteen-year-oldboy District(Binh Thuan Province),whose ho cam is knownto be Is still at homeand attendsprimary school. The particularlyfine, have now begunto producecloth to sell to highl A older children,girls of 15 and 17, have both minorities,who eitherdo not traditionallyweave or have given up leti hometo work as domesticsfor Kinh people weavingand favour hand-wovencloth. Whileweavers in Bac Bin nearby.They have contractsof one year and still producecloth for themselves,they also now travel long dista iAn 1-1.5million VNDper year.All four to SongBe and Dalatto selltheir cloth to M'Nong, Stieng,Raglai caildrenare illegitimate-it is unclearwhere and Ba Na customers. their fathers are living but they take no responsibilityfor her children.She works3 sao make some improvementsin their household of land but, as she has no buffaloto breakthe welfare in recentyears. In the Cham Commune land and plough,and no one in the villageto of Phan Hoa for example, it was reportedthat help her, she must rent machineryto do this job and this reducesher production.She also some householdshave significantlyincreased worksfor other farmers,eaming 15,000 VND their stock of animals (cows, goats) and have per day. Incomefrom her daughtershelps to managedto commercialisetheir handicraft supportthe wholefamily. production(see box above).

However,the poorest householdsin lowlandand coastalareas lack the resourcesto improvetheir economicwell-being, and are particularly vulnerableto economicshocks. Those householdslacking capital to invest in agriculturalproduction are among the poorer householdsin these areas. In addition, householdsin lowlandand coastal areas mustcover the costs for educationand healthcare,18 which poorer householdsstruggle to meet. Sudden expendituresto treat illness,or chronic ill health in the family, is a severedrain on the resourcesof such households.The poorest householdsin lowlandareas are likely to be indebted, and to resort to strategiessuch as land rental to cope, thus reinforcingthe cycle of debt.

As in the uplands,wage labour is usedas a coping strategyfor poor households,but creates increasingopportunities for the young and better-off to leave the community.Wage labour in poorer householdstakes young husbands away from their households,leaving wives to cope alone with raising childrenand agriculturaltasks; it also takes children of poor householdsout of school and into the working world at early ages. As in the uplands,key adults within the householdsmay be away from decision-makingwithin the communityover long periods of time. In many householdsit is becomingincreasingly common for young peopleto leave the communityon a more permanentbasis, seeking work in the cities. Better-off households,however, are not weakenedby the departureof a family memberas are the poorer households;rather, it is seen as an opportunityfor that person to make a better life elsewhere.Informants in Loc Binh Commune(in Thua Thien Hue) said that migrationto the south has becomeeven more popularfrom that communein the wake of the severe floods in November 1999. => Coastal and lowland communitiesare disadvantagedoverall by poor agricultural opportunitiesand harsh climatic conditions.Communes within new economic zones, built on marginal land, may feel the worsteffects of these poor conditions.

18 Thesetotal costsare higherthan in uplandareas, where minorities are exemptfrom tuitionfees (although'contributions' for school upkeep,books and clothes are still considerable)and somehealth costs.

27 => Withinthese communities,some householdsare managingto slowly develop their householdeconomies.

2.2 EthnicMinority Issues

2.2.1 Ethnic MinorityGroups in the ProjectArea

As a general rule, geographicremoteness corresponds to the degree to which minoritygroups have contactwith the Kinh majority.In turn, this influencesthe degree to which they are familiar with the Vietnameselanguage, Kinh cultural norms, and of course, governmentpolicies and programmes.

Physical space distances ethnic groupsfrom the Kinh majority; but so too does cultural difference.The Dao and H'Mong can be put with some confidence at the most isolated end of the spectrum;so, arguably,can many Co Tu and Xe Dang communities.At the other end of the spectrum are the Cham, who have lived for centuries alongsidethe Vietnamese,are very familiar with Kinh norms and share similar livelihoodstrategies, but retain a distinct languageand cultural identity.

The table below shows the ethnic groupswithin the project area accordingto the terrain they customarilyinhabit, and the level of familiarity with the lowland majority.The geographicalclassifications used are those commonlyaccepted by governmentto classify communes.

One of these categories, 'high mountainous'or 'deep-lyingand remote' requires some clarification.The H'Mong and Dao in Thanh Hoa and Nghe An tend to inhabit remote high mountainareas. In the south-centralregion, however, 'remote and deep-lying' may refer to inaccessibilityrather than elevation:this is particularly evident in plateauareas in Binh Phuoc and parts of Lam Dong. The degree of cultural remotenessamong such people is still high, but perhaps not as high as among the Dao and H'Mong.

Table11: EthnicMinority Groups in the ProjectArea Terrain 'Deep-lyingand Remote' Midlands** Lowlands Coastal (High Mountainand/or areas with poor access) Cultural 4 > Features greater physical greater contactwith Kinh live close to Kinh and cultural isolation but culturally distinct

NORTH-CENTRAL REGION _ Thanh Hoa Dao, H'Mong Muong, Tho, Thai NgheAn H'Mong Muong,Tho, Thai, Kh Mu, 0 Du SOUTH-CENTRAL REGION Thua Thien Bru-VanKieu, Co Tu, Ta Oi, Pa Co Hue Quang Nam Xe Dang, Co Tu, Xie Trieng, Co, M'Nong Quang Ngai Xe Dang, Co, Hre Binh Dinh Ba Na, Hre Phu Yen Ede, Ba Na, Cham Hroi Khanh Hoa Ede, Rac Glai, Xie Tneng Ninh Thuan Rac Glai, Gia Rai Cham Cham Binh Thuan Rac Glai, Chu Ru Cham Cham Lam Dong Co Ho, Cil, M'Nong, Ma, Chu Ru Nung*, Tay * Binh Phuoc M'Nong, Stieng Tay*, Nung* * denotesin-migrants. **North-central region only N.b. includedin this table are 30 ethnic groupsand nine 'smalllocal groups' (followingDang Nghiem Van, 2000) sources: SocialAssessment and MPI questionnaire

Althoughthere are 54 officially recognisedethnic groups in Vietnam, local ethnic identitiesare much more numerousand more complex. Local distinctions exist to subdivide the larger ethnic groups; in turn the ways in which peopleview their own identitiesare constantlychanging. At local levels: * People may use the name of their subgroup,but deny any associationwith the larger group within which they are officially included:examples of this are the Pa Co (within the project area they can be found in A Luoi District, Thua Thien Hue), supposedlya subgroupof the Ta Oi, but who considerthemselves as a separate group; and the Cil (in Lac Duong District in Lam Dong) who insistedthey were neither Co Ho nor M'Nong-the two groups into which ethnologytextbooks place them- 19 but purely Cil.

Table 12: Small Local Ethnic Groups in the Project Area Small Local Group Main Ethnic Group Ca Dong (or Ka Dong) Xe Dang Cham Hroi Cham Chau Ma Ma Cil Co Ho or M'Nong Dan Lai Tho Pa Co Ta Oi Pa Hy Ta Oi Rut Chut To Rin Co Ho source:Dang Nghiem Van, 2000

Ethnic groups may accept a commonly-usedname they wouldn't use among themselves in order to simply things for outsiders:for example,while the terms Man and Meo are commonly used among lowlandersto refer to the Dao and H'Mong respectively,some Dao and H'Mong do not like the use of these terms (both are considered derogatory);other people from the same ethnic groups will use these terms with outsiders in order to simplify matters.

While minoritygroups are ethnicallydistinct, they have many cultural and linguisticsimilarities. Some of these similaritiesmay be rooted in a common ethnic origin. Many groups can mutually understandeach other; they may also have similar customs or even shared traditions and festivals. The Muong, for example, are considered by some ethnologiststo be 'highland Kinh' who went to make a living in the mountainswhile their Kinh relativesstayed in the lowlands, and their languageis very close to Vietnamese.Similarly, the Rac Glai are the highlandcousins of the lowland Cham, and the two languages,it was claimed in a Cham commune,are

9 see Vien Dan Toc Hoc, Cac Dan TocIt Nguoio Viet Nam,pp. 129,306.

29 mutuallycomprehensible. Their cultural ties are also evident in religiousceremonies and festivals in which both groups traditionallyparticipate. Ethnic groups within the same languagegroup may share importantforms of social organisation;these may also, however,be found among groups which are not linguisticallyalike, as describedbelow.

Similaritiesmay also be rooted in generationsof interaction,even if there is no commonethnic link. Groups which live in close proximitytrade and learn innovationsfrom one another. They may have acquired methods of cultivation,forms of dress and customsfrom one another. For example, Kh'Mu women (though mainly the older women now) wear virtually the same dress as the Thai in Nghe An and Thanh Hoa-a slim embroidered skirt and tight bodice. In this case, Box12: AdoptingWet RiceCultivation in Tuong though, there is no ethnic link Duong betweenthe two groups.Rather, the HopThanh is a H'Mongvillage that was very recently settledin Xa LuongCommune inTuong Duong District, Kh Mu live alongsidethe Thai,Kh NgheAn. The villagersmoved from an area where they Muwomen do not weaveand thus farmedonly upland crops, and are nowin a location they purchasetheir cloth from wherethere theypurchase the ~~~~~~onlytwohouseholds are possibilities have begun forgrowing to do this. wet Whichrce. Soare far, Thai.Minorities of the Tay Nguyen thefirst householdsin thevillage to do so?'My own, and plateau(the Xe Dang,Gia Rai, Ba oneother household," says the village vice-leader. They Na, and Ede in the westernuplands haveshaped terraced fields Na, ~~~~~~~~~~Andwhydid they begin before andthe dug others? canals "Becausefor watering.my of the south-centralproject region) wifecomes from an area in Ky Sonwhere they planted havecultural influences that go back terracedfields. She knows how to dothis, and she is historicallyto the domination of the showingthe rest of us." Cham under the Champa kingdom. (LeThi Phi Van/Natasha Pairaudeau) This processof one group influencinganother-including but not confinedto the influencesthat the Kinh majority have on these other groups-is ongoing (see box above).

In areas of in-migration,there are ongoing changes in ethnic identities. Many of the northern minoritiesnow found in the south-centralregion of the projectarea, in Lam Dong and Binh Phuoc Provinces(Tay and Nung, but also Dao and H'Mong migrants), come with many of the same ambitions as their Kinh counterparts.There is some suggestionthat they may be taking more marginal tracts of land than their Kinh neighbours-but there are other reportsthat some are beginningto consider themselves as migrants first and minorities second.

=> There are 30 officially recognised ethnic minority groups in the project area. > In general, the cultural remoteness of these groups from the Kinh majority is also a function of their physical remoteness from lowland communities;however, there is also one group (the Cham) which is ethnicallydistinct from, but has lived alongside, the Kinh for many centuries. z Although distinct, there are many similaritiesbetween the ethnic groups, particularly those which share a common ethnic origin or have lived in close proximity to one another for long periods. Thesesimilarities provide opportunities to be built upon for planning purposes.

in 2.2.2 Ethnic Minoritiesas Stakeholders

2.2.2.1 GovemmentApproachto Ethnic MinorityDevelopment The fixed settlementprogramme, as described in the previoussection, defines the broad strokes of the government'sapproach to minority development.This includes discouragingslash and burn cultivation,while encouragingmore stable and settled agriculturalpractices for ethnic minorities;it involves taking steps to integrate ethnic minoritiesinto the economic and social life of the nation, by providing incentivesfor them to grow cash crops and bringingthem closer to social services such as health care and education.

In many areas this approach is highly appreciated by the people it targets, and great changes have taken place in many ethnic minority areas in recent years. Many minority groups practisingslash and burn cultivation have themselves recognisedthe limitations of these practicesin situationswhen there are too many people and too little land; many minorities also view an education in Vietnamese, and greater contact with other services of the State, as being importantto their children's future. In the field study, several villages reported volunteeringthemselves to be resettled to areas where irrigation is feasible, and where there is greater contact with social services.

However,there are also a number of shortcomingsto the Government approachto ethnic minoritydevelopment. The governmentapproach begins with the notion that many ethnic minority practicesare 'backward'(lac hau).20 This term is written into policy: the general objective of the 135 Programme,for example, reads:

'To quicklyimprove the materialand spiritual life of the ethnicminority people in the mountainous, deep-lyingand remote communes with special difficulties; create conditions for theseareas to overcomepoverty, backwardness and underdevelopment and integrate themselves into the overall nationaldevelopment, this contributingto the maintenanceof socialorder and safety, national security and defence'21

This approach implies that minoritiesare less advanced than, rather than culturally differentfrom, the national majority. In practice, it has resulted in the approaches and preferencesof the majority taking precedenceover the needs and prioritiesof ethnic minorities. Thus, for example, minorities who live in stilt houses are encouragedto build 'modern' houses on the ground, although their own architectural designs are more suited to the conditions in which they live. Moreover, many minorities have internaliseda notion of their own people as being 'backward'.

Furthermore, this approach ("improvingthe spiritual life of minority people") discourages beliefs and practiceswhich are felt to hinder the progress of minorities and their integrationinto national life. For many years, local cadres have acted on the basis that there are good customs (phong tuc), to be kept, and bad customs or superstitions (me tin) to be done away with. The motives underlying this approach may be well-meaning,as the example in the box below suggests.

20 In English,the term has a veryderogatory connotation; in Vietnamese,opinions vary as to whetherit is derogatoryor neutral. In the field visits, however,the researchteam agreedthat they wouldfeel uncomfortablereferring to minoritypeople as 'lac hau' in their presenceas it would be offensiveto them. 21 DecisionNo. 135/1998/QD-TTg of July 31, 1998to Approvethe Programmeon Socio-EconomicDevelopment in Mountainous,Deep-Lying and RemoteCommunes with SpecialDifficulties. However, in some areas these changeshave been imposed rather than encouraged; as a consequence,many ethnic minority groups are hesitant to speak of, or defend, their beliefs and traditions.

Box13: Illnessat theirHeels Inthis remoteXe DangCommune, illness is a greatthreat. In early1998, acute malaria caused nearly 100 deathsin the commune.At thetme therewas not yet a healthclinic or healthworker in the commune(a communedinic wasbuilt in late 1999).Faced with carrying seriously ill patientsa two-daywalk down to the districtcentre, with the risk of them dying on theway, most people tumed to traditionalexplanations and treatments,and thus to sorcerers,who cure patents of illnessby riddingthem of the particular'ghost' that has afflictedthem. These 'ghosts' are so feared-andwho would not fear a forcecapable of strkingdown 100 peoplewith fever-that in thepast, communities have sometimes moved away from the areawhere the ghost struck,in orderto avoidfurther ill fortune. The challengesto a programmeof fixedsettlement, then, are to makeeffective treatments accessible to peoplein remoteareas...and, some believe, to doaway with superstitions-sorcerers who have 'the ghost on the tip of theirtongue to threatenlocal people and make them move. NguyenVan Unh/Natasha Pairaudeau

> The CBRIP is being implementedin a context where govemment development approachesoften fail to value and make use of the local knowledgeof ethnic minorities.

2.2.2.2 Ethnic Dynamics within the Project Communes Communeswithin the project areas may be inhabited by a single ethnic group or a mix of ethnic groups. They may be: * Ethnically homogenouscommunes: Greater communitycohesion is the rule, although when some villages are very remote they may still regard themselvesas not closely attached to the centre. In addition, people of the same line or clan may have intereststhat diverge just as much as, if not more than, peoplefrom different ethnic groups. In remote communes,villages may be so distant from the commune centrethat it may be more convenientfor them to use a health clinic in another commune. Moreover,they may considertheir intereststo be quite apart from those of peoplein the communecentre to whom they regard themselves as only administrativelyattached, regardless of any ethnic tie betweenthemselves and people in their communecentre. * Ethnically heterogeneouscommunes in the north-centralregion: These are commonlycommunes where one ethnic group inhabits a lower elevation,which for logistical reasonshas becomethe commune centre, with the groups that inhabit higher elevations being more removed from the communecentre. In some communes the centres may be inhabited by Kinh people. Interestsof the different ethnic groups-who for the most part form their own villages-are likely to diverge. * Ethnically heterogeneouscommunes in the south-centralregion: Communes inhabited by large numbersof Kinh people, or ethnic groups from northern areas, suggest recent in-migration:as explained in the previouschapter, the interests and attitudes of these groups are likely to diverge considerablyfrom the interests and attitudesof indigenousgroups. Between indigenousminority groups in mixed communes in these areas, there may be greater levels of cohesion and consensus,as these groups will have lived alongside each other for longer periods of time, and be experiencingthe same difficulties. Some communes may be virtually homogeneous,except for very small numbers of ethnic groupswithin the populationthat are differentfrom the

32 majority ethnic groups. Unless these are northern minorities in communes in the south-centralregion (again, suggesting in-migrationto the communes),this indicates in-marriageof neighbouringethnic groups into the majority group within a commune.

In general, in implementationof the CBRIP, communityconsensus can be expected to be stronger in ethnicallyhomogenous communes, where community consultation will be assisted by a commonlanguage and shared cultural assumptions.However, ethnic links and shared histories also create similar interestsand shared assumptionsbetween different ethnic groups. Conversely,ethnic homogeneitydoes not guarantee consensusand shared interests within a community.

2.2.2.3 Representationof Ethnic Minorities within the Administration Although it is commonly held that ethnic minorities are under-represented in local administrations, data from the questionnaire does not support this view. Rather, it-showsthat the percentageof Kinh cadres in the projectarea is low in comparisonto the percentageof Kinh in the project area overall.

Table 13: Representationof Kinh Cadreswithin the Project Area In ProjectArea Percentage Percentageof Kinhpeople 51% Percentageof Kinhcadres 30% source:Social assessment questionnaire

Furthermore, the field study suggests that looking at representation merely in terms of ethnicity may be misleading. Havingnon-minority cadres in key positions in communePeople's Committeessometimes does reflect control being actively taken out of the hands of ethnic groups. This occurs especiallyin situations where cadres are from outside the communeand therefore have less personal commitment to the commune. In other cases, though, especially in remote communes,Kinh cadres have reached their positions having spent long periods within the commune (and often have marriedwithin the community). Local ethnic minority communities themselvesoften acknowledgethe need for administrativeand Vietnamese language skills: in some communesin remote areas, minoritycadres have themselves virtually 'contracted' the very few-literate-native speakers of Vietnamese in their midst to help them figure out regulationsand write administrative reports, becausethey have difficulty carrying out these tasks. These same people may eventuallyfind their way into the administration(see box below). They are likely to have a good understandingof communityneeds, and a much better understandingof the cultural prioritiesof ethnic minoritieswith their respective communes,than more recently appointed Kinh cadres.

33 Box 14: 'Seconded'to the CommunePeople's Committee The researchteam to Tra NamCommune (Quang NamProvince) found that severalKinh schoolteachers-all of whom have been living in the communefor over 10 years-have been informally'seconded' by the local Xe Dang peopleto help them with administrativetasks, and someof them are now being moved,with local support,into positionson the People'sCommittees or People'sCounals. Tra My Districtauthorities reported that in other communesin this district (within the projectarea: Tra Don,Tra Cang, and Tra Leng),school teacherswho are long-standingresidents have been similarlyelected to administrativepositions.

Mr. B. is a Kinh teacherwho came from Thang BinhDistrict up to Tra Nam Commune10 years ago, and stood as a candidatefor a seat in the communePeople's Council in late 1999. He was electedwith 90% of the vote. Mr. NguyenNgoc Lu, the (Xe Dang)president of the communePeople's Committee, remarked: 'Headmaster B. is a universitygraduate who hasten years of experienceworking in this area. It is certainthat B. can be a great help to the Xe Dang community[90% of the communepopulation] in implementingthe CBRIP,because he is respectedby all the communityand he has the best educational,social and cultural knowledgeof anyonein the commune.He will be a great help to us in his position".(Dau Quoc Anh/NatashaPairaudeau)

2.2.2.4 Languageand Literacy Literacy levels in Vietnameseamong ethnic minoritiesare low.

Few ethnic minoritieshave a written languagein which wide numbersof people are fluent. Those that have relativelyhigh levels of literacyin their own languages,such as the Cham, also have high levels of literacy in Vietnamese.

Thai is often used as a common languageamong ethnic minoritiesin markets in the north-centralregion of the projectarea, although knowledgeof Thai may not be widespreadamong H'Mongwomen. Betweenethnic groups in the south- central region, Vietnamese is more likely to be used as the common language,even though knowledgeof Vietnamese,particularly among peoplefrom more remote areas, may be quite poor.

Many minority languages are related, but not necessarily mutually comprehensible. The chart below lists the 30 ethnic groups in the project area accordingto the languagegroups to which they belong:

Table14: Ethnic Groups and LanguageGroups in the Project Area H'Mong Sino- Tay-Thai Viet- Mon-Khmer Malayo- -Dao Tibetan Muong Polynesian H'Mong Hoa Thai Kinh Kh-Mu Ba Na Gia Rai Dao San Diu Tay Muong O-Du M'Nong Ede Nung Tho Ta-Oi Stieng Cham Chut Co-Tu Ma Rac Glai Gie-Trieng Co Ho Chu Ru Xe-Dang Co Hre Van-Kieu Bru source: Dang Nghiem Van et al., 2000.

=> There is little to be gained from translationof written materials into local languages. > For programmingpurposes, oral materials could be translatedinto Thai in the north-central region to cover a wide audience > In other areas, working in Vietnamesewith oral translation at local levels would be the best approachto languagebarriers.

34 2.2.2.5 Local Beliefs Althoughundoubtedly less prevalentthan they may have been several decadesago, spiritual beliefsare still held strongly by many ethnic groups. The reasonswhy many ethnic minority groups have difficultiesin expressingtheir own beliefs have already been mentioned.Across the south-centralregion of the project areas, in particular,there are prevalent beliefs about spirits. These spirits are understoodto be attached to location; peopleare careful to avoid certain 'polluting' acts in some areas in order to prevent offence to the spirits, thus bringingmisfortune. These beliefs make the placement of buildingsand other types of infrastructure (especiallyhealth clinics, as death, bleeding, illness and childbirthare though to be very 'polluting') rather sensitive.

> Although the degree to which spiritual beliefs are now important to local communitiesmay vary greatly, they remain an importantconsideration in implementingthe CBRIP. If local people are not properly consulted,infrastructure projects may not be fully successful,for reasons that may be difficult for outsiders to understand.

2.2.2.6 TraditionalLeaders Virtually every village visited for the field study had some type of traditional leader or gia lang. This includes minority and Kinh villages.

The Vietnameseterm 'gia lang' actually refersto a number of differenttypes of traditional leaders. Gia lang literally means village elder in Vietnamese.However, the gia lang the Social Assessmentteam met and interviewedwere sometimesthe oldest man in the village, and at other times a clan or other leader. They include,for example: D A Muong gentlemanin his 80s who was the village historianas much as a key decision-maker.He had served as the head of the traditional Muong administrativeunit, as had his father before him. - An ethnic Cil man in his late 50s in Lam Dong Province.He plays no role in the official administrationnow, but is a clan head and was (before 1975)the village leader, as was his father before him. He is still consultedon all important village decisions. * A Cham elder who is also the Party leader of the commune.The village leader is at the forefront of decisions made in the commune, he says, but he also comes to him for guidance. In this Cham Bani (Islamic)community, priests also play an importantguiding role within the community.

All these 'village elders' commonly have the respect and trust of their local communities. Men and women frequentlyspoke of their traditional leaders as having greater wisdom than others in the community.This had come to them through expertise living outside the commune (such as would be gained by war veterans), or through specialised knowledgeand education. The ability of gia lang to articulate Box15: Knowledge and Expernence. Theelder of a H'Mongvillage in NgheAn was also the the needs and priorities of their headof theFatherland Front; part of villagers'respect for communitiesat communelevel is him was owed to the fact thathe had been a soldier in limited. The 'constituency' of these thewar, and therefore had seen more of lifeoutside the men rarely reaches the village communethan most 1 amSihanouk!' he prcdaimed,by beyond way of explaininghis role; 'Everyonemust ask me level. Moreover,the extent to which advice beforedeciding anything.' The twovillage vice local people are willing to give their gia leaders(brothers) were both his nephews. A mucholder lang a public profile varies from manwas presented as the manwho knew the village history;he was also said to be highly respected. communeto commune.Some communitiesstated firmly that the role of gia lang is quite separatefrom administrativeroles, while other gia lang have been elected to People'sCouncils, or otherwiseincluded within the official channels through appointmentto Party or mass organisationposts. In some villages, a gia lang led the village meetingheld with the researchteam: this procedurewas justified in these cases by the fact that the gia lang was also the head of the village People's Councilor Party cell. In other villages, interviewswith gia lang were conducted outside village meetings.

Gia Lang are traditionalleaders of communitiesthat customarilydesignate decision-makingroles to men. All of the gia lang encounteredin the field research were men. Their attachmentto the more conservativeelements of their own traditions may mean that they are not necessarilyadvocates of stronger roles for women within their communities.

=> Despite the evidencethat traditional leadersplay an active role in articulating local needs and priorities, CBRIP designneeds to be very sensitiveto the fact that these leadersare not always givenhigh public profiles within their own communities. > The support of traditionalleaders withinthe CBRIPmay work in oppositionto the promotion of greater inclusion of some groups in the project cycle. 2.2.2.7 MatrilocalTraditions Severalgroups within the project area practicetraditions of matrilocalityand matrilineality.Matrilocality means that men go to live, upon marriage,in their wives' households;matrilineality is the practice of handing propertyfrom mother to daughter (rather than from fatherto son). In these circumstances,a young woman usuallyhas the power to activelychoose her partner (or, where arrangedmarriages still occur, of the bride's rather than the groom's side to make a proposal).

These practices are prevalentwithin several ethnic minoritygroups in the south-central region of the project area. In Cil and Cham ethnic groups included in the field study, men customarygo to live with their wives upon marriage, and women's property rights are actively recognisedwithin the community.Other groups have combinedpractices: these may be characterisedby residenceand property going to either sex, accordingto convenience,or by a maternal uncle holding an importantrole as decisionmaker in family matters.

22These combined practices are taken by someobservers to indicatethe gradualreplacement of matrilinealwith patrilineal practices(see Dang Nghiem Van, 2000). Table15: Matrilocaland MatrilinealEthnic Groups within the ProjectArea Practice EthnicGroups matrilocaland matrilineal(mau Cham, Cham Hroi, Cil, Ede, M'Nong,0 Du, Rac Glai he) patrilocal/patrilinealand Ba Na, Chu Ru, Co Ho, Hre, Kh Mu, Pa Hy, Pa Co, Ta Oi, Xe matrilocal/matrilineal Dang,Xie Trieng

The custom does not extend to women being the leaders within the community. During the field study, the Vietnamese term used to describe this tradition was mau he, meaning matriarchy; however, these communities are not really matriarchal. For example, traditional elders in both Cham and Cil communities are male; Cham religious leaders are male and men clearly play a leading role in community decision-making. It was observed anecdotally, however, in both Cham and Cil communities visited, that women within these communities are particularly confident and active within the community.

Box 16: BuyingHusbands, Selling Sons Mrs.K'Tu is fromthe Cilethnic group and lives in a communeset in lowhills roughly two hours' drive (or, for her,a day'swalk) from Dalat City. She is 41 yearsold and hasnine children. Her family residence card (hokhau) is in her husband'sname, as is theirred bookrecording the family's agricultural and forestry plots.This is despitethe fact that in ethnicCil custom,she says, Men have to followtheir wives (phai theo vo).She was the only daughter in her parents'house, so by Cil customshe shouldhave taken her parents' house.Her parents did notagree with hermaniage, however, so sheand her husbandeloped, and bought landfrom his parentson which to builda house.When asked if it is moredifficult to raisegirls or boysin her community,the respondsimmediately, Girls, of course,because you haveto buy sons(phai mua con trai)!'.Couples who have only sons, by contrast,have 'lost their root' (mat goc) and are in a weakposition as theylose their sons at marriage.Her sons, however, often retum to her householdto helpout. She is happyto havegirls: when they marry and have children, the house is full,busy, and happy. Yes, she says, divorcesare not unheardof in thiscommune, but people rarely look to thelaw whenthey divorce. ff a man goes,he goes,but the womenkeep the children and the propery-he leavesempty-handed. NguyenThi NghiaAVatasha Pairaudeau

Significantly, this practice makes women sometimes less vulnerable than in patrilineal and patrilocal traditions when the family breaks down. Matrilineality and matrilocality strengthen the circumstances somewhat for unwed mothers, divorcees and young widows. Importantly, divorced or abandoned women within these traditions keep their children and their property, unlike in patrilocal traditions in rural areas where women may have to leave their children to their estranged husbands, and move back to their own parents in their home villages. Young widows in matrilocal communities may still be faced with raising a family on their own, but, instead of living with their in-laws, they remain with their own relatives. Unwed mothers may meet with the disapproval of the community, but they are more likely to have a property base of their own.

Women's customary property rights, however, are not recognised by the state. In communities visited for the field research where mau he is practised, ho khau residence registration books name the husband as the household head; red land allocation books have also been filled out in the husband's name.

= Matrilineality and matrilocality are customary forms of social organisation that strengthen women's position within the community and provide good support to women in vulnerable positions. It is in the interest of the CBRIP to recognise and support such practices wherever possible.

37 2.3 Poor and Vulnerable Groups The first sectionof Part Two has already describedthe types of households that can be expectedto be poor or vulnerablewithin the different geographic regions of the projectarea. These include: * Householdswhich lack the resourcesto succeedin intensiveand/or commercial methods; * Swiddenfarmers; * Householdsand villages marginalisedby distance; * Indigenouscommunities in areas of in-migration; * Inhabitantsof new economiczones.

Thesecond section has shown reasons why ethnic minorities can alsobe countedamong the most poor and vulnerablepeople in the projectarea.

In addition,some types of householdsmay carry a heavier burdenof poverty than others.This is especiallytrue for those householdsthat do not fit in with social norms. It is often these social circumstancesthat contributein the first place-either directly or indirectly-to poor accessto resources.

Includedin this categoryare certain types of women-headedhouseholds. This includes: UnwedJ mothers: Women who have been left by the fathers of their children to raise the childrenthemselves-with no legal acknowledgementof the relationship-are ostracised,to varying extents, by their own communities.In additionto this, they often face the practical difficultiesof raising childrenwhile being the sole adult labourer in the household. 'Young widows: If they are living in separatehouseholds from their parents (or in-laws),they face many of the same challengesas unwedmothers in supporting a family on their own; they are likely to receive more social and moral support from their communities,however, than unwed mothers.Once widowed,rural women are unlikelyto remarry. * Divorcees:Divorce is rare in rural communities,for the very reason that it leaves rural women with few viable alternatives.When rural women do divorce, they customarilyreturn to their own parentsor home villages.2 3 Althoughsome rural divorces are conductedwith referenceto national law, with divorceesobtaining an equal share of property and custody of children, it is not uncommonto find women who are disenfranchisedfrom the communityand have had to leavetheir children with their husband'sfamily. Such cases may be particularlyprevalent among more remoteminority groups (with patrilocaltraditions) who rely to a greater extent on customarypractice rather than civil legislation.

In addition,social norms,age, and acceptedgender roles within the household result in some householdmembers bearing the bruntof poverty within their own householdsmore than other householdmembers.

23 In communitieswith matrilocaltraditions, women are likely to be lessvulnerable should they be divorcedor widowed.This is explainedin furtherdetail in the sectionon ethnicminorities.

38 Women in all householdscarry the heavier burdenwithin the household in coping with poverty.The dual burdens of agriculturallabour and houseworkfall overwhelminglyon the shouldersof women. Among other problems,this leads to physical overworkand limitedtime for social activities or participationin meetings and decision-makingfora. Those who must sell wage labour earn less than men, and considerablyless in lowlandand coastalareas (see table below).

Table 16: WageLabour Rates by Genderwithin Project Communes Zone Men(VND) Women(VND) highand remote 13,973 10,771 Mountainous 14,703 11,745 Lowland 13,765 10,000 Coastal 12,800 10,263 averagerate 13,810 10,695 N.b.rates given for oneday's labour excluding lunch source:Social Assessment questionnaire

Children often bear the brunt of poverty in poor rural households.In upland ethnic minority areas, children have difficulty continuing school beyond primary levels due to languagedifficulties (they must often repeat classes),distance from secondaryschools, and household labour demands.In lowlandareas, labour demands as well as higher costs of education prevent poor childrenfrom continuing their educationto higher levels. In turn, children are the first to suffer the effects of poor nutritionand poor access to adequatehealth services.

The disabledare likely to be among the poorest members of the community. In rural communitiesthe disabled are unableto contributeto householdwell-being to the same extent as the able-bodied.They often find it difficult to marry, remainingin their parents' householdsinto adulthood, or, if they do marry, they may find partners who are disabledthemselves. These individuals,but also these types of households, are routinelyamong the most disadvantaged.

Elderly households,although rare in rural areas, may be particularly vulnerable. In upland communitieswhere traditions of extendedfamilies remain strong, it is uncommonto find elderly couples living apart from younger generations, but in some cases this may occur. The departure of younger generationsfrom poor lowland and coastal communitiesto seek work elsewhere results in greater numbers of elderly householdsin these types of project communes.

/nI order to ensure project benefits reach the most marginalised,the CBRIP must explicitly address the needs and priorities of the poorest and most vulnerable householdsand people in the project area. => Rural poverty is often associated with a lack of available labour within the households,and thus the poor householdsidentified here will often be the least able to spare labour to work on the CBRIP. PART THREE: KEY PLAYERS IN INFRASTRUCTURE DEVELOPMENT

PartThree describes the rolesof key players in implementinginfrastructure works within communes.Different players are involvedin infrastructureworks, depending on whether the works are: . Formalgovernment projects, built with externalfunding and assistance; . Works financedand implementedby local people,but motivatedby government;or . Works financedwithout government fundingand built entirelyby local people.

The first sectionbriefly describes these three typesof infrastructureworks, and aims to clarifywhat is meantby the term 'capacity'in relationto these works. In the secondsection, key playersin the differenttypes of infrastructureworks are identified,strengths in implementing,operating and maintaining governmentor externally-fundedprogrammes are highlighted,and elementsintegral to successfulorganisation and implementation of infrastructureworks from grassrootslevel are emphasised.

4n

3 PART THREE: KEY PLAYERSIN INFRASTRUCTURE DEVELOPMENT

3.1 'Projects'and Local Works

When authoritiesand local peoplewere asked to speak about the small-scale infrastructureworks with which they had been involved,it became evident to the researchteam that 'projects' (du an) were being automaticallyunderstood as governmentinitiatives, while many other works were going completelyunmentioned. The researchteams walked along footpaths,crossed simple bridges,and saw traditionalcommunal houses and basic village classroomsthat local peopledid not think to count as 'infrastructure'.Once the teams expressed an explicit desire to discuss 'things peopledo themselves',as well as governmentprogrammes and those projectsthat lie somewherein-between, the picturechanged significantly.

Infrastructureworks within the communesfall roughly into three differentcategories:

1. Projectsbuilt with external assistance.These include those funded through governmentprogrammes, those fundedwith bi- or multilateralassistance, and NGO projects. 2. Locally-financedworks motivatedby governmentpolicy. This includes physical infrastructurethe constructionof which has been motivatedby governmentcommitments to provide the necessaryservices. Commune schools and village classrooms2 4 were the main examplesencountered in the field study. 3. Works financedand built entirely by local people.Examples of such works were found in every communevisited in the field study. Simple,locally-built infrastructureworks are constructedwith no referenceto governmentfor several reasons (see box below). Box 17: What Motivates People to Built their Own Simple Infrastructure? Simple infrastructuremay be financedand built entirelyby local people: * to fulfil customary community religious or cultural purposes: suchas communalhouses (nha rong) or religiousbuildings; * as traditional schemes with local technologies and materials: these, suchas bamboogravity drinkingwater systems,are the precursorsto the physicalinfrastructure now providedin other places through governmentprogrammes; * in the absence of govemment infrastructure construction: communeswhich do not yet have electricalconnections, for example,and villagesand householdswhich cannotafford the costs of connectingfrom the communecentre, may use small,privately funded hydropower generators to bring powerto their communities.In someplaces these amountto privateelectricity networks running throughoutthe village.Wells are also often built privatelyby individualhouseholds, or by clustersof households;or * because they are not funded under govemment programmes: bridgesare includedby govemment within roadnetworks, but manyremote upland villages are accessedby crossingstreams or rivers situatedoff the main road networks.Wooden or bamboobridges (or sometimesferry services25) are sometimesconstructed by the local communityin these cases. Footpathsare not routinelyfunded through govemmenteither, and are thus clearedand maintainedentirely at local level.

24 Primaryschools covering grades 1-5 are located in or nearto communescentres. Some communes have lower secondary schools(grades 6-9) but this is lesscommon. Increasingly, classrooms are being built in villagesin orderfor youngchildren to beable to attendthe lowest grades without having to travellong distances. Village classrooms (lop hoc) are satellites of the centralcommune school (truong hoc), and teachers in villageclassrooms are supervised by the headteacher of thecentral school. 25 A servicealso organised and paid for by villagersthrough local contributions, in the casewitnessed by the researchteam.

ZI While the capacity to implement, operate and maintain government or external projects was relatively low in many communes, another type of capacity-the local capacity to organise and implement infrastructure works from grassroots level-was quite strong in communes visited in the field study. Moreover, communeswith no road access (or, to a lesser extent,seasonal road access),where the benefits of infrastructureprojects and economictransition have been the slowest to reach,are the very regionswhere self-reliance,and the incidenceof simple works organisedwithin the community,are the highest.

> An opportunityto build uponlocal self-reliance exists alongside the needto improvelocal capacity to implementgovemment projects.

3.2 Key Players in the Different Types of Infrastructure Works

3.2.1 In Projects Built with External Assistance

These projects are most often characterised by plans and priorities that have been determined from higher levels before they are implemented in the villages. The exceptionsare some NGO projectswhich tend to operatemuch more closely at communeand village levels, beginningwith the drawingup development plans.

District Departments of Planning and Investment currently play the lead role in planning infrastructure projects within their respective communes. They rely on their respectiveInfrastructure Bureaux at district level (Phong Giao Thong va Xay Dung Co Ban) for implementation.District-level People's Committees are in charge of project managementthroughout the project cycle,and villages and communes must refer to them for any maintenancethat cannot be done at local level.

Cadres working within district Infrastructure Bureaux who were interviewed in the field study were technically trained, but inexperienced in community development approaches. Moreover,there appearsto be little co-ordination betweentheir technicalwork and social mobilisationefforts of mass organisations.

Evidence of very limited implementation of government programmes in the most remote communes was attributed in such communes to the low motivation of district-level authorities to travel to these communes. There are no financial incentivesfor district-levelcadres to reachthe distant communes,and long periods away from the district may take them away from small-scale businessor other income-earningactivities used to supplementstate salaries.

As a consequence of top-down management and directives from district level, Commune People's Committees lack experience in management of government projects. However,there is widespreadrecognition at district level that more responsibilityneeds to be placedin the handsof the commune. Districts, however, are unsure how to go about this. Many district level cadres expresseda wish to hand greater responsibilityover to communes,but are uncertainabout communes'abilities to take on this work. Those districtsthat have tried to hand

42 responsibilitiesto the communeshave in some cases taken it back again as they felt the communeswere not handlingthe duties well. A top-down approachalso creates heavyworkloads for the district. Said one InfrastructureBureau chief, "I do everythingfrom design to audit-but I would preferjust to manage."

Nonetheless,the basic structuresare in place to pass responsibilityfor management,implementation and accountingdown to communelevels. Communeshave accountsin districttreasuries, and their own accountantsat communelevel. They may also, as mentionedbelow, have personswho work at communeand village levels on ad hoc managementand inspectionboards.

On a project-by-projectbasis, a personor personswithin the communeare appointedto be commune-levelrepresentatives (can bo giao thongva xay dung) of the district InfrastructureBureaux. Those appointedto these positions are Commune People'sCommittee members(s), who work part time and are paid expresslyfor this work (in Thanh Hoa Province,payments of 40,000-180,0000VND per month were reported).These peopleare primarilycontact personsat local levels, responsiblefor organising local labourcontributions.

In some projects,there exist inspectionboards, with representativesfrom village and communelevel. This is in line with Decree 24. However,these did not exist in every village nor for every project, and in many cases inspectionwas being done by cadres alone.

Communemass organisationsdo not appear to be currently involvedin any way in infrastructureprojects, except in some NGO projects.They do however play a role in mobilisingand informingpeople on other issues;the Women's Union, Youth Union, and FatherlandFront were repeatedlystated to be the most active mass organisations.

Positionson People's Committeesand in mass organisationsin some communesare held by a small group of closely related people.The research teams visited several communes in which the head of the Women's Union was the wife of the Commune People's Box 18: PowerSharing in the Village Committeechairman, and other In this H'Mongvillage, the village leader'swife has been relativesled the mass organisations. appointedas the head of the Women'sUnion. The village The sameThe patternpattensam was often ften ntednoted atFarmers'at rleaderhimselfAssociation, is thehead the Youthof the UnionFatherland ... and Front, the Veterans'the village level. While it is clear in some Association.He has the highestlevel of literacyof any communeswhere cadres are closely personin the village, and he daimed that for this reason, relatedthat they represent the he is the only one really capableof holdingany of these interestsof a narrow group of people, in some minority communeswhere proficiencyin Vietnameseis low, this matter is more difficultto judge (see box above).

Women are especiallypoorly representedwithin formal institutionsat village and communelevels. It is very rare for women to hold positions other than posts within the Women's Union.

4 1 Table17: Women'sRepresentation in Commune-LevelInstitutions Overall percentageof female leaders in projectarea 12% Percentageof communesin the projectarea with more than 2 female leaders 10% Highestnumber of female leadersin a projectcommune 4 Lowest numberof femaleleaders in a proiectcommune 0 source:social assessment questionnaire"

=> The will to hand greater responsibilityfor implementationand managementover to commune-levelauthorities exists in many districts, and should prove to be an asset in implementationof the CBRIP, as long as the project provides clear guidelinesas to how this responsibilitycan be handed over effectively. => Players exist at communelevels and below, whose experience-albeit limited-is appropriateto tasks required in implementingthe CBRIP from communelevel. This experiencecan and should be built upon. = Strengthsof mass organisationsin mobilisinglocal people for other types of activities are anotheropportunity for implementingthe CBRIP well. => Becauseof the risk that cadres at communeand village levels may only represent the interests of a narrow group of people in some communes,the inclusion of non-officialsin managementand maintenanceof the CBRIPis vital. 3.2.2 In Locally-FinancedWorks Motivatedby GovernmentPolicy

In most of the communesvisited, primary level classroomshad been built by local people using their own funds. The classroomsare built on the understanding that, if the local people provide the structure,district education authoritieswill provide a teacher.

Organisationand managementof these works is generallydone from village level. In the examples discussed in the field study, managementof these works was undertakenby village leaders, with the active involvementof teachers and parentsof school children in design and constructionof the classrooms,and subsequent maintenance.In some villages, only parents of school-aged children contributedto construction; in others, all householdswere asked to make a contribution.The quality of such classroomsvaried from very simple shelters (see photo below)to a sturdy wooden Thai stilt house which had been purchased (at considerablecost) by villagers from a household which had left the village. Photo: A very basic education:a teacher and his classroomin Tan Hoa Commune,Dong Phu District, Binh Phuoc Province.The classroomwas built by Tay and Nung in-migrants.(photo: Dau QuocAnh)

> This arrangement,when the provision of teachers can be guaranteed,appears to work well as a way of encouraginglocal people to construct and in some cases cover the costs of simple infrastructure.It also implies that without such guarantees,people will be unwilling to construct schools or clinics. => Villagesconstructing such works are able to develop workablemaintenance systems for these projects.

26 Responsesbased on askingcommunes to state genderof leadersof the People'sCommittee, People's Council, Party and mass organisationsat communelevel..

44 3.2.3 In Works Financedand Built Entirely by Local People

Most works financedand built by local people are organisedfrom village level, but they may also be organised inter-village,or by a whole commune.The latter is more common in communesof uniformethnicity, where a common cultural or religious need creates closer communitycohesion.

Box19: 'Respected People' Whenasked about the possibility of appointingor electingnon-officials to projectmanagement boards, respondentsin VinhSon and Tra My saidthat it is possible,and they do it alreadyfor worksthey construct themselves.Many people suggested that 'respectedpeople' (nguoi tin nghiem)who were trusted locally shouldbe selected for thesepositions. 'Respected people' include village elders (gia lang) and clan leaders(chu ho), but also people with high levels of educationincluding young graduates who have retumedto theircommunities. Such people may have publicly pointed out weaknesses in the administrationin the pastwhen others felt unableto doso. Saidone man, 'it is onlythese people who can preventcorruption.'

Examplesof this type of projectwhich were discussedin the field study often had well-organisedmanagement boards. Positionswere elected and the makeup of the boards dependedon the type of works constructed.Thus, to serve a cluster of Ba Na householdsin Vinh Son Commune (Binh Dinh), the householdsconcerned elected the most educated man among themselvesto chair the board, a woman as a cashier and a village elder (gia lang). As this was an arrangementbetween householdsthat did not extend to the whole village, it was felt there was no need to include the village leader. In Tra Nam Commune (Quang Nam), a bridge constructioncommittee to repair a bridge across a stream to serve a whole village includedthe village leader, a gia lang and several villagers who did not hold official administrativepositions.

Local consultationon these works tends to be extensive, and local people are well aware of many aspects of the constructionof these works. As they are organised and implementedfrom local levels, general knowledgeexists within villages about the works, and meetingsappear to be organisedand well-attended. The small scale of the works and levels of local awareness make it relativelyeasy to come to local agreementsregarding contributions and to assign responsibilitiesfor maintenance.

Although local architecturalskills were strong in many of the communes visited, many communitiesalso have experience of contracting outsidersto constructinfrastructure works. Ethnic minoritygroups may even contract Kinh peopleto build their 'traditional buildings' (such as some Thai and Muong minority houses and Cham mosques).

Photo: Eachof the three mosquesin this Cham Bani communein Binh Thuan Provincewas built with local contributions.Contractors were hired from outsidethe communeto carry out the work. Communeand villageauthorities managed and supervisedthe project.(photo N. Pairaudeau) Photo:A village meetinghouse in a Ba Na community,Vinh Son Commune,Vinh Thanh District, Binh Dinh Province:It was constructedin 1996 with local managementand communitycontributions of moneyand labour. The cost of materialswas coveredfrom two sources: people'scontributions (20,000VND/household) and the govemmentforest protectionfee.27 A village meetingwas held to

27 There is a total of 30 hectaresof forest undervillage protection;each protectedhectare earns the village50,000 VND per year.

AF; discuss the construction, facilitated by the traditional leader and village leader. Fifty households attended, both husband and wife in some cases. A management team was elected with a leader, vice-leader, and cashier. Kinh builders were hired to carry out the construction, and villagers actively participated by preparing the ground and transporting materials. Construction was completed within two months. The management team continues to guide people in maintenance of the communal house. The house is used for all village meetings as well as for traditional festivals. Youths practice every day at a football ground in front of the house. (photo: Dau Quoc Anh)

Levels of maintenancefor this type of works are high. For those works that rely on local architecturaland design skills, repairs are easily made to these works as the knowledgeof repairs exists at local levels and such repairs can be carried out entirely using local materials.

People may be more willing to get involved in projectsthey fund and build themselves,because they know what the money they commit to the projectis used for. Some informantsstated candidly that they were not particularlywilling to contributetheir labour or financial contributionswhen called upon to do so by commune cadres, as they perceivethat local cadres are rich through taking cuts from these programmes.

While these locally-constructedinfrastructure works operate very well from an organisationalstandpoint Box 20: Childrenare Vulnerablewhen Waters are High and involve high levels of Childrenare particularlyvulnerable when river crossingsare inadequate.A young girl pupil in a villageof Xa Luong Commune, local participation,local Tuong Duong District,Nghe An, had drownedin the previous rainy people expressed concerns seasonwhen the small woodenferry used to carry passengersto about the quality and safety and from the village had capsizedin high waters. One of the researchteams was strandedfor severaldays in a mountainous of some of them. With higher district of Phu Yen Provinceduring the torrential rainsthat levels of investment,they felt accompaniedthe secondmajor flood in centralvietnam in late they could for example, build November;the first newsto reachthis district of deathsdue to the y c floodingwas, tragically,of childrenswept away by flood waters.All bridgesthat would be safer of these childrenwere on their way to schools outsidetheir villages, than the precarious bamboo ones, piped water and electricitysystems that would require less maintenance,and better schoolhousesand classrooms.

Photo: Indigenous bamboo water supply systems in a Xe Dang village, Tra Nam Commune, Tra My District, Quang Nam Province, and in a Thai village, Xuan Phu Commune, Quan Hoa District, Thanh Hoa Province. (photo: Dau Quoc Anh)

Photo: Locally-made bridge in Son Thai Commune, Khanh Vinh District, Khanh Hoa Province. Seemingly precarious, this bridge, built by local Rac Glai and Xie Trieng communities, is nonetheless impressive: it spans 200 m and is 10 m above the river bed. (photo: Dau Quoc Anh)

In all interviewsregarding local works, though, people voiced strongly their satisfactionwith the process.The common characteristicsof these works are that: * They are initiated from local level; * Local people are consulted and better-informedabout these works than they are about externally-assistedprojects; * They make good use of local skills in constructionand expertise in design; * Although local resourcesare very limited, local people proved themselves very willing to make labour contributions,and to continuewith contributionsfor maintenance; * Local people were able, and willing, to make some small cash contributions.

aR Precedentsexist for the involvementof non-officialsin managementand maintenanceof infrastructureworks, and local people can easily identify capable people within their communitiesto do this. = Local-levelconsultation and high levels of awarenessof infrastructureworks within the communitycontribute to successfulprojects. = Experiencealready exists in many communesof contractingoutside expertise to assist in infrastructureconstruction. People with this experiencecan be identified and incorporatedinto the CBRIP. = In most communesthere is existinglocal knowledgeof constructionusing traditionaldesigns that are well-adaptedto local conditions,and of maintenance of these works using local materials. = Greatertransparency increases local people's willingnessto be involvedin project implementationand to contributeto maintenance.

47

Part Four: Participation Issues

In Part Four, participation issues related to the CBRIP are addressed. The first section examines how different beneficiaries (identified in Part Two) may benefit from infrastructureprojects more, or less, than others. The second section looks at the involvement of beneficiariesin each stage of the project cycle.

48

4 PART FOUR: PARTICIPATION ISSUES

4.1 Small-ScaleInfrastructure: Who Benefits?

The CBRIP is intendedto be 'community-based'.Thus, central aims of the project are to ensure that activities supported have broad-basedsupport of and ownership by the communities,and to make as many householdsas possible within the selected communesthe primary beneficiariesof the project. This requires special attentionto ensure that decision-makingis open, transparent and driven by the communities,and that projectbenefits reach all segmentsof communepopulations, includingthe poorest and most vulnerable.This chapter examinessome of the main challengesto achievingthese aims. Box21: Selection of ProjectCommunes Thetwo research teams were frequently placed in the 'lineof fire' wheni cameto discussingwith district authoritiesthe communeschosen for inclusionwithin the CBRIP. Mention by theresearch teams of the communesincluded in the project-whichdistrict-level authorities appeared in mostcases to be hearingfor the firsttime-were often met by expressionsof surpriseor dismay.It was notdear to districtauthorities why somecommunes on thelist of 'poorcommunes' had been chosen for inclusionin theCBRIP, while other communeson thissame list hadnot. In some districts, requests were madefor all poorcommunes to be includedin the project;in others,the claimwas madethat poor communes excluded from the CBRIPwere in factpoorer than those included.

ThanhHoa was the one province visited where clear criteria had been laid out by theDepartment of Planning andInvestment for inclusionof communesin the CBRIP.In that province, two large donor-assisted projects are planned,and 110communes regarded as the'poorest' communes in theprovince are includedeither in the WorldBank-assisted CBRIP (25 communes),or are servedby a CIDA-supportedproject. However,it shouldbe notedthat while some district authorities made convincing arguments for theirpoor communesto beincluded in theCBRIP, either now or at a laterdate, doubts were raised in otherdistricts as to districts'abilities to makethese choices wisely. One district chairman argued at lengthover the inclusion of oneof thetwo 'poor communes' in hisdistrict in the CBRIP,and not the other. Discussion with informants workingwith an NGOin the excludedcommune, though, reduced the credibility of hisposition on thematter. Theseinformants daimed that the exdudedcommunes already receive the lion'sshare of NGOand other donorassistance-so much so thatthey were frustrated by the overlapin the effortsof differentagencies, and whatthey regarded as a wasteof funds.The districtchairman's strategy, they claimed, was to 'saturate'one communewith aid, before moving on to othercommunes in need.They doubted the wisdom of thisstrategy. It has alsobeen mentioned at higheradministrative levels that, while consultation of levelsfurther down is important,final decisions on projectcommunes must rest finally with Hanoi, in orderto ensurethat decisions are notmade on the basisof biasor privilegeat morelocal levels. The provincewith the greatest uncertainty about which communes had been selected for theproject was (in the experienceof the SocialAssessment team) Binh Phuoc. Our contact with Binh Phuoc provincial authoritiesin relationto the questionnairerevealed that the provincehad drawn up a listof projectcommunes differentfrom the listwe obtainedin Hanoi(and had initially sent out the questionnairebased on thislist). Binh Phuoc'slist included all communeswithin the provinceclassified as 'poor'by MOLISAand CEMMA. Taken thatit is importantto ensurethat communesin nationalparks (and possibly buffer zones) are not includedin the project,and that every commune moves through real processes of consultationand participation of local people,it will be importantfrom the startthat the names of thecommunes included in theCBRIP are clear andagreed at all administrativelevels.

4.1.1 How Road Access to Communes Influences Infrastructure

Communes in highland areas with road access have had an influx of funding for infrastructure projects in their more remote villages. This occurs primarily under the resettlementprogramme, and, where present, NGO programmes.Poor villages within this type of communeare not being excluded; indeed, in one

49 commune(which has had infrastructureworks built both through an NGO and through the fixed settlement programme),people in the central (and better-off) villages questioned,"Why do infrastructureprojects always go to the H'Mong and Kh Mu [in the villages at the communeperiphery], and not to us?"

However, in the most remote highland communes-those without road access-government infrastructure projects are very few. Infrastructurein such communesbuilt under the resettlementprogramme or 135 is more limitedthan in more accessibleupland communes.People in Tra Nam Commune,Tra My District (who had no projectsunder the fixed settlement programmeand only one 135 project introducedin late 1999), put this down to the lack of incentivefor district-level administratorsto carry out these projects in communessuch as theirs, which lies a two-day walk from the district centre ('district cadres get paid the same amount if they drive up to a closer communeand implementa project there as they do if they walk two days to get here. So why should they begin their work here?'). These projects thus reach the 'remote' communeswith road access, but have not yet reached those areas that are truly remote. 4.1.2 Remote Villages' Access to Shared Commune Infrastructure

The location of shared commune services in the commune centre makes access difficult for distant villages. Communeclinics, primary and secondary schoolsand communal meeting houses are generally built in the administrative centre of the commune. It is to the communecentre, too, that the main roads are built and that electricityconnections, where they exist, arrive first. This pattern follows logicallyfrom the fact that the most central and accessiblevillage in the commune routinelyserves as its administrativecentre. However,poor transportation from villages to communecentre-which was a feature of all of the highland communes and some of the lowlandcommunes visited-means that the more distant villages are poorly served by this system. Their distancefrom the commune centre usually indicatesthat they are the poorer villages in the commune;the distributionof infrastructurebenefits also meansthey are the least well-servedby the communeinfrastructure.

Some efforts have been made to place schoolhousesand classroomswithin reach of the most remote villages. In many villages, simple classroomshad been built for the earliest grades (anywhere from one to three grades) so that the youngest children do not have to undertake long journeys in the early years of their schooling. Children from farther villages still face difficultiesin getting to higher primary and lower secondary school grades in the communecentre.

50 Clinics, however, are often located far from remote villages. The central point in the communeis oftenjudged too Box 22: The Road to the Clinic far for peoplefrom remote villages A communetypically has one clinic situated in a central to reach,especially when the sick location.These clinics are oftendistant from outlying haveto be carried on foot (see villages.Tay and Nung migrants from Cao Bang began to settlein Tan HoaCommune in DongPhu District of Binh box opposite).Informants from PhuocProvince in 1988,and new families were continuing remotevillages said thatthey to arrivein 1999.The Tay andNung villages, Dong Xe, would liketo havea secondclinic DongTrac and Trang Tranh, are far fromthe centre(12 km,7 kmand 17 km respectively),communewhere the withintheir commune, closer to healthclinic is situated.The roadconnecting the villages to their villages. thecentre is passableby bicyclesand motorbikes in dry weather,but in wetweather the commune centre can only bereached on foot. Several pabents from the three villages havedied on these roads before they reached the clinic. 4.1.3 Access for Households Lastyear two peoplesuffering from acute malaria-a manin within the Village histhirties and a youngboy-both diedon the road.In each casethe family was obligedto carrythe patient on a stretcher(suspended in a slingbetween two bamboo poles Access to infrastructureworks betweenstretcher-bearers walking on foot) down to the is poor for some households clinic.In the same year, a woundedman was caried down to the clinic,but did notmake it. Peoplein thesevillages within the villages. Thisoccurs viewroad and bridge improvements as keyelements in oftenin villageswhere groups of receivingtimely treatment. (Dau Quoc Anh) houses are concentratedin clusters distant from one another, or houses are widely scattered. It is especiallydifficult to spread the benefits of water supply systems equitably in such situations.The case study in the box below illustrates-among other problems-the difficultiesthat may arise due to widely spread householdsor householdclusters.

Box 23: The Trouble with Water Supply in a Mountain Village The firstproblems arose in the meetingto preparefor construction.People couldn't agree: neither on where to placethe water tank, nor on mobilisinglocal labour. Many people felt that building the tanknext to the village leadershouse would be too costly-and anyway, they said, it isn'tthe village centre. Once the villageleader explainedthat this was a technicalrequirement, they reluctantlyagreed, but they sIll hadtheir doubts. Getting localpeople to workon the projectwas difficult:some people were not willing, as they thoughtthat their houseswere too far awayfor themto bother;they would benefit less, they said, so whyshould they have to workas muchas everyoneelse? Emotions ran highenough on this issue that somepeople-those living at the peripheryof the village-refusedto goto themeeting even to discussthe maKter.One of thesepeople, Mr. C.,went to hisfield instead. When he was informedthat hewould have to pay moneyif he hadn'tattended the meeting,he repliedthat hewas poor and didn't have money, and he wouldn't use the water. The secondproblem arose in implementabon.The village leader didn't have a clearidea of howmuch water was availableand howmany households would be ableto havepipes to theirhouses, but hewent ahead and gotthe whole community to contributetheir labour for construction.When the construction was completed, it tumedout there were no water pipes leading to the sixfarthest households. These families complained that they hadhad to workas muchas othersbut didn't receive the samebenefits; it hadnot been made clear to themat thestart that they would not benefit directly. The thirdproblem arose in maintainingthe tank and pipes.Each person had to pay2 kg of rice/yearfor maintenanceand replacingmaterials. However, the vice-leaderwas the only person charged with managing thiswork. He was responsible for protectingand repairing the tank;if anyonebroke anything, that person was obligedto payfor therepair. The vice-leader thus received2 kg of ricefrom each villager and extra from anyoneresponsible for specificdamage. He spent the moneyhimself without asking the secretary or the cashier.Many people said, 'But there has been no damageor repairs,so whydo we still haveto pay?'(Tran QuySuu)

51 4.1.4 Insufficient Investment and Foregone Benefits

Local people cannot benefit from infrastructure projects if the amount of investment is not enough to solve the problem at hand. In some cases, externally assisted small-scaleinfrastructure Box24: Water Supply Receives has failed to solve problemsproperly for the Attention-but NotEnough-In Phan community becausethe financing of the HoaCommune solution was not sufficientto solve a very In thisCham commune in BinhThuan solutionwasnot ufficien to solv a veryProvince, localpeople have relied in the expensiveproblem (see box opposite). paston diggingwells and using water from a streamto get waterfor drinking.The communeis notfar from the sea, however, Similarly, local people cannot benefit from andthe water is verysaline; it is notsuitable schoolhouses and clinic buildings unless for drinking,only washing. Households rely the corresponding services are appropriate on purchaseddrinking water, which they andohh atIoabuy about 2 kmaway for 200-500 VND per and of high quality. Informantsstated that thung(one thung is 20litres). This is services in clinics could, in their view, be sufficientfor a familyof 5-7people for24 improved,and their remarks suggested that the hours. However,there are many of healthcare remainslargely a householdswhich simply cannot afford to currentview of healthcare remainslargely a payfor water.What do theydo? They just curative one. When asked if a health clinic was drink salty water. amongtheir priorities for improvingthe Assistanceto improvethe watersupply has infasrur ibeenextended to thecommune through infrastructure in their communes, a typical bothdonor (UNICEF-) and NGO projects. responsewas, "Yes,we shouldbuild/upgrade Thishas consisted, though, of limitedefforts the clinic, but only if we can make sure there with limited funds-deeper wells in new are enough medicinesthere." Obstaclesto locations.The newsources areenough ~~~~~~~~~~~salinetoo.Local cadres reckon of supplythe only are way providing appropriate basic educationfor ethnic to supply fresh water to the communeis to minority children in Vietnam have been widely pipe it in-from a considerabledistance documented. away.This is anexpensive venture that they havenot as yet beenable to have funded,and clean drinking water remains a 4.1.5 Ways in Which Poorer Households key problemin the community. May Benefit Less (NguyenVan Linh/NatashaPairaudeau)

The poor can only benefit from schools and classrooms as long as they can afford to send their children to school. In the lowland and coastal communes visited, where parents must pay the costs of schooling after primary levels, it was observed that children in poor families rarely continue schooling after grade four or five, becausethe costs of tuition are prohibitive.Although minority children are exempt from tuition fees, their parents still have difficulty meetingthe costs of schoolbooks, clothing and contributionsfor school maintenance.

The costs of connection of villages and households to the national electricity grid is out of reach of many villagers. Though many commune centres within the sample were connected to the national electricitygrid, very few outlying villages have been able to afford to pay for the connectionfrom the communecentre to their villages. In some communes,wealthier householdscan afford the sum requiredof them, but their poorer neighboursclearly cannot. As the connection is paid for out of the accumulated contributionsof the whole village, projects are stalled as a consequence. Several informantsremarked upon the disparities between rural and urban areas with regard to connecting to the national grid: "City people are much wealthier than us, and they can afford the relatively low costs they pay for electricity connections.We are very poor in comparison,and we cannot be expected to afford the same costs."

52 Box 25: Who can Afford to have Electricity Installed? In Village7 in Thach Tuong Commune(Thach Thanh, Thanh Hoa), one householdhad spent nearly 5 millionVND to install a privatehydropower network. Meanwhile, villagers were beingasked for a contribution of 200,000VND per householdto connectto the grid, but some householdsin the villagecould not afford this. This rate had already beensubsidised for peoplein village 7, due to theremoteness of their village (it is the farthestvillage from the communecentre).

Althoughthe benefits of improvedroad access are widespreadwithin a commune, better-off households tend to benefit more than the poor, simply because they have the resources to do so. The better-off can afford to purchase new seed varieties, fertilisers and pesticides which allow them to make better use of irrigation; they are more likely to own motorbikes which, when roads improve, give them better access to the outside; and they are more likely to have commercial interests which will be improved significantly when central roads are upgraded.

4.1.6 Migrants and Infrastructure Benefits

In areas of in-migration, migrants are quick to make use of any infrastructure improvements. Migrants are especially drawn to areas once improvements to infrastructure related to their commercial interests are made (market centres, roads, irrigation). In Bu Dang District of Binh Phuoc Province, it was observed that migrants had bought land and installed themselves where roads have been built or improved; a market centre that was completed two years ago has become a hive of migrant commercial activity.

4.1.7 Infrastructure of Specific Benefit to Vulnerable Groups

Safe river crossings are particularly beneficial to school children. The danger that unsafe crossings pose particularly to children on their way to school has been discussed in Part Four.

Kindergartens are especially beneficial to ethnic minority children. Several studies within Vietnam have shown the effectiveness of kindergartens in improving the quality of learning for ethnic minority children in subsequent grades. This view was reiterated in the field study: numerous ethnic minority informants either said they wanted a kindergarten to help their children get a head start in schooling, or that they appreciated the recent construction of a kindergarten in their community.

Kindergartens (mau giao) and day care (nha tre) are of great benefit not only to the children who attend them, but to their parents and older siblings. This is true in many areas where people continue to depend on upland cultivation for their livelihoods, but it also applies to areas where opportunities for day labour are growing. In both cases, adults may spend entire days a long distance from the home. It is common practice in highland areas to solve the problem of childcare by getting older children to look after their younger siblings. The older children often forego school in order to do this. Local kindergartens and day care centres allow parents to work more freely, and children (both the younger and older siblings whose parents can afford the costs) to attend school. Such services are particularly beneficial in single adult households where it is unlikely another adult child carer is present.

5.3 Water supply projects reduce one of women's work burdens.Across the communesvisited, collectingwater is primarily a woman's task, and women may spend several hours a day collectingwater when there is no local supply. It is one of the many tasks that contributeto women's heavy workloads. In those communities where water supply has already been improved,women and other membersof the communityrepeatedly stated that the improvementshad made the greatest change to women's lives.

Accessible clinics make a big differenceto the elderly. Elderly people interviewedin the communessampled routinelyranked clinics as a high priority.

Toilets in public buildingsare particularlybeneficial to women. Female teachers pointed out that schoolhousesand other public buildings do not have public toilets in many communes.They said this creates unhealthyareas around their schools and is a great inconveniencefor themselves.

Public village meeting houses may help to include more marginalised householdsin processesof consultation.Except for those minoritygroups which have traditionalmeeting houses within their villages, at present most village meetings are held in the village leader's house, or the 'nicest' house in the village (which often belongs to the village leader). Some respondentssaid they would like to have public meeting houses at village level (rather than meeting in a private house), becausethis could prevent public benefits (like a village television)from becoming private by being located in one household.Others said meetings held in a public place might encourage broader participation(at least of men) in public meetings.

4.1.8 Benefitsand Potential Lossesfrom Road Improvements

Better roads can lead to a multitudeof improvementsfor communes. Improvementsboth in commune-to-districtroads and in inter-villageroads can bring a number of benefits.The sick can reach treatment more easily, children's access to schools is improved,and, when paved, roads double as places for drying rice, cassava and other produce. Photo: Ricedrying on the road throughThach Thanh District,Thanh Hoa Province(photo: Dau Quoc Anh)

54 Most significantof the changes roadsbring, though, are increases To reach Dak Nhau Communein Bu Dang district involvesa drive in the prices of agricultural of upto three hourson a very rough and rutted dirt road. When it produce, and lower input costs, rains it is very difficultfor trucks or other four-wheeledvehicles to due to better transportation (see reachthe commune.When the steel-and-woodbridges are out (whichhappens when heavy trucks-logging trucks, some said- box opposite). travel over them) it is virtually impossiblefor four-wheeled vehiclesto make it to Dak Nhau which is situatedat the edge of However, better roads can benefit the district.Farmers in the commune havebeen plantingcash loggersthe expenseat of the rest crops for the last 10years-first cashew (introducedto the area loggers at the expenseof the rest ten years ago) to which they have now addedcoffee (introduced of the community.It was clear in in the last 5 years).As the productionof cashcrops has twofild loationsthatllegalloggg increased,farmers have grown less and less rice, to the extent two field locations that illegal logging that manyhouseholds now rely entirely on purchasingnce. was taking place. It was not overly However,poor roadaccess meansit is very expensiveto buy rce disguised in either commune,despite in the commune-it costs 34,000 VND per kg. Whenthe bridges the fact that in at least one of the are out, the trucks cannotget through at all and the price of rice canshoot up to 5,500VND/kg-which is whereit stoodwhen the communesthe local forest protection researchteam visited. officers (Kiem Lam) were heavily armed. In both cases, the activities of loggerswere actually damagingthe roads, creating discontentamong locals not involved in such activities. In Xa Luong Commune,Tuong Duong District, loggers had turned the recently surfaced road into muddy pits, by continuallydragging logs down it. Local people resentedthis: they had contributedlabour to maintainthe road, but felt they were doing this not for their own benefit, but only to have the loggers destroy the road. In Bu Dang District in Binh Phuoc Province,the road throughthe district had two major steel and wood bridges which had collapsed and are no longer passable by four-wheeledvehicles. Although some informantsdenied it, others said this was due to heavy logging trucks passing over the bridges:while the truck companieswere called upon to pay for the repairs, this processwas very slow and in the meantime, local people had to suffer the consequences.

Box27: How to MakeOld Wood On each of the two dayswe drove up to the most remotevillages of Xa Luong Commune(Tuong Duong District, NgheAn Province),we saw, on average, 10-15 buffaloesbeing led down the road, pulling heavy timber logs behindthem. The dirt road has been levelled in-the last year, but was alreadydeeply pitted in some places. Curiously,the men leadingthese buffaloeswould lead the animalsthrough the pit, allowingthe logs to linger some time in the mud;they then sloshedmuddy water over the ends of these logs. This, we were told, is 'instant ageing':through this process,it is hoped, logs will look old enoughthat, if the logger is stoppedon the way down the hill, he can convincethe authoritiesthat he is haulingacceptable, old wood and not timber that has been recentlyfelled.

4.1.9 Local Priorities

In order to gain a better understandingof local priorities for infrastructure,the researchteam asked district and communeauthorities and people in the villages to rank their prioritiesfor subprojectsprojects proposedunder the CBRIP. This was done using a set of ranking cards much simplified from the list of eligible works. They were asked to rank projects based on the prioritiesthey perceivedfor the community (for authorities)or for themselves (for individuals).After ranking, peoplewere then asked to explain the reasons behind the choicesthey had made. It was intendedthrough this exercise to find out if different administrative levels and different types of people (men, women, different ethnic groups, the

I55 elderly, the poorest households)had different prioritiesfor infrastructure,based on their position,their gender, their age or their socio-economicstatus.

Distinct priorities based on gender, age, or even economic status did not clearly emerge, however, from the rankings. It could be argued-and the research teams spent several long eveningsdiscussing this and making adjustments-that the rankings needed to distinguishmore clearly between individualand household interests. However,when this was tried, it didn't make much difference: individualrespondents continued to rank prioritiesbased on the interestsof their whole households,rather than (what an outsider might view as) their individualinterests.

Moreover, regardless of who was being interviewed, the question was answered based on the infrastructure works people did not yet have in their villages or communes. Gender, age and economic differencesemerged only in the reasonspeople gave for ranking their priorities as they did.

Although the ranking exercise did not produce the type of information that was hoped for, two important lessons emerged: * Although when pressed, local people could speak of problemsspecific, for example, to women or to the elderly, when referringto their own householdsmost rural people find it difficult to separatethe interestsof individualswithin their householdsfrom the interests of the household as a whole; * People's priorities are practical needs based on local conditions.

4.2 Local Involvement in Small-Scale Infrastructure

Part Three, section two, describedhow different players were involved in different types of infrastructureworks, and identifiedthe organisationalstrengths that each of these works relied upon. This section assesses in more detail the involvementof beneficiariesin all stages of the project cycle.

The project cycle includes a number of processes.These include: information dissemination,planning and design, implementation,management, maintenance, handling conflicts, and financial oversight. Consultationshould run throughoutthe project cycle. Barriersto, and opportunitiesfor, local involvementare identifiedfor each step. 4.2.1 Consultation

Consultationof local people is a critical componentof the CBRIP design. The most basic tenet of the project is that local people be allowed to decide for themselves what types of infrastructure they require. Villages,and then communes,would need to reach agreementon project proposals. As the arrangementimplies competition between villages within one commune for the same resources, this is particularly important. Local consultationwould continue to be important through all stages of the project cycle. Public 'consultations'do not alwaysserve to includepeople in decision- Box28: Being Inforned ... or Deciding? making. Communeand village SomeWomen's Union cadres attending commune-level meetingsto transmit informationfrom meetingssaid they were dissatsfiedthat, althoughthey higher levels down should in principle are invitedto localmeetings, they are presented with decisionsthat have alreadybeen madeand are not be different from local level askedfor their opinions.Afltough this difficultywas consultations.Local informantssaid, presentedas a genderissue, villagers attending village though, that when they are 'consulted', meetings(both men and women) have also said that when they attend local meetings,they are informedof this is basicallyno different from decisions,but not asked to decidethemselves. This meetings in which informationis suggeststhe problemis not only one of gender,but is transmittedto them with no opportunity morewidespread. to respond.

True local consultationis strongestat village level. Consultationof local people is stronger and more effectiveat village level, particularlyfor those works carried out with no external assistance.In minorityvillages, these processesare facilitated by the fact that they take place in local languages.

The ability of remotevillages to have their concernsrepresented at the communecentre is weak. This is due to distance,which in turn leadsto the concentrationin uplandcommunes of cadres' positionsin the hands of people (and ethnic groups)from the communecentre.

The advice and views of traditionalleaders (gia lang) are importantin guiding local people'sdecisions. Many gia lang play importantroles as 'key consultants'in village-leveldecisions. However,degrees of local comfortwith making this person a visible public figure vary: some gia lang are kept quite separatefrom the administration,and their communitiesclaim that, while their opinionsare important, they do not appearat administrativemeetings but are consultedmuch more informally.This, for some groups,seems to be a comfortableway to maintaina strong customarycommunity figure without his stepping into the area reservedfor the official administration.Other gia lang, it was observed,play leading roles in meetings-but this is primarilywhen they hold positions in mass organisationsin additionto their traditionalpositions. Either way, their influenceremains strong and their wisdom and experiencerespected: in one village, for example,a consensus could not be reached if the gia fang was away from the village.

Women are frequentlyexcluded from village decision-making.Women do not attend village meetings in many Box 29: Who Should Come to the Meeting? communes.Often, only householdheads The recent Decreeon GrassrootsDemocracy, issued are invited and household heads are in May 1998,stipulates that 'people", 'voters" or "householdheads' shouldmeet to discussvillage and seldom women. This practice was found communeconcems. Giventhe choicebetween three to be prevalentin many communes options,local cadresfall back on whatthey have visited. customarilydone, which is the third option, to call householdheads to meetings.Taken that household headsare rarely women,an opportunityto encourage However, women with relatively high women's participationin local-levelmeetings through education are visibly more involved in policy directiveshas been missed. communitydecision-making. Better educationfor women makesa real difference in enablingthem to participate in local decision-making,and in bringingother women in their communitiesinto the circle of decision-making.As the case below illustrates, even if they are not activelyadvocating the inclusionof other women in village or

57 communeactivities, they have a positiveeffect on other women within their communities. Box 30: EducationWidens the Circle Althoughseveral women attended the meetingof villageleaders, the women in this Kh'Muvillage were rather reticentto takethe leadin discussions.One exceptfon was Mrs. M. Sheis in her 30s,-the daughter of a former villageleader, and, unusually for womenin thiscommunity, her father supported her to continueher studiesup to universitylevel. She spent a yearin universityin Hanoi,a rarething in thiscommunity even among men. Unfortunately,after the firstyear, her family could no longerafford to supporther and she was obliged to returnto thevillage. It cannotbe said,though, that her educationwas wasted: i hasopened opportunities to her withinthe village and commune, and has made her a greatbenefit to her community.She is a Women's Unionleader and a cashierwithin the village, and she plays an activerole in villagemeetings. People- regardlessof theirsex-defer to her becauseshe knows a fair bit morethan most people about affairs outside the commune.A moresubtle influence she exerts was witnessed when the research team and village cadres sat downto a jointmeal. In other villages, women routinely disappeared at this point:she stayed,and because of her presence,two other women came back, pulled up stoolsand joined in thediscussion.

Households which spend long periods away from the village may also be excluded from decision-making. Householdsfor whom seasonal migration,wage labour or work on upland fields are importantlivelihood strategies,and single adult householdsin which one adult carries a particularly heavy work burden,are often unable to attend importantmeetings. It is not thought appropriate in many places for wives whose husbandsare away for long periods to attend meetings in their place- or indeedthey may not feel comfortabledoing so; peoplewho wage labourfar away or work on distant uplandfields may spend periods of several days to weeks outside the village at certain times of year, and thus may not be present at village meetings. Young widowed women or single mothers28 may be the recognisedheads of their households,but they are also among the busiest people in the village and the least likely to be able to attend meetings. 4.2.2 InformationDissemination

The CBRIP would require the disseminationof informationto explainthe ground rules of the project to local people. Although the results of consultationssuch as this Social Assessmentare intended to bring local-level views into the process of project design, ultimately these rules will have to be transmitted,in some formn,down from higher to lower levels.

Knowledgeof governmentpolicy and programmesremains very poor. This is true regardlessof ethnicity,though levels of knowledgeare particularlypoor among ethnic groups which live in the most remote areas, or the most remote villages within a commune(see box below). In the latter cases this is due both to sheer distance and to languagebarriers. Box 31: Gender and Knowledgeof GovemmentPolicy and Programmes Womenin remoteareas, and fromsome ethnic groups, are disadvantagedboth by lowerlevels of Vietnamese languagecomprehension than men, and pooreraccess to alreadypoor information flows: they are muchless likelythan men to attendvillage meetings.

2z This appliesequally to widowedmen with young childrenbut the incidenceof such householdsis muchlower (it is much more commonfor youngwidowers to remarrythan youngwidows). Nonetheless,people are hungryfor information.Everywhere the field teams visited, local people said they wanted Box32: We Want to Know... greateraccess to informationabout 'Wewant to knowwhat is happeningoutside practicalthings that concernthem, and thecommune. We want to knowif farmersin greater knowledgeof how otherpeople otherparts of thecountry have the same across the countrysolve problemssimilar to difficultiesasus, and what they do to solve thosethey face themselves. (womanin DaSar Commune, Lam Dong, Binh Phuoc) 'We do notget enough information here. When Basicinformation about government the floodshappened [in central Vietnam in policiesand programmesmust travel November1999], we did notknow here for a throughevery administrative throughevery levelto ~~~~~~~~~~centre."week,unfil someone went to the commune reachlocal people: once it getsto the (womanin Xa LuongCommune, Tuong Duong village,it mayhave been reinterpreted District,Nghe An) several times. Informationfrom higher levels is routinelytransmitted though communemeetings to village leaders, and then from village leaders to heads of household,who are then charged with carryingthe informationhome to household members.This informationflow was the rule in all the communes studied,and there were few examplesof villagers (besidesvillage leaders) attendingmeetings at communelevel.

Among the limitationsof this system, leaders from remotevillages may not be able to attend the meetings.Within many communes,villages are connectedto the communecentre by dirt paths or roads which become impassablein wet weather. If an importantmeeting occurs on a miserableday, leaders from these villages may simply miss out on the meeting, or have to 'catch up' later. This greatly reducestheir understanding(and, consequently,the understandingof their constituents)of the matter discussed.In turn, commune-levelauthorities have difficulty reaching these villages. They too are faced by poor roads, but they also have limited budgetsfor travel within the communes-one cadre said he paid out of his own pocket for transport to villageswithin the commune.This means communecadres may visit remote villages very infrequentlyto informthem of meetings.

In addition,village leaders may have difficulty understandingand retainingthe informationtransmitted to them at communemeetings. This problem is often described with reference to ethnic minoritiesas a problem of 'low levels of education' (trinh do thap). The problem is not one of low aptitudes,though, but of village leaders having difficulty retaining informationtransmitted in Vietnamesewhen their language abilities are poor. Furthermore,it is not just ethnic minorities who are unableto retain informationtransmitted at meetings. Informantswho were native speakers of Vietnamese also claimed to have difficulty in retaining information,as there is little opportunityto ask questions in these meetings,and thereby to check that the messageshave been properly understood.

Furthermore,household heads must routinely rely on one personto pass on informationto them that will have an importantimpact on their families.A village leader who passes on the informationclearly and diligently is a great asset to his community and highly valued; conversely,a leader who is less than diligent in passing the knowledgehe gains on to his constituents(or is unable, for a number of reasons, to do it well) can severely limit the participationof householdsin his village in state programmes.

.rq Moreover, household heads may not transmit this information clearly down to other household members. The most marginalisedmembers within households are likely to receive the least informationabout governmentprogrammes and policies. They may learn of government programmesand policies, if they hear about them at all, at several removes from the 'source', and thus they tend to be the least informed people in the community.

Aside from the information transmitted through formal meetings, other channels exist through which information is received from beyond the commune. Newspapersand magazines are routinely receivedin communecentres Box33: Video Nightin Na Be and,to a lesserextent, by village Na Beis a H'Mongvillage situated at onedistant end of leaders.These materials, though, do Xa LuongCommune (Tuong Duong District, Nghe An), not circulatewidely within communes far fromthe electricalgrid that serves the central part of thecommune. Nonetheless, villagers-who moved to or villages,a reflectionin ethnic thisplace four yearsago from a higherarea up the minorityvillages of low levelsof literacy valley-are eagerfor informationand will go to great in Vietnamese.Increasingly, radios endsto obtainit. They received a televisionfrom the govemmentresettlement programme which they have andtelevisions are appearingand managedto tumon usingan old carbattery. The village werefound in the field studyeven in is a long,long way from a televisionsignal, but villagers the mostremote of villages.An are unfazedby this:videos are routinelyrented and broughtup to the village,with anevening's increasein radioshas beenmade entertainment-consistingprimarily, it appears,of action possiblethrough a government films-enjoyedby largegroups of peopleboth from Na programmethat enablespeople in Beand the neighbouring village, Imountain areas to buy them at subsided prices. Some televisions have also been given to villages through the resettlementprogramme, but others have been acquired by the better-off householdsin the villages, and connectedthrough sheer determination(see box above).

Despite the existence of electronic media in many communes, however, very limited development-related information appears to be picked up through these media in rural areas. Locals appear to be using radios and televisions primarily for entertainment.

4.2.3 Planning and Design

The CBRIP aims to address the needs and priorities of ethnic minorities and other vulnerableand marginalisedgroups. This would include incorporatingthe local knowledgeand cultural priorities of ethnic minoritiesand the needs and priorities of marginalisedgroups into project planning and design. The preparation documents also suggest the possibility of using traditional and regional designs for public buildings such as schoolsand health centres.

Issues of cultural importance to ethnic minorities are too often overlooked during project preparation. Spiritual or customary beliefs may not seem important, or indeed may not even be evident, to outsiders, but they are likely to be very importantto local people.As one (ethnic minority)district chairmanstated, this is not merely a sentimentalissue being raised, but a practical necessity if projects are to be successful: "If planners listen to minority people and respect their customs, infrastructureprojects are much more cost-effective.In the past, many projects have

An been built without asking the views of local people, and then the projects have never been used."

Recognisingand paying attention to these cultural differences may be an especiallydifficult task for contractorsand technicianswho are accustomedto working only within the bounds of their particularareas of technicalexpertise. As the same minority chairman(above) remarked;"Technicians need to respectthe customs of local people when Box 34: OneWoman's Castle... they build infrastructure. For Two yearsago in thisvillage in BinhPhuoc Province, the local authoritesbuilt a newwooden house wfth a sturdytile roofand a tile manyethnic groups It is floorfor an elderlyM'Nong women of 78 years.The researchteam importantto consultpeople andthe headof theWomen's Union (who guided us throughthe beforehandabout the best house)all agreedit lookedlike a ratherpleasant place to live. The personwhose opinion really counted, though, thought otherwise. She g hasnever moved into the newhouse, prefemng to stayinstead in a they tried in this communeto smalltatch housewith a traditionalpit fire anda bedalong the wall buildthe clinicon a uponwhich her rowof valuablejars is linedup. This humble house hadto be rebuilt,in fact,as it hadbeen taken down when the new graveyard,local people housewas built. The authorities were well-meaning: they had meant protested-this is not to helpher, as an elderlywoman living on herown, and to honour superstition(me tin); it is an her, as the motherof warhero (liet sy). Instead, though, they took out of hercontrol something that mattered to hervery much. As shesaid, belefimportant that mustbe 'if thegovemment wants to buildme a house,that's fine with me, but respected. theycan at leastask me what kind of houseI want." By the same token, design features may be culturallyspecific. As modern as a new house or building may appear to a contractor or an outsider, if it does not fit in with local requirements,customs and ideas of comfort, it may not be used and may end up a well-meaning,but ultimatelywasted effort (see box above).

Despite some strong local design and constructionskills, traditionaland regional designsare not used in governmentinfrastructure works. Local designs are tolerated for schools and classrooms,but were never seen in any of the communes used in clinics or other works built with governmentassistance.

However, some projectapproaches have been able to incorporatethe needs of specific groups into project design. Consultationon project design with those interest groups most directly affected has had very positive results. This specific type of consultationwas not found to be widespread,but one single example may illustrate how effective such consultationscan be in producingdesigns that work for the people who use them most. In preparationfor an NGO-supportedwater supply project in Tuong Duong District, NgheAn Province,it was decided that, as women in Tuong Duong, like women across the country, are the main collectors and users of water, they might have specific ideas and requirementsfor the planned project. This assumption proved to be right. They emerged from a preparation meetingwith precise plans for where the taps and showering rooms should be placed.

4.2.4 Implementation

Important elements of subproject implementationdrafted into the CBRIP design include an emphasis on the use of local labour and local materials wherever possible. Once subproject choice has been finalised at communelevel, it is anticipated that bidding will go out to contractors. Local people are routinely underpaid for their labour. When local people are asked to contributelabour or cash for infrastructureprojects, the rate calculatedfor a day's labour is well underthe market rate. This unduly penalises the poor. Poor people rarely have cash availableto pay the contribution, so they usually pay in labour. It is only better-offhouseholds which can afford to pay in cash.

Conscripted labour is sometimes confused with programmes which have a budget to pay for local labour. In communesvisited, local people routinely contributetheir labour (10 days per year is the standard) to road maintenanceand sometimesto other locally-agreedprojects (school maintenance,etc.). For some specific donor and governmentprojects, however, budgets have been set aside for local labour, but authorities had required peopleto work their 'voluntary' days before reaching into the budget set aside for this purpose.

Currently, communes must experience delays if they want to make use of local Box35: Red Tape and Trees materials (especially wood). It takes Onedistrict-level leader expressed his frustration with a longtime to get permissionto use theprocesses that hold up projectimplementation in his localmaterials (such as localwood) for distrct.He is anxiousto get roadimprovements completedbefore the rainsstart through one section of constructionof infrastructure.Other thedistrict. However, the roadrequires two bridges for buildingmaterials have to be brought whichwood is essentialand readily available in the area. in on roadsthat are oftenvery poor. However,in orderto beable to cut thesetrees, he must go the provinceto requesta letterof permission.He These delays sometimes push the mustthen take this letter to the ForestryProtection scheduleinto the rainyseason, Office,and then backto the province,before he cancut delayingwork until conditionsare dry downthe trees. I enough again.

Delays in disbursement of funds and cumbersome appraisal and approval procedures can hold up project construction. Funds often do not become available until after the dry season (the best time for construction).Poorly-timed disbursementresults in projectscarried out in a hurry. This reducesopportunities for local consultationand involvement,as well as reducingthe quality of the infrastructureworks.

While communes in remote upland areas can be expected to have little experience of hiring contractors for infrastructure construction, in more accessible uplands and many lowland areas, such experience may already exist. Contractingfor infrastructureworks financed locally has already been mentionedin Part 3. Some communesmay also have experienceof bidding proceduresthrough programmeswith external aid (see box below). Box 36: Training and Infrastructure Bidding in Loc Binh Commune, Thua Thien Hue In this coastal Kinh commune,40 km from Hue city, communecadres have receiveda considerableamount of training relatedto managementand participatoryplanning. Several commune-level cadres attended an 18- month coursein law, policy and economicmanagement skills in Hue,and a two and a half month coursefor communemanagement held by the province.Commune Women's Union and FatherlandFront cadreshave also participatedin trainingworkshops in the district.Three Women'sUnion cadres, who stoodfor seats in the CommunePeople's Council last year, participatedin a 4-daytraining course in women'scapacity and leadershipat the grassrootslevel. This was held by the districtWU (throughthe NCFAW. All cadresin this communeare educatedto at leastgrade 5.

In 1999,in the courseof an NGO-fundedproject in all 6 villagesof Loc BinhCommune, the communeproject managementteam (PMT) had a chanceto makea biddingselection for constructionof a kindergartenvalued at 75 millionVND in Tan An village.This PMTwas headedby the CommunePeople's Committee secretariat and vice-headedby the headof the communeVVU and representativesfrom all villages.The PMTprepared the invitationto bid and, after carefulconsideration, they chose a constructioncompany of Phu Loc District. The PMTalso invited technicalexperts from districtlevel to be consultantsfor technicaldesign monitoring, planningand appraisaltogether with the NGO'sproject officers. Meetings were held to gamer ideas from local people, whoalso actively monitoredconstruction. Finance was reportedto the NGOand publicisedin the commune.

The head of Loc BinhWU remarkedthat the financialregulatons of govemmentare muchmore complicated than those of the NGO. She added that, with the experienceacquired, the communeshould be able to manageimplementation of the CBRIP. (Dau Quoc Anh)

Contractors cannot always be relied upon to give a fair deal to local labourers. Two problemswere cited in this regard.Complaints were heardfrom some communes that contractors hired outside Box37: PayCut workers-claiming the work was 'too In onecommune, an NGOsmall-scale technical'-but local people professedthey infrastructureproject had been implemented. could have done parts of this work Accordingto the distnct,the policedepartment of the provincein questionwon the contractfor themselves,but weren't hired. In a more the work from the NGO.The NGO had insisted extreme case, contractors profited from a uponlocal labour being used.A rate of 10,000 labour budget allocatedby paying local VNDper personper day had been set, and the moneyfor wage labourwas includedin the labourersless than the amount available, budgethanded to thecontractor. The and not disclosing sums to local people. contractor,however, paid local labourersonly 5,000VND per day. It appearslocal peopledid not know until laterthat they had been duped; althoughit was not madeclear whether any 4.2.5 Management official complaintwas made,no actionwas taken to correctthe error. CBRIP documentsanticipate that skills and capacitiesof communeleaders to manage a process of participatoryplanning and implementationwill be limited, and that mass organisationswould be involved in the process of planningand managementof infrastructureworks, with independentbodies overseeingthe projects.

The existence of commune-level management boards is very uneven. Decree 24 stipulatesthat boards for management,finance and supervisionmust be formed within the communeonce a project is approved.In practice,their occurrenceis irregular.In remote communesthis may be explained by the fact that districts frequentlytake over managementwhen, they judge the communeslack the experienceto do so.29 The remarkof one district-levelcadre is typical, "We would

29 It shouldbe notedthat thispractice is withinthe provisions of Decree24 (Regulationof voluntarycontributions). See article 9, para.3: 'Wheresuch a commune[faced with particular difficulties] fails to meetconditions for the establishment of the project managementboard, a districtlevel project management board shall be set up to organisethe management of eachcommune's projects". like the communesto take a more active role in infrastructureprojects and we have tried to encouragethem to do so. But they don't have enough experienceand we end up doing most of the work ourselves."

Many districts expressed a genuine desire to hand more project responsibility over to communes. It cannot be too hastily assumed that, just becausemany districts retain managementresponsibilities, this reflects a desire on their part to continueto control the process.While such motivationsclearly do exist in some quarters,genuine expressionswere also heard from district authorities of the desire to hand over more work to the communes.

Non-officials rarely play a recognised role in bodies set up to manage and supervise projects. As seen in Part 3, membersof commune and village managementboards for governmentprojects are almost exclusivelydrawn from the ranks of cadres. These managementboards are sometimes no more than an extension of the Commune People's Committee.Nonetheless, on several occasions local people expressedthe desire to include people who are not part of the administrationin the supervision of governmentprojects; and they recognise, through the practice of putting these people on the boards of works they built themselves,that it is effective. 4.2.6 Maintenance

CBRIP documentsrecognise that much local infrastructurefails to be adequately maintained,and that more in-depth knowledgeis needed about local capacitiesfor ongoing operationand maintenance.

Maintenance is particularly poor for infrastructure built under the resettlement programme. Local infrastructureprojects built under the resettlementprogramme form part of a packageof incentives to encourage peopleto either move location and/or change their production patterns. Such projects are very much viewed locally as a gift from government,with little incentivefor people to maintainthe projects,and little sense of ownership.

Although nominal boards exist in some communes to inspect projects at handover, communes feel they do not have the expertise to judge the quality of construction of projects handed over to them. The example in the box below illustratesthe types of difficultiesfaced by commune-levelboards at project handover. Box 38: "If it looks nice then it's fine with us": quality Very few government inspectionin a rural commune infrastructure projects have Constructionin villages-be it an electricity system,a school or formal local-level committees in maintaineda clinic-needsthereafter. to be inspectedIn Tan Hoa before Commune it is handedin Dong over, Phuand charge of maintenance. Structured District, Binh Phuoc Province, there are always 2 people from maintenancecommittees which the village and one from commune level in the inspectionteam gather contributionsfrom local (bannghiem thu): none of them, though, has expert knowledge in designor construction.Many types of construction,such as peoplewere more prevalent among electricity networks, puzzle villagers as they have had no NGO-assistedprojects. experienceof them before. Even commune cadres, who have workedin industryor constructionfor manyyears, feel they do not haveenough expertise to do a properquality inspection at Local people do what they can to handover,so they know what they are getting: .... We only pay make repairs, but for any repair attentionto the appearance;if it looks nice then it is fine; we do they cannot fix themselves, they not know about durabilityor absorbability.We only know what they the constructionunit side says to us; we have never been must appeal to district level or trained in construction inspection... says the head of the hire technicians. Appealing to the Industry-Transportation-ConstructionCommittee (Tuong Ban district causes delays as money AtCong present, Nghiep-Giaoinspection Thong-Xayconsists of Dung) simply ofhanding Tan Hoaover Commune. written must be found from district level. documents:the quality of the constructionis not checkedby the Hiring techniciansprivately may be communepeople as theydo not have the expertse to do so. costly, but also, local people may not (TranQuy Suu) be assured of quality (see box below).

Box 39: Troubles with Maintenance In Tan HoaCommune, people speak of twomain problems with maintenance. The first is thatlocal people can onlymaintain simple works such as levellingthe roadsurface or whitewashinglocal buildings. Other, more technical,works such as moistureprevention, or repairingwires, have to bedone by hiredpersons. These technicianscost dearly. Moreover, local people do nothave the ability to judge the technical quality of theirwork: they havefound in thepast they have contracted shoddy work which in tum hasneeded frequent repairs. Secondly,they say they are too poor-as recentmigrants to thiscommuPnity who are just establishing themselves-tocontribute money to maintainingworks. They can only contribute their labour. (Tran Quy Suu)

4.2.7 Conflicts and Grievances

The CBRIP hopes to minimise the occurrenceof conflicts, irregularitiesor grievances,by emphasisinglocal participationand disbursing funds at the lowest administrativelevels. Nonetheless,people cannot be expected always to have the same interests, and conflicts or disagreementswill no doubt arise in the project process. In anticipationof and preparation for these occurrences,workable mechanismsfor conflict resolution are required.

Local people hesitate to air their grievances publicly. Many people prefer to avoid conflict and confrontationand, even if they have a legitimatecomplaint, may not air it publicly for fear of losing face. Even when a village complaintdoes get a public airing, the decision agreed upon may not be carried through (see box below). Box 40: Financial Transparency in a Mountain Village In onemountain village, villagers were unhappy with the way the villageleader was handling the village accounts.A watersupply project had been completed not long ago with the assistanceof an NGO,and a committeehad been set up to handlethe maintenance of the system:each household is requiredto pay 10,000VND per year for maintenance.The village leader had taken these contributions from people, but ratherthan handing the money on to thevillage cashier, which he was supposedto do, hehad kept the money for himself.The cashier had raised the problemwith the CommunePeople's Committee, and there was an investigation.The investigation resulted in a decisionstating that the village leader and the Partysecretary of the villagemust compensate 1,700,000 VND to thevillage. But there had been no change,and the compensationwas not forthcoming. The village leader has still notgiven the cashier any village money to handle-8 millionVND from another source was provided in Septemberbut the cashier has not seen any of thissum. Villagers spoke about the casein whispers,checking their doors to see if the villageleader was anywherenear. 'He goesand drinks the money",said one woman, 'He's drunkall thetime." This woman and her husbandhoped to havea villageelection soon, as thatwas the only way they saw of gettingrid of this leader.Her husband (though himself related to thevillage leader) said this might be difficult,election results, he said,depend on the numberof relativesthat a candidatehas; the morerelatives he has,the greaterhis chanceof winning. (HaThi PhuongTien/Natasha Pairaudeau)

Legal procedures through which villagers can voice complaints or make denunciations to the commune level and higher have some limitations. Proceduresfor citizens to express their grievances in Vietnam are defined by the Law on Complaintsand Denunciations(No. 09/1998/QH10of December2, 1998). Citizens' concerns are supposedto be receivedby the Commune People's Committee and, accordingto the law, Commune People's Committeesshould set out one day a week as 'citizen receivingdays' for this purpose. Chairmenof Commune People's Committeesare given the authority to settle complaints-about themselves as well as about responsiblepersons undertheir management(Chapter 11, Section 2, Article 19). Thus, a grievance against a People's CommitteeChairman would have to be directedto the chairman himself; taken the strong cultural preferencefor avoiding direct confrontations,launching a complaint in this way would be extremelydifficult.

The law does not include clear legal procedures for making complaints to the village level administration. Villages do not legally constitute a level of administration,and thereforeprocedures described under the Law on Complaints and Grievancesdo not apply to this level. However,the law does state that: "The State encouragesthe conciliationof disputes among populationsbefore they are settled by competentagencies, organisationsand/or individualsin order to restrain complaintsfrom arising from the grassroots"(Chapter I, Article 3). In addition,the Decree on GrassrootsDemocracy (Decree No.29/1998/ND-CP)states that village or hamlet meetings shall be organised "with the participationof all voters or household heads"to "elect or dismissthe village of hamletchief(s); [and] elaboratevillage or hamlet conventionsand rules" (Chapter 13, Article 14, Paragraph4). Thus, villages are left to their own devices to handle disagreementswithin their own communities.

Finally, as with other government policies, villagers in remote areas are unlikely to be aware fully of the Law on Complaints and Denunciations. 4.2.8 Financial Oversight

The CBRIP project would provide funds directly to a large number of communes. This calls for a carefully designed system to minimiseleakage and misuse of funds. Views are divided as to whether communescan handle a project budget. For Box41: Different Perspectives the most part, districts are not always District cadre: 'No, communesdo not have convincedthey can, though most stated that experienceto handlethe largesums of t moneyinvolved in infrastructureprojects-say hey think they should. Communeswere more up to 200 millionVND. You can't hopeto give confident that they would be able to do so, if the subprojectbudgets to the communes.' they receive support in financial management. Communecadie: In my commune[a remote minoritycommune in a mountainousdistrict of Many of them already handle amounts over Thua Thien Hue] we annuallyhandle over 300 and above the budgetsthey would be receiving millionVND.' for subprojectsunder the CBRIP. There was (stakeholderworkshop, Nha Trang) widespreadagreement among local peoplein the communesthat disbursingthe budget to the communeis completelyappropriate.

There is no public display,and very limited local knowledge,of actual finances and budgetsat communelevel. Communebudgets should be publicly posted (accordingto Decree 29 on GrassrootsDemocracy), but this had not happenedto date in any of the communesvisited. Communeauthorities are frequently aware of partial budget details only for the projectsthey are involved in (see box below). Box 42: Sheet Iron and Nails-but No Accounts The biggestconstruction project to take place in the Rac Glai and Xie Trieng ethnic villagesthat make up Son Thai Commune(Khanh Vinh District,Khanh Hoa Province)in the last five years has been an initiativeto build or upgradeevery house (all 125 households)in the commune.This has beencarried out through the govemmentfixed settlementprogramme. Village peopleco-operated with each other to build or repair their houseswih the material providedby to them by the authorities.Approximately 90% of householdshave now repairedtheir existing houseswith sheet iron and nails. The remaining10% of householdshave new houses built by the govemmentvalued at 7 millionVND each.

Althougha large part of the work was locally managed,no knowledgeexists within the communeof the financialaspects of the project. Eventhe CommunePeople's Committee did not know how much moneywas fundedfor each household,or for the entire commune.They didn't know the pricesof constructionmaterials; the only thing they were asked to do was to sign receipts,to show how manytons of sheet iron and nails they had received. (Dau Quoc Anh)

67

PARTFIVE: PROPOSEDSTRATEGY

The final part of this reportlays out a strategy for incorporatingthe goals of the Social Assessmentinto the CBRIP.This strategyis designedto:

* Ensurethat vulnerablegroups will benefit from the project; * Incorporatethe specificneeds and incorporatethe local knowledgeof ethnic minorities; * Build uponthe recognisedstrengths of the full rangeof stakeholderscurrently involved in implementinginfrastructure works.

The strategyhas been developedso that it is consistentwith existingpolicies and laws of Government.Where existing laws are not consistent,suggestions have also been made as to how policychanges could form a firm foundationfor implementationof the CBRIP, and for other projectsbased on similar principles.

5 PART FIVE: PROPOSEDSTRATEGY

5.1 CommuneClassification

The CBRIP preparationdocuments suggest classifyingthe project communesby degree of existing capacity to implementgovernment works, with an appropriate level of external support then given to each category.

This study has shown that communecapacities differ by type rather than degree. Very remote communeshave high levels of self-reliance,but very little experience and low levels of understandingof governmentpolices and development programmes;all upland minority communeshave 'lower capacity' in working within governmentprogrammes, but relatively high capacity in managementand implementationof their own works usingtheir own forms of social organisation; challengesto the implementationof the CBRIP in in-migrantcommunities has more to do with low communitycohesion than low capacity.

This communeclassification recommends distinct types of support for two types of communes:communes with no vehicle access, and communes inhabitedby both recent in-migrantsand indigenousgroups. The basic participatoryapproach, with special attentionto ethnic minority needs outlined in the training requirements, should apply in all other communes,including the more accessibleminority communes,lowland communes and coastal communes.

Aq Table 18: CommuneClassification Commune Prevalence Characteristics Specificrisks for SpecialSupports Classification CBRIP Type A: 10 - communesin the - risks of benefitsbeing - subprojectproposal to be Communes communes processof rapid captured by migrants reviewedby with in- social change; at expense of anthropologist/ethnic migrantsand - difficultiesin indigenousgroups; minorityspecialist before indigenous coming up with - risks of improved approval can proceed(see groups plans that express infrastructureattracting 'subprojectselection'); the needsand further migrantsand - CommunityFacilitator prioritiesof the increasingpressures (CF) must facilitateall communityas a on indigenousgroups commune meetingsfor whole; and the environment. subprojectselection to - language ensure both migrantsand barriers most indigenousgroups are likely in reaching represented; indigenousgroups - close monitoring(studies rather than spot-checks) required throughoutproject cycle Type B: 9 - situatedfrom - simplytoo far away - CFs for these communes Communes communes severalhours' to 2 for district cadresto hiredon basis of with no days' walk from justify, on their commitmentto reach vehicle distnct centres; salares, walking the communes;financial access - least experence hours or days it takes incentivesfor CFs in these of implementing to reachthese communes; govemment communes; - close monitoring(studies projects:few fixed - similar disincentives ratherthan spot-checks) settlementor 135 for contractors; requiredthroughout project projectsreach - highercosts of cycle here; implementationdue to - high levels of self transportation reliance; difficulties - language barriers in reachingmany people _ Type C: all other Standardprocedures recommended in this proposalapply to all these all other communes communesexcept for those laid out for workingwith ethnic minoritiesin project ethnic minorty communes. communes

* These numbersrequire further clarificationby MPI. NB: See Appendix 11.

Type A communesare communesin which more than 20% of the community is made up of in-migrantsarriving in the last five years. Communes made up of 100% migrants are consideredto be Type C communes.Special supports for both Types A and B apply for in-migrant communeswithout vehicle access (if any). 5.2 ProposedGround Rules for CBRIPSubprojects

5.2.1 Ground Rules to Ensure CommunityInvolvement

The basic ground rules for subprojectsneed to be kept as simple as possible. Complicatedrules and regulationsdecrease local capacity to carry out subprojects effectively,thereby limiting local participation.

70 Table 19: Ground Rules for CBRIP SubprojectImplementation Stageof GroundRules ProjectCycle Information 1. CBRIPinformation that is in the publicinterest must be publiclyposted in villagesand communes. Dissemination Subproject 2. Village-levelconsultations on subprojectchoice must be heldwithin each project commune. Choice 3. At leasttwo consultations must be held:one with women in the village,and one with all adultsin the village. 4. Inthe firstyear of projectimplementation, communes must make a provisionalchoice on howall threetranches will be spent. Meetings 5. All votersmust be invitedto all project-relatedmeetings in the villages. 6. Representativesfrom every village must be presentat commune-levelCBRIP meetings. 7. Decisionscan only be passedif 2/3of thoseinvited to villageand commune meetings are in attendance , andif over50% vote in favour. Management 8. Communesmust form a CommuneManagement Committee (CMC) for the durationof the project,comprising cadres with signing powerand financial experience, as wellas an electedrepresentative from each village. The CMCwould form the coreof the MaintenanceCommittees as needed(where there are commune-levelworks). 9. Villagesmust form Village Management Committees (VMC) for theduration of village-levelsubprojects, which would form the coreof MaintenanceCommittees upon completion of construction.Each Village Management Committee must have at leastone non-official as a member. Design 10. Contractorsmust consult communes on subprojectdesign and location; Commune Management Committee must approve subproject designand location. 11. Targetedconsultations must be heldon designwith those interest groups most directly affected by a particularsubproject. This may include,for example,women (for water supply), parents of schoolchildren (for schools), or the elderly(clinics). Labour 12. Localpeople must be paidfor their labour on the CBRIPsubprojects; payment must be at the marketrate. 13. The CommuneManagement Committee must approve the hireof anymanual labour from outside the communeto workon subprojects. Maintenance 14. MaintenanceCommittees (and water users' groups where relevant) must be established for eachsubproject. Conflict 15. Complainant(s)must present a petition(oral or written)to theCommune Management Committee, Village Management Committee or Resolution MaintenanceCommittee. Petitions may be madeat any time. 16. The Communeor Village ManagementCommittee must respondwithin one week of receivingthe petition.

' This is in linewith the GrassrootsDemocracy Decree, which stipulates that meetingsfor works to be directlydiscussed and decided by the peoplemust have the participationof at leasttwo-thirds of the peoplesupposed to attend(Chapter 1I1, Article 8, Paragraph2).

72 5.3 RecommendedChanges to CBRIP Design

5.3.1 Changesto the List of EligibleWorks

5.3.1.1 Revised List of Eligible Works The table overleaf shows recommendedchanges to the list of eligible subprojects under the CBRIP. The original draft list, "EligibleWorks for Commune Infrastructure,"31 appears in italics.

31 Summaryof ProjectDesign Features, Annex 1, Attachment2.

71 Table 20: Recommended Changes to List of Eligible Works Revislons/Changes ProposedCommune-Level Subprojects 1. Improvements/upgradingof existingroads not involvingchanges of alignmentor > Clarifywhether communes with no vehicleaccess would be road extensions(improvement in critical sectionswith gravel,drains culverts,etc.; and allowedto improvetracks so that they would be passableby Irish crossings<50 m, single span bridgesand truss bridgesof <15 m). four-wheeledvehicles. 2. Footpath and track improvement(including provision of footbridges<30 m; cuttingof > Add improvementsto existing (transportation)waterways. steps in steep sections;and widening). => Covercosts of spans up to 200 m. 3. Improvement/upgradingof temporaryrun-of-river diversion schemes (serving 5-20 > Checkassigned costs: US$3000may be too low for reservoirs. ha) with gabionsand stone masonry,main canal and essentialcontrol structures up to a maximuminvestment of US$2000per ha. 4. ConstructionAmprovementof small reservoirsserving <30 ha up to a maximumtotal investmentcost per hectare of US$3,000. 5. Flood conlrol works serving at least 50 ha of agricultural land or 1)00 families. 6. Gravity drinkingwater schemes(with appropriatesanitation) serving a minimumof > Allocatehigher sumsfor water supply. Costs can go 30 families at a cost of Checknumber of householdsper well. It is highly dependent upon local geography:in some places it may be logicalfor only two households-if they form one cluster in a widely scattered village-to share a well. Ten householdsper well is rather high. 8. Connectionof the communesto the nationalelectricity grid providedthis requires<5 > Add connectionto nationalelectricity grid from communecentre km of high voltagelines (400V)and one transformer. to villages. 9. Construction of small hydropower schemes of 5-50 kW serving between 20 > Clarifynumbers of householdsper hydropower generator: 20 seemsvery high.

74 RovislonslChanges and 200 families at a maximum cost per family of US$50-200. Rov_slon_ _Changes 10.Replacement of existing'leaf schools' with permanent structures. = Permitvillage-level classrooms (lop hoc)-serving the lower 11.Repairs and upgrading of existingschool buildings and health centres. primarygrades-to be builtor upgradedusing the grant.32 As in 12.Construction of newkindergartens and primary and lower secondary schools 12.,the districtmust undertake to provideteachers. includingcommune semi-boarding schools, provided district undertakes to provide = Ensureboth day care centres (nha tre) andkindergartens (mau teachers. giao)may be builtor upgradedusing the grant. = Allowcommunes with poor internal access to builda second communeclinic if this significantlyimproves access for remote villages.The districtmust undertake to provide/trainhealth staff. > Includetoilets in the designof all schoolsand health clinics. 13.Provision of public toilets and water supply to existingmarkets in communecentres => No change. at a costof Checkcost: too high.33 => Definepurpose: to be usedfor a widerange of community activitiesfrom public meetings to culturalevents to agricultural extensionactivities. > Addvillage-level meeting houses. > Includepublic toilets in the design,to be incorporatedif thereis no publicfacility in adjacentbuildings. > Incorporateplaying grounds into layout of standarddesigns. 34 ProposedDistrict-Level Works 1.improvements/upgrading of existing roads (including Irish crossings, bridges and Addto communelist as "inter-communeworks" (see below). culverts). 2. Extensionof thenational electricity grid to onecommune or a groupof communes wheresuch extension exceeds 5 km or requires< 1 transformer. 3. Improvements/upgradingof existing irrigation systems.

32 It is notclear in the threeoptions above whether this would be possible. Some people took 11/ to meanthat these classrooms (which are frequently made of leaves)could be replacedby permanentstructures-but then pointedout that this was illogicalas it meansthey would haveto builda leaf housein orderto be eligibleto upgradeit to a permanentstructure! 3 This statementis based on informantsin several(mountainous) communes putting the constructioncost of a 'Kinh' houseof bricks with a tile roof at 20-50million VND. 34 The existenceof playinggrounds or soccerfields was seasonalin manyvillages visited: temporary playing fields come into use oncecrops have beentaken off the land, and revertback to agriculturaluse in the plantingseason. This maydepend then on the availabilityof suitableland.

75 5.3.1.2 Inter-communeWorks Under current proposals,district-level works would allow districtswith three or more project communesto be providedwith funds equivalentto 20% of funds allocatedto the communesin the district, to the three types of inter-communeworks serving the population of two or more communes.35

The Social Assessmentteam has two reservationsabout the district-levelworks:

1. The district-levelworks mar what is otherwisea relativelysimple project design. The simpler the projectand the rules of the game, the more likely they are to be understoodand taken up in the way in which they were intended. 2. The district-levelworks go against the principleof encouraginggreater controlat communelevel. The proposal puts both decision-makingand managementback at district level. This may result in the failure of efforts at decentralisation,as it reintroducesan opportunityfor districts to 'take over' from communes as soon as they judge communesare not up to the job; it detracts from participatoryfeatures of the project design elsewhere.

Recommendationsfor district level works: > Re-namethem 'inter-communeworks' and add them to the commune list of eligible works; => Allow communes (ratherthan districts) to decide whether they want these works undertaken; => Place project managementand financial managementin the hands of the districts; > Require the same proceduresfor public posting of informationin the communes (as used for commune-levelprojects).

5.3.1.3 Further Remarks The inclusionof only shared public infrastructure works is sound. The inclusion of infrastructureworks directed at individual households(electricity connections or wells to individual households,for example) is to be avoided,as this lessens the risk of benefits being concentratedin the hands of dominant personsor groups.

Costs of infrastructureworks vary significantly betweendifferent geographic zones. The Ministry of Constructionacknowledges this by issuing guidelines on different investment levels for infrastructurein mountainous,midland and lowland areas.36 Financialentitlements issued to communesonly on the basis of population cannot be consideredto be equitable as mountainousareas will not be able to build as much with their money as lowland areas.

The final Vietnameseversion of the list of eligible works must be checked carefully.Translation of the draft list from the English (which appears to be the

35 At the time of writing it was not yet clear whetherthis would involveonly projectcommunes, or other communesadjacent to projectcommunes. This needsto be clarified,because if communeswhich are not projectcommunes are eligibleto be included in proposalsfor inter-communeworks, measures need to be introducedto bringthem into the decision-makingprocess. 36 See Vien Kinh Te Xay Dung, Suat Dau TuXay Dung Co Ban, Hanoi, 1993.

77 original)to the Vietnamesecontains a few key errors (for example,costs for small reservoirsare given as cost per ha in English, but as flat rates in Vietnamese; 'kindergartens'in English is translatedinto Vietnameseas 'day care', etc.). The wording of the final documentsneeds to be checked carefullyto avoid problems arising merely from mistranslation.

5.3.2 Phasing

Project pilotingand phasing is essential.It is the key instrumentthrough which the CBRIP: * can respondto community needsand preferenceson an ongoingbasis; and * can thereby make necessaryadjustments to projectdesign and processes.

Errorsin carryingout pilots include: * 'piloting'too large an area; problemsoccur across large areas that could have been prevented; * 'piloting'only part of the process,so that problems cannotbe correctedin later stages of the project.

Recommendationsfor phasing: > Determinethe numberof phase one communesby the numberof Community Facilitatorswho can be effectivelytrained in the first year (under20); FFeed lessonslearned through phase one and each subsequentphase into a mandatoryreview of project experienceat the end of each year, and make necessaryadaptations; => Use communesintroduced in earlier phases as training sites for subsequent phases (and use skilled CommunityFacilitators from previousphases to assist in training).

Below is a suggestedtimetable for phasingthat would allow project completion within five years:

Table 21: Suggested Project Phasing YearOne YearTwo YearThree YEARFOUR YearFive Communes Communes Communes YEARUFOU Communes 2001 2002 2003 COMMNES 2005 Phaseone 48 48 48 4 Phasetwo 240 240 240 Phasethree 244 244 244 Totalcommunes 48 288 532 484 244 Community 12 72 120 108 61 Facilitators I_I_I

Recommendationsfor choosing phase one communes: => Include the communeswhich participatedin the field study, stakeholder workshopsand selectivesurvey; => Include at least three Type A and three Type B communes; => Choose-as far as this is possible-communes close together; And/or begin work in 2 or 3 provincesonly in Year One.

78 5.4 Project Preparation

5.4.1 Selectionof CommunityFacilitators

It is estimatedthat grantswould be availableto train andemploy some 120 "CommunityFacilitators" (CFs) to serveas key resourcepersons in providing supportto projectcommunes. Good facilitators skilled in communitydevelopment are vital to the successof the CBRIP,and a numberof difficultiesare to be expected in seekingout suitablecandidates.

Table22: Recommendationsfor HiringCommunity Facilitators: Recommendations Justification > Advertiseand interview for positions; * Tendencyin Governmentto targettraining => Hireon the basisof previousexperience or supportto peopleon the basisof theirnot trainingin community-baseddevelopment, havingatiended training courses before, andcommitment to traveland to spend providinglittle opportunity to buildon longperiods in assignedcommunes and existinglearning; remotevillages. . Existenceof manypeople with some = (Thesemay be younggraduates or older experienceand/or training in community peoplewith lengthy work experience; they development,but littlechance to buildon maybe peoplein government,NGO or it.37 privatesectors.) => Paycompetitive salaries with financial . Districtcadres lack motivation toreach very incentivesfor those CFs working in the distantcommunes and remote villages due to mostremote areas. lowsalaries and lack of travel allowances. Provideaccess to motorbikesfor transportationto projectsites when this is necessaryand feasible. > Providebudgets for travel. = Encouragequalified ethnic minorities to . Lowawareness ofissues related to local apply; knowledgeofethnic minorities; => AssignCFs who understand minority . Languagebarriers in communeswith languagesto relevantcommunes. ethnicminorities. =, Activelyrecruit women; . Travelrequirements may prevent women > Requireall CFs to traincommune and fromapplying for CF positions; villageWomen's Union leaders (or other . Musthave backup mechanism for ensuring appropriatewomen in the communes)to maleCFs can includewomen in project facilitatewomen's meetings (see guidelines consultations. for CFs).

Suggestedguidelines for CFs: * Readthis SocialAssessment; * Hold targetedconsultations with mostmarginalised groups on projectchoice and projectdesign, or instructlocal massorganisations or otherappropriate individualswithin the villagesand communesto do so; * Travelto and work in all villagesin theirassigned communes; • Reviewand reviselist of infrastructurethat is of particularbenefit to marginalised groupsas partof their normalwork;

37 For example,the NationalCommittee for the Advancementof Womentrained approximately 10 womenin each provincein Vietnam(from variousoffices at provincial,district andcommune levels) to stand as candidatesin electionsfor the People's Councilsin November1999. Topics that were includedin this trainingare directly relevantto the CBRIP:stakeholder analysis, consultation,and bringing gender perspectives to local councilsand administrations;the NGOTraining Project regularly holds similar coursesdown to communelevel, and cadresfrom someof the 532 projectcommunes can be expectedto have received training from this source.

79 * Do seasonal/workcalendars with communitiesin their assigned communesto determine appropriatetimes for meetings; * Identify and train women (either Women's Union staff or other appropriate women) to run women-onlymeetings; * Attend biannualmeetings with all other CFs to exchange experiencesas team members.

Recommendationsfor assigningcommunes to each CF: = Assign a cluster of communes,adjacent or accessibleto one another, to each CF (one CF may work in more than one district in some areas);38 > Assign communesaccording to workload and remotenessof communeand villages (in accessible lowlandand coastal communitiesit may be reasonableto have facilitatorsoversee more than 5 communes;in extremely remote communes,CFs may not be able to reasonablywork in more than two communes); => Hire CFs full-time.

Recommendationsfor ensuringthe timely inclusionof community developmentprinciples into the project: => Hire CFs as soon as possible; > Have CFs in place and trained before informationdissemination begins (they will be key players in it). 5.4.2 Training

5.4.2.1 Support Required Recommendationsfor training in community-baseddevelopment approaches: => Training for CFs in communitydevelopment tailored to the CBRIP; => Training for Commune ManagementCommittees in project management; => Basic training for district technicians in community development.

Training for CommunityFacilitators should be project-specificand practical,with emphasis laid on convincing CFs of the value of argumentsfor participationwithin the context of the project.

Recommendedtopics for training of CFs: => Principles, objectives and ground rules of the CBRIP; = Content of this Social Assessment; = Awarenessof relevant ethnic minority issues: notions of 'backwardness',the importanceof local knowledge,awareness of the value of local designs, constructionskills, and forms of social organisationfor managementand maintenance; > Awareness of relevant gender issues; =, Awareness of issues related to in-migrationand indigenousgroups; => Awareness of poverty and marginalisation; > Skills to elicit opinions of marginalisedgroups; > Strategiesfor conducting participatoryconsultations;

38 The CF can also facilitatenegotiations between adjacent communes if an inter-communeproject is chosen.

90 => Strategiesfor speakingthrough translation; > Techniquesfor organisingtargeted consultationsof interest groups (especially women), and for training others (mass organisations,key village people)in these techniques; > Understandingof bidding processes; X Understandingof financial managementsystems; => Developmentof projectmaintenance systems; => Understandingof all GOV policies relevantto the CBRIP.

Recommendedtopics for training CommuneManagement Committees: => bidding procedures; => quality control; => financial management.

Recommendedtopics for training districttechnicians: > basic awarenessof conceptsand benefits of participation; =* how to incorporateparticipation into their work. 5.4.2.2 Study Visits Recommendedstudy visits: => For district techniciansand administrators,provincial PMU, and CFs: to visit sites in projectcommunes where local people have built infrastructureworks without governmentassistance, to learn from design, construction,and organisationfor managementand maintenanceof such works; > For CFs and communesin second and subsequentproject phases:to visit and learn from communesin previousphases. 5.4.2.3 TrainingProviders Constraintsto findingsuitable training providersfor CFs are: > A lack of experienceamong trainers themselvesin providingthis type of training; => A small numberof appropriateinstitutions; and z> Training providers'own need to rely on outside expertiseto assist in developing adequatetraining courses.

Training providerscapable of taking on this training, provided outside support is available, include Ho Chi Minh City Women's Union Training School, Hue University and Hue AgriculturalCollege. Please note these suggestionsare not based on in- depth assessmentsof these institutions,but on the basis of these institutionsbeing referredto the Social Assessmentteam.

Recommendationsfor training provisionin community-basedapproaches: => Hire two training providersfrom two urban sites in the project area. This could comprise one site in HCMC and one in Hue, with the advantagesof reducing geographicalconstraints in the southern part of the projectarea, and addressing north-southcultural differences. => Providetraining providerswith outside expertiseto assist in puftingtogether a CBRIP-specifictraining package:the NGO Training Project is a suggested resource with the expertiseto do this.

Ri Training providers should be encouragedto take the training to provincial or even district DPls, and conduct it there, rather than bringing CFs down to the cities. This could help to: * incorporateparticipation into the mainstreamof the project; * allow trainers to see the context in which they are training people; * allow visits by CFs to some of the project communes.

5.4.3 Policy/Advocacy

5.4.3.1 Policy Changes

Table23: SuggestedPolicy Changes in Supportof the CBRIP Policy RelevantArticle ProposedChange Supportto social goalsof CBRIP? DecreeNo. 29 all articlesthat refer to > Makeinvitation of Wouldprovide policy communeand village all voters backingto encourage meetingsand provide an compulsoryto the inclusionof optionto invite'people', community womenat community ,voters', or householdheads meetings. meetings. to meetings. DecisionNo. Articleone, General => Deletethe word Projectsuccess 135/1998/QD-TTg Objective:"...create 'backwardness.' requirespolicy that of July 31, 1998to conditionsfor theseareas to explicitlysupports Approve the overcome... backwardness." ethnic differenceand Programmeon makesuse of local Socio-Economic knowledgein Developmentin subproject Mountainous,Deep implementation. Lyingand Remote Communeswith SpecialDifficulties Lawon Complaints Chapter2, Section2, Article > Personor agency Wouldcreate better andDenunciations 19: "ThePresidents of the otherthan the conditionsfor local (No.09/1998/QH10 People'sCommittees of communePC peopleto voice of December2, communes... are competent Presidentshould grievances,if 1998) to settlecomplaints about be givenpower to necessary,against the administrativedecisions settlecomplaints communePC: this will and/oradministrative acts of relateddirectly to bevery importantif their own [SAemphasis] as this person. the communePC is to well as thatof responsible bethe mainagency personsunder their direct implementingthe management." project.

Regulationson use -> Streamline Neededto make of localmaterials in procedures. constructionof infrastructure infrastructurebetter construction timedto includelocal labourand allow local materialsto be used.

Bidding -> simplify Will assistcommunes procedures; in understandingand usingbidding procedures. LandCertificates => Clarifythat both Strengthenswomen's andHousehold maleand female legalposition. RegistrationCards namesare In areaswith |__ __permitted on matrilocaland Policy RelevantArticle ProposedChange Supportto social .______goalsof CBRIP? thesedocuments; matrilinealtraditions, > Includeboth givesstate recognition spousesnames to traditionalproperty on documents. and inheritancerights. Policyon cadres' => LobbyMinistry to Necessaryto enable salaries(Ministry of allowhigher CFsto reach Finance) salariesfor CFs in extremelyremote remoteareas than communes. permittedunder regulations,and convincethem of valueand importanceof this.

5.4.3.2 Advocacy Issues In order to ensure that marginalisedgroups benefitfrom the planned infrastructure, two key points must be addressed:

1. Poor and marginalisedgroups cannot benefitequally if they cannot pay for the servicesfacilitated by the CBRIP; 2. Benefits will be non-existentor greatly reduced if soft services (of teachers, extensionists,health workers) are either not provided or are of poor quality.

The CBRIP should integrate its work with:

Measuresto improvethe quality of educationfor children in the project communes,and access to educationfor poor children,including: * Improvingthe provisionand quality of teachers in remote areas; * Promotingthe provisionof kindergartenteachers and teachers for village classrooms (lop hoc); * Promotingteacher training for ethnic minority teachers, particularlyin the early grades; * Lifting regulationson minimum numbersof children required to hold secondary school classes; * promoting multigradeclasses, and teaching of Vietnamese as a second language; * promotinggirls' education; * promoting measuresto reduce the real costs of educationto poor households.

Measures to reducethe real costs of health and to improvethe quality of healthcare,particularity preventative health, in the projectarea.

Initiativesto improve the socio-economicsituation of the poor in the project communes,including extending credit and agriculturalextension services.This will allow the poorest householdsto make good use of roads, irrigation,etc., constructed under the CBRIP to improvetheir householdeconomies.

83 5.5 Rolesand Responsibilities

Community Facilitators: * To ensure that a participatoryprocess is actually used in all phases of the project cycle.

Village Management/MaintenanceCommittees: * to ensure villagers in their village are fully representedin all project processes; * to ensure appropriatenessand sustainabilityof the infrastructure. CommuneManagement/Maintenance Committees: * to ensure that all villages are fully representedin all project processes; * to ensure appropriatefinancial management; * to support village managementand maintenancecommittees; * to ensure the appropriatenessand sustainabilityof the infrastructure.

Commune People's Committee: to respondto any grievancesthat cannot be resolvedwithin the communeor village managementcommittees

District DPI: * to conducttechnical appraisal and approval of communesubprojects; * to supply technical support, as requested by the projectcommunes, during construction; * to manage inter-communesubprojects if these are selected by the project communes.

District People's Committee: to respond to any grievances that cannot be resolvedwithin the project communes.

Province DPI: to ensure effective linkages betweenthe CBRIP and other donor and government programmes;especially in the provisionof basic social services (health and education),agricultural extension services and credit.

MPI: * to ensure the highest quality of project preparation,particularly recruitingand arrangingfor the training of the first group of CFs in phase one; * to ensure the community-basedapproach is carried through project implementationmonitoring and evaluation.

PA 5.6 The Project Cycle

5.6.1 InformationDissemination

5.6.1.1 Disseminationof Project Information Recommendedstrategy for disseminationof CBRIP information: > Use multiplechannels to disseminateinformation; > Emphasiseunderstanding by the target audience; => Target for an audiencethat thinks in very practicalrather than highly conceptual terms.

Recommendedtopics for informationdissemination: = Descriptionof CBRIP; r Summaryof key governmentpolicies; > List of projectcommunes and criteriafor choice; = Project objectives; > Financialarrangements; = Ground rules.

Recommendedchannels for disseminationof project information: => Verbal:through CFs, communeauthorities, village leadersand mass organisations; => Printed factsheets:published in clear simpleVietnamese, 39 distributedin communes; => Informationvideos: live-dubbedin local languages; Provincialnewspapers and Phu Nu.40

5.6.1.2 Disseminationof SubprojectInformation within the Project Communes Ruilef . CBRIPinformation that is inthe public interest must be publicly posted in villages and communes.

Recommendedprocedures for public postings:

- Should be posted in public places (in the communecentre at the UBND or a centralschool) and within each village in the projectcommunes; > Should be postedas soon as the projectbegins-following disseminationof projectinformation; > Shouldbe updatedat regular intervalswith relevant information; > Shouldcontain minimal but sufficientinformation and be simple and clear; => Must be recordedin a standardformat (to be designedand suppliedto each projectcommune).

Recommendedinformation to be includedin the public postingsat various stages of the projectcycle: > each village's subprojectproposals; = final projectschosen by commune;

39 The Cham are trie one minority group within the project area who have a script that is actively used and quite widely understood; however, high levels of literacy in Vietnamese among the Cham mean it is not necessary to print the information in Cham: Cham who are illiterate are likely to be illiterate in both languages. 40 "Woman"-a popular women's magazine

85 contractorschosen, and their addresses; > design proposals(map showing position of project,diagram of construction); > statementof CMC approval of design; = statementof CMC approval of hire of any outside labourers; => wages local peopleare paid and days worked; = names, gender and ethnicity of memberson the CMC and VMC; => names, gender and ethnicity of memberson the MaintenanceCommittee; > amount of budget assignedfor the subproject,and amountspent to date; = levels of maintenancefees or contributions.

5.6.2 SubprojectSelection

Rule2. Village-levelconsultations on subprojectchoice must be heldwithin each project commune. Rule3. At leasttwo consultabonsmust be held:one with women in the village,and one with all adultsin the village. Rule4. Inthe firstyear of projectimplementation, communes must make a provisionalchoice on how all threetranches will bespent

Recommendationsfor the processof subprojectselection: => Every village producesa proposal; > All village proposalsare presentedat commune-levelmeetings; => CMCs (with one representativefrom each village-see sectionon management below) vote on final choice; Village proposalsare reviewedto draw up provisionalchoice for subprojectsin years two and three.

Recommendationsfor local consultationprocess: CFs, in conjunctionwith communeand village mass organisations,should: =-. Ensurethe ground rules are adhered to; => Identify,with local people,the most convenienttimes for holding consultationsto ensure maximumattendance; => Arrange targeted consultationsfor women, and encouragewomen to attendopen meetings; = Encouragecommunities to consider projectsthat are particularlybeneficial to certain groups (old people, women, children)within the community;

- Encourage communitiesto discuss potentialbenefits of all types of eligible works.

Recommendationsfor subprojectselection in Type A communes: => CFs must facilitate all communemeetings for subproject selectionto ensure both migrants and indigenousgroups are represented; => Proposals must be reviewedby ethnic minority specialistlanthropologist.Approval cannot be given until specialist has judged that subprojectselection reflects interestsof both indigenousand migrantcommunities. Recommended that this specialistbe from Hanoi-basedresearch institute,and hired for duration of CBRIP to conduct this work.

5.6.3 SubprojectAppraisal/Approval

Recommendationsfor subprojectappraisal:

Rr District technical staff to carry out technical appraisalof the subprojects; = Appraisal proceduresshould be simplifiedto ensure timely implementation; > Costs of technical appraisal should be a clearly defined percentageof project costs, but should not be paid for directly by the projectcommunes out of their CBRIP tranches.

5.6.4 Selectionof ManagementCommittees

Rule8: Communesmust form a CommuneManagement Committee (CMC) for the duration of the project,comprising cadres with signing power and financial experience, as wellas anelected representativefrom each village. The CMC would form the core of theMaintenance Committees as needed(where there are commune-level works). Rule9: Villagesmust form Village Management Committees (VMC) for the duration of villagelevel subprojects,which would form the core of MaintenanceCommittees upon completion of construction. EachVillage Management Committee must have at leastone non-official as a member.

CommuneManagement Committees should: * Be responsiblefor all duties at communelevel, includingadministration, financial management,overall project managementand organising all commune-level consultations.

Village ManagementCommittees should: * Be responsiblefor all duties at village level, includingorganising village-level consultations,local labour arrangements,and maintenancecontributions.

Recommendationsfor promotingrepresentative Management Committees: CFs, in conjunctionwith communeand village mass organisations,must: => Encourage communitieswhich have constructedinfrastructure works in the past with no outside assistanceto model their managementcommittees on pre- existing committeesfor the previousworks; => Encourage the communityto put those people they trust and respecton the committees:village elders;women; those people respectedin the community; teachers (for schools) and health workers (including nursesand indigenous midwives)for clinics; > Encourage peopleto considerthe capabilitiesof, and give active roles to, the elderly or disabled on these committees; = Overseethe process of electionsfor these committees; => Ensurethat the final choice reflects the ground rules for management committees. 5.6.5 Financingof Subprojects

Recommendationsregarding financial transfers in support of the community- based approach: => Transfer all subproject budgetsto commune level. All communesin the project are able to handle budgetsfor the subprojects,if provided with adequate supports;transfers to this lowest administrativelevel, with reporting of the financial 'bottom line' directly to the community,will help to reduce leakage and misuse of funds;

R7 => Ensuretimely disbursementof grants. Essentialfor ensuring construction takes place in the appropriateseason, when materialscan be brought in and local people are free from agriculturaltasks to participate.Late disbursement results in rushed implementationand poorquality construction,and reduce opportunitiesfor local participation; = Flexibilityof funding to accountfor higher infrastructurecosts in remote areas.

5.6.6 ProjectDesign

Rule 10: Contractorsmust consult communes on subprojectdesign and location; Commune ManagementCommittee must approve subproject design and location. Rule11. Targetedconsultations must be heldon designwith those interest groups most directlyaffected by a particularsubproject. This may include, for example,women (for water supply),parents of schoolchildren (for schools), or the elderty(clinics).

In orderto promotesubproject design that respondsto local needs and priorities, CFs, in conjunctionwith communeand village mass organisations,must: = Encourageopen discussionsabout how marginalisedgroups within the communemight benefit and how remotevillages might benefit; => Encouragecommunes to exploreoptions for buildingshared commune infrastructure(schools, health clinics, markets)in locationswhere they will benefit the greatest numberof peoplein the commune,and not necessarilyin the administrativecentre of the commune; z Encouragedesigns that benefitas many householdsas possiblewithin the communeor village; => Encourageminority communitiesto point out ways in which infrastructure constructionmay violate local beliefs, and to suggestappropriate changes; = Encouragepeople to state their design preferences,especially in terms of traditionaldesigns; => Emphasisethe benefits of using local materialsfor better maintenance.

5.6.7 Use/Maintenance

Rule14: Maintenance Committees (and water users' groups where relevant) must be establishedfor eachsubproject

Requirementsfor MaintenanceCommittees: => Membersmust be elected; > There must be a realisticfinancial plan for sustainablecontributions; => There must be an effectivecollection plan that takes into account different degrees of usage of infrastructureand poverty of villagers; > There must be a commitmentfrom higher levels for assistancewith maintenance in times of great need.

Recommendationsfor incorporatingthese requirements: = Allow communitiesto decide for themselveson levels and kind of contributionfor maintenance;

PA => Developa simple list of estimated costs for maintenanceand repairs to assist communitiesin calculatinglevels of contributionsthat are both locally affordable and sustainable;this can be done by the economic team; = Provide a packageof advice on troubleshootingand simple repairs for local people.This can be disseminatedthrough CFs and printed on simple leaflets.

5.6.8 Conflict ResolutionMechanism

Rule 15:Complainant(s) must present a petition(oral or written) to theCommune Management Coummittee,Village Management Committee or MaintenanceCommittee. Petiions may be made at any time. Rule16: The Communeor VillageManagement Committee must respond within one weekof :ireceivingthepefition.

Existing legal instrumentsregulating complaints and denunciationsat district and communelevel 41 => District-and commune-levelPeople's Committeesshould have one day free per week when they can receive complaintsfrom the community; > District/communePeople's Committeechairpersons are responsiblefor settling complaintsor denunciationsof their own activities or illegal actions, as well as those of people and agenciesunder their jurisdiction; => The FatherlandFront and citizensare responsiblefor supervisingthis process.

Recommendationsfor complaintsmechanisms to work effectively: => Local people must be made aware of the mechanisms; = Local people must be encouragedto use the mechanisms.

If faced with inaction or an impasse,complainants should then be encouragedto go to higher levels using GOV laws (below).

5.6.9 Monitoringand EvaluationMechanism

The mechanismproposed here is a system to monitor and evaluatethe community- based approach. It is designed to:

* Monitor whether genuine community participationis taking place and whether benefits of the infrastructureworks are reaching not only the better-offmembers and more easily accessibleareas, but also the most marginalisedpeople within the communitiesand the more remote areas of the project communes; * Anticipateand identify problems:areas where these processesare not happening (or not happening well); and * Feed this informationback into a system for improving CBRIP design in subsequentphases.

41 Law on Complaintsand Denunciations(No. 09/199&'QHof December2, 1998).

Ra 5.6.9.1 Reporting Reportingon a community-basedapproach must originate from communeand village levels.

It is recommendedthat reportingbe made: = To the local community(and by the local community)through public postings(see section on informationdissemination); > To higher levels: photocopiesof the original public postingsand somewhat more detailed communereports to be sent up through all administrativelevels.

Village and CommuneManagement Committee reports to higher levels: > Must be created at clearly defined intervals; * Include photocopies of the public postings; and > Include a brief report on problems encountered in involving the community in the subproject. District and provinciallevels should answer the same questions in their reporting.This feedback should be the trigger, along with other aspects of monitoring,for identifyingproblems, and for providingappropriate additionalsupports as necessary. > Should be signed by all Management(and subsequentlyMaintenance) Committeemembers; => Must be sent up (as a photocopyof the original)through all administrative levels; districtand provincial reportscan then be attachedto these as they move up throughthe administrativelevels; > Receipt of satisfactoryreports could act as a trigger for disbursementof subsequenttranches.

Four suggestedstages after which reportingshould be requiredare: 1. subprojectselection; 2. design; 3. constructionfinished and maintenanceplan developed; 4. six months after works are put into use.

5.6.9.2 Monitoring Two types of communesrequire very close monitoring:

Type A (in-migrantcommunities): to ensure that decisionsmade reflectthe needs and prioritiesof the communityas a whole, that both indigenousand in-migrant groups benefit from the subprojects,and that infrastructureconstruction does not place undue pressureon the indigenousgroups.

Type B (extremelyremote communes):in order to ensure that the community- based approach-and all other project activities-are reachingthese communes.

Adequate monitoringof these two types of project communeshould consist of: > In-depthperiodic studies of a sample of the project communeswithin that classification:TORs for each study should cover the issues outlined for each type on the box above.

qn The communesample chosen for Type A should cover all relevant provinces in the project area and be representativeof all relevantethnic groups; = The communesample chosen for Type B should cover all relevant provinces in the project area, and communeswith ethnic minority as well as Kinh in-migrants should be chosen.

The risks involved in implementingthe CBRIP in Type C project communesis somewhatlower. It is therefore recommendedthat monitoring in these communes consist of periodic spot-checksto ensure:

=> that the poorest and most vulnerable groups are benefitingfrom the subprojects; => participationof beneficiariesin all stages of the project cycle.

Monitoringshould occur periodicallythroughout the project cycle.

Monitoringexercises should be done by an external agency:for example, the Departmentof Sociologyat the Institute of Social Sciences (Hanoi),or the Departmentof Anthropology,Hanoi University.It is preferableto have an institution that is Hanoi-basedand has a long term (5 year) contractwith MPI to conduct this work. 5.6.9.3 Indicators While monitoringexercises for Types A and B will include much more qualitative assessments,both types of monitoringshould includethe collection of a numberof basic indicators.

Suggested indicatorsto measureproject benefits: 1. Percentageof 'hungry' and 'poor' householdsin villages where subprojectis implemented.42 2. Percentageof 'hungry' and 'poor' householdswith water points. 3. Percentageof 'hungry' and 'poor' householdswith land irrigated through CBRIP subprojects. 4. Percentageof subprojectsimplemented that benefit particularlyvulnerable groups (water supply, classrooms,kindergartens and day care).43 5. (For areas of high in-migration)Reasons why indigenousgroups (ho goc) are satisfied/notsatisfied with the progress of the project. 6. Reasonswhy women-headedhouseholds in the communeor project villages are satisfied/notsatisfied with the progress/resultof the subproject. 7. Reasonswhy poorest householdsare satisfied/notsatisfied with the progress/resultof the subproject.

Suggestedindicators to measure participation: 1. Percentageof women on Management(and then Maintenance)Committees. 2. Percentageof ethnic minoritieson Management(and then Maintenance) Committees. 3. Percentageof non-officialson Managementand MaintenanceCommittees. 4. Percentageof costs spent on local labour.

42 Baselineavailable in PMU database(types D and C households). 4' Baselinedata for these indicatorsis availablein the PMU database.

QI 5.6.9.4 Evaluation It is recommendedthat: > An internal phase one evaluation(review and adaptation-see below) be conductedat the end of year one; > Similar reviews occur at the end of each year; = A final evaluationbe conductedby an independentagency.

The aims of the final evaluationshould be: => To evaluatewhether the aims of this Social Assessment have been met; => To make recommendationsfor future projects similar to the CBRIP.

5.6.10 Audits Districtsmust be responsiblefor auditing commune'saccounts for the subprojects.

5.6.11 Review and Adaptation

In order to ensure that the project design respondson an ongoing basis to local needs and priorities:

=> A project review should occur at the end of year one; > Similar reviewsto addressproblems and make necessaryadaptations should then occur at the end of each subsequentyear.

Processesthat should feed into project adaptation include: => Commune reports; => District and provincial reports; => Resultsof monitoring exercises.

5.7 For the Economic Team...

The Social AssessmentTeam recommends:

= Checking and revising costs of eligible works; => Factoringin higher costs involved in reachingextremely remote and upland communeswith seasonal road access, when calculating real costs of subprojects.These should not be calculated merely on a per head basis; => Developinga simple list of estimated costs for maintenanceand repairsto assist communities in calculating levels of contributionsthat are not only locally affordable but also sustainable; > Clarifying what levels of costs would need to be met by higher levels in emergencies. 5.8 Project Risks

The CBRIP rests on the premisethat local communitiesmust be included in the process of the CBRIP, and should reap the benefits.This section anticipates

92 situations in which the reversesituation may be created: that local communities,or importantstakeholders, may actually lose out. Anticipated negative impacts, and measuresto mitigate them, are presented in the table below: Table 24: Project Risks Risk Reasonfor Risk MitigationMeasures Poorco-operation and Participatoryapproach may be Clearsupport role defined for involvementof higher perceivedas a threatto the districts. administrativelevels administrativesystem in place; (districtand province) Withfinances and managementassigned to communelevel, district feels it is beingbypassed. Strengtheningand Weakimplementation of Community-basedapproach enrichmentof dominant community-basedapproach centralto the projectat all times; groups;further will resultin the CBRIPbeing Investtime (and money) in marginalisationof the implementedin samemanner trainingand choosing good mostvulnerable as top-downprojects. communityfacilitators and incorporatinglessons from phase one. Constructionof Improvedinfrastructure may Specialattention to monitoring infrastructurein in- attractfurther in-migrants, andresearch of in-migrant migrantcommunes may increasingpressure on communitiesbeyond the end of aggravateor createsocial indigenouspeople. the subprojectcycle (see section problems on monitoring). Lossof incomeandlor Landlost and/or houses Requirecommunity approval of landfrom irrigation reallocateddue to irrigation projectdesign, and records of projects. whowill loseout andhow they will be compensated; Landacquisition plan prepared withinWB. Infrastructureworks Poorquality construction or Economicteam to do realfigures unusableafter several inadequatemaintenance. on costof goodquality years construction; Economicteam to support maintenancethrough advice on calculatinglong-term costs of maintenance. Localpeople Poorerhouseholds cannot Advisecommunes to grade overburdenedby affordcontributions at same contributionsaccording to maintenance levelas better-offhouseholds; householdpoverty level/ability to contributions pay; Communecapacity to payfor Assurecommitment for periodic majorlong-term repairs may contributionsfor necessarylong- be limited. termmaintenance from state budget. Riskof difficultiesin Projectfunds are a loanto the GOVhas agreed and should CBRIPloan repayment, GOV adhereto 20/20Agreement. andof moneydedicated to socialservices being used for repayment. APPENDICES

Ia