Implementation of Water Supply & Sanitation Programmes under PRSPs Synthesis of research findings from sub-Saharan Africa Tom Slaymaker & Peter Newborne WaterAid Malawi: Steven Sugden, Linda Milazi, Boyce Nyirenda, James Longwe, Nelly Magelegele and Researchers: Nebert Nyirenda, Amalia Sevilla and Olivier Stoupy Researchers Uganda: Juliet Kanyesigye, Joseph Anguria, Edison Niwagaba and Tim Williamson WaterAid Zambia: Eric Gutierrez, Saviour Mwambwa and Researcher: Webby Wake WaterAid Tanzania: Dominick de Waal and Rehema Tukai WaterAid: Belinda Calaguas and Mary O’Connell Research Team August 2004
Transcript
Implementation of Water Supply & SanitationProgrammes under PRSPs
Synthesis of research findings from sub-Saharan Africa
Tom Slaymaker & Peter Newborne
WaterAid Malawi: Steven Sugden, Linda Milazi, Boyce Nyirenda, James Longwe,Nelly Magelegele and Researchers: Nebert Nyirenda, Amalia Sevilla and Olivier Stoupy
Researchers Uganda: Juliet Kanyesigye, Joseph Anguria, Edison Niwagaba andTim Williamson
WaterAid Zambia: Eric Gutierrez, Saviour Mwambwa and Researcher: Webby Wake
WaterAid Tanzania: Dominick de Waal and Rehema Tukai
All rights are reserved. Readers may quote or reproduce from this publication, but as copyright holder, ODI andWaterAid request due acknowledgement
Front cover photograph: Community well, Uganda (ODI/Marialivia Iotti)Back cover photograph: Children collecting water, Uganda (WaterAid/Geoff Crawford)
This report is a synthesis of findings from a research project entitled ‘WatSan & PRSPs: strengthening design, finance anddelivery of water supply and sanitation (WSS) programmes under PRSPs in sub-Saharan Africa’ funded by the UK Department forInternational Development (DFID) - project reference ENC0103 0671/0675/023A.
The project was coordinated by ODI Water Policy Programme in partnership with WaterAid and in collaboration with theWater and Sanitation Programme-Africa (WSP-Africa), as well as other researchers and policy specialists in the region. Theauthors are indebted to the rest of the project team upon whose work this report is based. Responsibility for the interpretationspresented here is that of the authors alone. We are also indebted to DFID for funding for the project although viewsexpressed here are not necessarily shared by DFID.
Table of contentsAcronyms and Abbreviations
I. IntroductionBackground to the researchStatus of WSS under PRSPs in sub-Saharan AfricaFactors behind poor integrationScope of the studies
II. Key themes and issues emergingPRSP preparation and implementationSources of funding to the sectorNational budgeting and processes of sectoral resource allocationPlanningEquitySustainabilityAccountabilitySanitationMonitoring and evaluationPerformance benchmarking
III. Zambia case studyIntroductionBudgetary problemsProgress in budgetary reformsFactors affecting sectoral investment under PRSPs
IV. Malawi case studyIntroductionWater sector under the MPRSPOff-budget and off-plan workingManagement of the water sector within districtsEquitySustainability
V. Uganda case studyIntroductionProgress in budgetary reformsProgress in implementationData problemsTowards better planning and budgeting for WSS interventions within districtsTowards coordinated and focused monitoring and evaluationDistrict performance assessmentNational performance measurement framework
VI. Conclusions and recommendationsFactors behind poor integration of WSS within PRSPStrengthening sectoral processJoint sector reviewsCoordinated and predictable donor supportImproving performance through monitoring equity and sustainability
Core principles of Poverty Reduction StrategiesWatSan in PRSPs: Key observations and opportunities from initial studiesEquity assessment toolSustainability snapshot: themes and rating systemSustainability snapshotRural local government levelsCalculating water point density per 1000 populationAn indicator for equity?Planning & budgeting processPossible indicators in a sanitation snapshotPossible tools for monitoring and measuring equity, sustainability and sanitation performanceThe political cost of failing to perform in Mubende district
Scope of WatSan & PRSPs research by WaterAid/ODIFrom PRSP priorities to resultsShare of revenueNew water points installed between 1998 & 2002Distribution of water points in Mwanza and PembaICWP densities at enumeration level in Mulanje south districtPercentage of functional handpumps dependent on installation date and maintenanceGOU W&S budget allocation, excluding donor projectsWater Point Density in sub-counties in Tororo DistrictICWP Density in Subcounties in Wakiso DistrictWPD by Parish in Kasanje, a poorly served subcounty in WakisoWPD by Parish in Masulita, a well served subcounty in WakisoWDP for successive years in Kwapa subcountyWDP in successive years in Nawanjofu subcountyImproved Community Points Density Mapping – Salima District
Sectoral share of PRSP budget, 2002/04Comparison of resource envelope against expendituresTotal authorised provision vs actual spending (Zambia national budget, 1999-2003)Levels of capital investment: authorised provision vs actual spending (MEWD, 1999-2001)Authorised provision vs actual spending of provincial departments of Water Affairs, 1999-2001Comparison of WSS capital expenditure allocations between four key spending bodiesHow WSS budget is allocated in the PRSPComparison of allocations under the PRSP and the budgetComparison of D-WASHE allocations in the annual budgets, TNDP, and PIPOriginal MPRSP costingsPillar 1 – Sustainable Pro-Poor GrowthShare of HIPC to total revenuePro-Poor Expenditure budget allocationActual PPE expenditure in 2002/03 against originally budgeted amountsComparison with other sectors for disbursement of PPE funds to date in 2003/04Actual PPE in 2002/03 against budgeted amounts for PPE in 2003/04Changes in PPE budgets for two years between 2001/02 and 2003/04Summary of recurrent and development expenditure (2003/04)The MoWD development budget by providerThe MoWD development budget by degree of controlThe MoWD development budget by useProviders of water points between 1998 and 2002 in area surveyedNumber of new water point installed via MoWD budgetNumber of new water point installed by NSAsSources of funding for 1,218 water points installed in Salima and Machinga (1998 -2002)Functionality of water points by technology at district levelInconsistent data in local governments and DWDIncreasing inequity the deeper you digWater Point Density in Kasanje subcountyDeveloping a district balanced scorecard in the WSS
African Development BankCommunity-Based ManagementCanadian International Development AgencyDistrict AssembliesDistrict Development FundDistrict Development PlansDepartment for International DevelopmentDistrict Water, Sanitation, and Health EducationDepartment of Water AffairsDirectorate of Water DevelopmentEnumeration AreaEuropean UnionFinancial yearGross Domestic ProductGravity-fed SchemeGeographic Information SystemsGlobal Positioning SystemHighly Indebted Poor CountriesHeadquartersImproved Community Water PointInternational Monetary FundJapan International Cooperation AgencyKreditanstalt für WiederaufbauLocal governmentLocal Governments Budget Framework PaperLocal Government Development ProgrammeMinistry of Agriculture and CooperativesMalawi Social Action FundMillennium Development GoalMonitoring and evaluationMinistry of Energy and Water DevelopmentMinistry of Finance and National PlanningMalawi KwachaMinistry of Local Government and HousingMinistry of FinanaceMinistry of Finance and Economic DevelopmentMinistry of Water DevelopmentMinistry of Energy and Water DevelopmentMalawi the PRSPmedium-term expenditure frameworkMinistry of Water, Lands and EnvironmentNon Governmental OrganisationNorwegian Agency for Development AidNational Roads AuthorityNon-State ActorsNational Water Supply and Sanitation CouncilNational Water Development ProjectOverseas Development InstituteOperation and MaintenanceOffice of the PresidentPoverty Action FundPoverty Eradication Action PlanParish Development CommitteePublic Investment ProgrammeParticipatory Poverty AssessmentPro-Poor ExpenditurePoverty Reduction and Growth FacilityPoverty Reduction Strategy PaperSave the Children
vi
SSASWAPTATNDPUNDPUNICEFWPWPDWSPWSSWSSDWWF
Sub-Saharan AfricaSector Wide ApproachesTraditional AuthorityTransitional National Development PlanUnited Nations Development ProgrammeUnited Nations Children’s FundWater PointWater Point DensityWater and Sanitation ProgramWater Supply and SanitationWorld Summit on Sustainable DevelopmentWorld Wide Fund for Nature
1
Introduction
I. Introduction
Background to the researchThis working paper presents a synthesis of findings from aresearch project entitled ‘WatSan & PRSPs: strengtheningthe design, finance and delivery of water supply andsanitation programmes under PRSPs in sub-SaharanAfrica’. The project was coordinated by the OverseasDevelopment Institute (ODI), in partnership withWaterAid and in collaboration with the Water andSanitation Program (WSP)-Afr ica, as well as withresearchers and policy specialists in the region.
The research detailed here addresses both policy andpractice issues surrounding the implementation of PovertyReduction Strategy Papers (PRSPs) from a water sectorperspective. The importance of access to improved waterand sanitation services for poverty reduction is wellestablished and recently enshrined in the MillenniumDevelopment Goals (MDGs). The water sector thereforeprovides a good example and test of PRSP delivery anddevelopment at national level, towards international goals.
The emergence of PRSPs in the late 1990s reflects agrowing consensus on the importance of poverty reductionas a central objective of debt relief, government expenditureand donor support, and the PRSP framework hassubsequently become a centre piece for policy dialogue inall countries receiving concessional lending from the WorldBank and IMF. A number of core principles of PRSPs canbe identified (see Box 1).
These reflect a growing recognition that developmentstrategies imposed from the outside are generally noteffectively implemented. They also represent part of abroader shift in aid provision away from individual projectstowards more programmatic forms of aid, and away frompolicy conditionality towards a more partnership-basedapproach.
Box 1: Core principles of Poverty ReductionStrategies
Source: Evans, A. (2002)
• country-driven: drawn up on the basis of broadconsultation and participation;
• results-oriented: each PRSP is to include performancetargets and a system for monitoring, to link pro-poorpolicy and results;
• comprehensive: since poverty is a multi-dimensionalphenomenon, each PRSP is to set out a statement of thefull range of interventions necessary to addresspoverty;
• prioritised: choices of the best anti-poverty interventionsare costed in relation to the available public revenues,as managed through the national budget;
• partnership-oriented, with collaboration between stateand non-state actors, and between donors andgovernments;
• based on a long-term perspective: PRSPs are pluri-annual, with also production of second andsubsequent PRSPs in a cycle (see further below in thissection II).
Poverty Reduction Strategies are intended to be:
This growing disillusionment with project aid andpolicy-based lending in the 1980s and 1990s led toproposals for a new type of budget support. Central tothese partnerships is the idea of mutual obligations througha structure of mutual accountability. PRSP frameworksare intended to form a basis for implementation ofnationally owned strategies, to build accountable andcapable government and ensure effective poverty reduction(ODI, 2002). PRSPs and associated processes are becomingincreasingly important in framing the policy agenda withindeveloping countries and focusing it on poverty reduction.As such PRSPs in many ways form the ‘keystone of a newaid modality’ (DFID, 2002).
Following the drafting of interim and full PRSPs incountries across sub-Saharan Africa, the spotlight of interesthas shifted to PRSP implementation. Successfulimplementation of Water Supply and Sanitation (WSS)elements in PRSPs will be important for translating theMDGs into action at country level, specifically those goalsrelating to water supply and sanitation which wereconsolidated at the World Summit on SustainableDevelopment (WSSD) in 2002.
The purpose of this project has been to review progressof PRSP implementation in terms of WSS objectivesthrough in-depth case studies in three countries (Zambia,Malawi and Uganda). These studies have examined theextent to which promises made in PRSPs are beingdelivered: specifically whether resources allocated in PRSPsto WSS priorities, are converting into expenditure on WSS-related outputs, and thereby contributing to povertyreduction objectives.
Status of WSS under PRSPs in sub-SaharanAfricaThe research reported here built upon earlier reviews ofthe status of WSS under PRSPs in sub-Saharan Africa(SSA). The Water and Sanitation Program – Africaconducted an initial desk review of WSS in PRSP initiativesin SSA in 2001. Further detailed studies were subsequentlyconducted in five African countries: Zambia, Malawi,Uganda, Madagascar and Kenya (ODI/WaterAid 2002) anda regional workshop was convened in Nairobi in 2002 tocompare country experiences and identify common issuesemerging (see Box 2).
The scoping studies and workshop concluded that WSShad been inadequately reflected both in terms of the processof PRSP preparation and the content of emerging PRSPsin SSA, with the exception of Uganda where the level ofpriority afforded to WSS in the PRSP was significantlyhigher (ODI, 2002). Whilst engagement with theseprocesses is increasingly important for stakeholders indifferent sectors, it seems water sector stakeholders havefailed to mobilise and engage with PRSP processes aseffectively as their counterparts in other sectors.
2
Implementation of Water Supply & Sanitation programmes under PRSPs
Factors behind poor integrationThere are a number of factors behind the poor integrationof WSS in PRSPs. Experience suggests that successfulprocesses rely on a combination of institutional capacityto develop and implement policy, and political commitmentto poverty reduction (Booth, 2003). Donors can play animportant role in building institutional capacity, andimproving the incentives for government actors to engageby providing support through budget systems but ultimatelyif the political will is lacking then the resulting strategiesare unlikely to be effective. PRSPs essentially provide amechanism whereby donors and civil society can work toensure poverty reduction remains on the agenda. As suchit is increasingly recognised that political processes are justas important as technical aspects (Piron & Evans, 2004).
A major premise in the PRSP ‘experiment’ is thatparticipatory processes consolidate and even generatepolitical commitment. While experience in Ugandasuggests that civil society lobbying can be instrumental inbuilding political preference for WSS (WaterAid Uganda,2003), a complex range of factors affect the emergence ofopportunities for influence (Williamson, 2003a). WhilePRSP processes provide opportunities for political dialoguethey are ultimately technical planning instruments andcannot by themselves address deeper issues of political andsocial change (Piron & Norton, 2004).
It is interesting to contrast the generally poor integrationin practice of the WSS sector with other sectors, notablyeducation and health, which almost universally havesignificant mention in PRSP documentation and aresubsequently benefiting more in terms of resourceallocation in budget processes. Williamson (2003a) examinesthe integration of WSS under PRSPs in Uganda, Zambiaand Malawi and compare it to generic experiences in the
Lack of Integration of WLack of Integration of WLack of Integration of WLack of Integration of WLack of Integration of Water Supply and Sanitation (WSS) with Poverater Supply and Sanitation (WSS) with Poverater Supply and Sanitation (WSS) with Poverater Supply and Sanitation (WSS) with Poverater Supply and Sanitation (WSS) with Poverty Reduction Strategies (PRSs)ty Reduction Strategies (PRSs)ty Reduction Strategies (PRSs)ty Reduction Strategies (PRSs)ty Reduction Strategies (PRSs)• The WSS sector has not been effectively incorporated into most PRS processes in the countries in the region, with the
possible exception of Uganda;• A key reason for this exclusion is the weak poverthe weak poverthe weak poverthe weak poverthe weak poverty diagnosis with the sectorty diagnosis with the sectorty diagnosis with the sectorty diagnosis with the sectorty diagnosis with the sector. The workshop emphasised the need to
improve diagnostic procedures, and also explored new approaches, such as the concept of Sustainable Livelihoods;• Weak diagnosis also reflects the weak base of sector information, which focuses largely on physical facilities and tends to
neglect sector-level monitoring and evaluation (M&E) systems. Even in national poverty impact assessments water-relatedindicators are poorly defined, and sanitation is generally not covered at all;
• There is an absence, or weak representation, of water and sanitation in PRSPs compared to education, health andagriculture, due to the weak links between central ministries, local authorities and sector actors at the local level inthe weak links between central ministries, local authorities and sector actors at the local level inthe weak links between central ministries, local authorities and sector actors at the local level inthe weak links between central ministries, local authorities and sector actors at the local level inthe weak links between central ministries, local authorities and sector actors at the local level inplanning and priority setting;planning and priority setting;planning and priority setting;planning and priority setting;planning and priority setting;
• The WSS sector has not effectively used the programmatic and sector-wide approaches which would enable a betterintegration with PRSs and the medium-term expenditure framework (MTEF). This is partly due to the sector’s institutionalcomplexity and the fragmented nature of its financing arrangements.
OpporOpporOpporOpporOpportunities which PRSPs present for Improving Sustainable Wtunities which PRSPs present for Improving Sustainable Wtunities which PRSPs present for Improving Sustainable Wtunities which PRSPs present for Improving Sustainable Wtunities which PRSPs present for Improving Sustainable Water Supply and Serater Supply and Serater Supply and Serater Supply and Serater Supply and Service Coveragevice Coveragevice Coveragevice Coveragevice Coverage• The PRSP is not a one-time document, but is part of an ongoing and evolutionaran ongoing and evolutionaran ongoing and evolutionaran ongoing and evolutionaran ongoing and evolutionary processy processy processy processy process, by which a well-defined set of
procedures presents an opportunity for WSS to be integrated into PRSs and development programmes;• While seizing this opportunity, it is important to ensure that the PRSP and the linked HIPC resources are not seen as
substitutes for the implementation of the critical sector reforms which will enable investments to benefit the poor in asustainable manner;
• PRSPs have not triggered water sector reform, but have shown the need to convert broad reforms to coherent implementationimplementationimplementationimplementationimplementationstrategies which link water sector initiatives with planning and budget processesstrategies which link water sector initiatives with planning and budget processesstrategies which link water sector initiatives with planning and budget processesstrategies which link water sector initiatives with planning and budget processesstrategies which link water sector initiatives with planning and budget processes.
Box 2: WatSan in PRSPs: Key observations and opportunities from initial studies
education and health sectors, providing recommendationson how WSS actors can better align themselves with thebudgetary process. The national budget is the mechanismfor implementing PRSP objectives through provisions andactual expenditure under the annual budget (PRSPs areframed for 3-5 years).
WSS in Uganda, for example, achieved higher budgetarypriority than in Malawi and Zambia and health andeducation have tended to gain higher priority than WSSin all the study countries. It is possible to identify a numberof key factors underlying success and failure of WSS ascompared with health/education (Williamson, 2003a):• Progress in sector reforms – in general health and education
sectors are further ahead in the development of SWAPsthan WSS. Institutional fragmentation, which manifestsitself in plans produced by different parts or “sub-sectors” of the sector (rural WSS, urban WSS, waterresources management etc.) which are not planned incoordination and not mutually supporting, is an obstaclein the water sector. By avoiding some of the usualdivisions between sub-sectors, Uganda has made someprogress in moving towards a SWAP which has in turnmeant WSS has been better able to engage with PRSPprocesses.
• Political commitment to poverty reduction and budgetaryreforms – greater political commitment to povertyreduction in Uganda provided a greater incentive forindividual sectors to align themselves with the PRSP.Experience in Uganda suggests that civil societylobbying can contribute to ensuring WSS is a politicalpriority. In terms of political ‘clout’, health andeducation sectors are generally more powerful withlarger budgets and thus in a stronger position whenlobbying for budgetary resources. Whether they havemade a better case for investment in terms of povertyreduction, or it is simply taken for granted that such
3
Introduction
expenditure is poverty reducing, is not clear.• Greater on-budget funding – a greater proportion of
funding available for education and health comes fromgovernment revenue whereas WSS has historicallydepended heavily on project funding, i.e. external tonational revenues. In Uganda there has been a shift infunding available for WSS from projects towardsprogramme-based support. This has resulted in strongerlinks to the Ministry of Finance (MoF) and greaterincentive to engage in the budgetary process. In contrastthe WSS sector in Zambia and Malawi remains heavilydependent on multiple fragmented donor projects.
If the PRSP and budgetary process can demonstrate thatstrategic, pro-poor interventions will generate moreresources for the sector, then there is a greater incentivefor sector stakeholders to engage. Donor behaviour is thuskey – varying levels of commitment to budget supportamong different donors send mixed messages to watersector stakeholders.
Despite heralding the Ugandan example in terms ofalignment and prioritisation of WSS within the PRSPand budget, it should be noted that sectoral reforms inUganda have yet to yield substantial improvements inefficiency and effectiveness. Case study analysis reveals,for example, that funds are not yet reaching the mostdeserving communities (see later). Neverthelessestablishing a single overarching framework for all sectoractivities is an important first step. There is a growingrealisation in the water sector that PRSPs are here tostay, cannot be ignored and that constructive engagementis therefore essential.
Engagement in the PRSP process is not however a pre-requisite for tackling inefficiency and ineffectiveness inthe water sector; this should be pursued in its own right.However if these issues are addressed then the sector islikely to be able to make a stronger case for funding under
Figure 1: Scope of WatSan & PRSPs research by WaterAid/ODI
PRSP Priorities for WSSPRSP Priorities for WSSPRSP Priorities for WSSPRSP Priorities for WSSPRSP Priorities for WSS
Budget AnalysisZambia andTanzania
Targeting of investmentand subsidies Tanzania
Equity of water pointdistribution Malawiand Uganda
the PRSP. While sector alignment is more difficult in theabsence of a strong PRSP process, as is the case in Malawiand Zambia, it is not impossible. Even in the context of aweak PRSP, a coherent sector programme is arguably betterthan the usual status quo of fragmented, inefficient donorprojects (Williamson, 2003a).
Scope of the studiesThe second phase of research has looked at progress to-date in PRSP implementation in three countries – Zambia,Malawi and Uganda. WaterAid also carr ied out anadditional study in Tanzania, and other linked studies havebeen commissioned by WSP-Africa in the region.
Each case study focused on different aspects of PRSPimplementation but together they cover the full spectrumof issues tracing the process from identification of prioritiesthrough to poverty reduction outcomes (see Figure 1).
The case studies summarised here sought to examinekey stages along the journey from Water Supply andSanitation ‘allocations’ in PRSPs to budget allocations andactual expenditure on sector outputs and in so doinghighlight the kind of events or factors which may disruptand delay the flow of funds and their translation into desiredpoverty reduction outcomes.
Alongside the key issue of how promises in PRSPs areconverting into expenditure on WSS-related outputs andthereby contributing to poverty reduction objectives –including amongst the poorest countries – three questionswere included in the research:
1. Are WSS sectoral plans/processes converging with PRSplans/processes and is resourcing of PRSs convergingwith national budgeting processes?
2. How may capacity for action by central anddecentralised government be strengthened to deliveron promises made in PRSPs, and how can external
Composition ofallocation
Composition of spend
Services and facilities provided
4
Implementation of Water Supply & Sanitation programmes under PRSPs
donors support this effort?3. From a donor perspective, how can budgetary support
to PRS be combined with contribution to waterpolicies and achievement of water-related povertyreduction outcomes on the ground?
ZambiaThe study in Zambia focused on the process by whichresources are allocated to WSS priorities under theZambian PRSP. It examines the composition ofallocations under different budget-lines and comparesthese with actual expenditure. The study highlights anumber of issues and problems surrounding decision-making, allocation and release of budgetary funds,resulting in underutilisation and poor targeting of availableresources. It identifies a number of bottlenecks relatingto the highly centralised nature of resource allocationprocesses and notes that these problems are furthercompounded by the existence of multiple donor projectswhich are poorly integrated with national and districtlevel plans. The study goes on to examine possible meansof improving efficiency and effectiveness throughdecentralised planning and resource allocation.
MalawiThe study in Malawi examined the flow of sectoralresources from national to district level and from districtauthorities to beneficiary communities. It focused inparticular on factors affecting the equity of water pointdistribution; constraints to sustainability; and the linkagesbetween access to water and poverty. The study shows thatthe process of budgetary reform still has some way to goin Malawi and highlights shortcomings in sectoral planningand budgeting processes – which remain poorly defined –partly as a result a large proportion of funds flowing intothe sector remain off-budget and beyond the control ofgovernment. In theory Malawi is currently undergoing aprocess of administrative and fiscal decentralisation but inpractice the Ministry of Water Development (MoWD)remains highly centralised. District authorities have littlecontrol over the allocation of sectoral resources and analysisof the location of existing water points which shows glaringgaps in the map and raises serious concerns as to the equityand sustainability of current investments. The studyproposes a number of ways in which these issues can startto be addressed.
UgandaThe Uganda study examines whether PRSP and nationalsector goals are actually being achieved on the ground byUgandan local governments in rural areas, and howimprovements in planning, monitoring and evaluation inthose local governments could potentially improve theefficiency and effectiveness of service delivery. Despitethe sectors slogan of “some for all, not all for some” the studypresents strong evidence that water services are beingdelivered increasingly inequitably, whilst sanitation andsustainability remain secondary concerns. It goes on toexamine the underlying factors and why WSS planningand M&E systems contribute to this breakdown betweenpolicy objectives, planning and implementation.Ph
The country studies show that WSS related objectives andcostings in PRSPs are in many instances being lost, ordisregarded in the process of national budget preparationand execution. ‘Activation’ of the pluri-annual promisesin PRSPs ultimately depends upon satisfactory operationof the annual budgeting cycle. If existing service gaps areto be addressed and progress towards MDGs enhanced,there needs to be improved coordination within the sectorand the final destination of sectoral resources needs to becarefully identified.
Uganda, with its Joint Sector Review process supportedby donors, and its relative advances in development ofsystems for sector performance assessment and performancebenchmarking of local governments, points to possible waysforward.
PRSP preparation and implementationAs noted above Uganda has progressed much further interms of budgetary and sectoral reforms than Malawi orZambia. The strength of the overall PRSP process and thecredibility of the Ministry of Finance, as the driver of theprocess, are key factors affecting the opportunities andincentives for sectors to align themselves with the PRSP.A credible PRSP framework includes an institutionalenvironment which provides incentives for alignment, clearroles and responsibilities of different institutions in planning,budgeting, implementation and accountability and review,and also standards for donor behaviour (Williamson, 2003a).
Sources of funding to the sectorAn important function of PRSPs is budgetary. A key role,together with the medium-term expenditure framework(MTEF) is as a platform for financing. Anti-poverty policiesand programmes set out in PRSPs are typically organisedby sector or theme, itemised into a list of actions and thencosted for funding. PRSPs are intended to attract fundsfrom national government and external donors.International donors are encouraged to apply their fundsby means of programmatic funding, instead of grants toindividual projects, thereby supporting a sector wideapproach. WSS interventions are more likely to receivefunding if water supply-related needs are clearly articulatedand translated into programmes of action. Poor integrationof water and sanitation objectives within the PRSPdocument would be expected to have lasting negativeconsequences for mobilising financing for the water sector.
However research revealed that sources of funding tothe water and sanitation sector are typically diverse andpoorly coordinated and flows into the sector are oftenextremely difficult to quantify precisely. The sector remainsdonor driven and dominated by major projects which aremostly funded externally. Many of these projects remain‘off-budget’ or ‘off-plan’, or both, reflecting a continuedlack of donor agreement on policy and funding prioritiesboth between and within sectors. Consequently nationalgovernments struggle to control and account for funds
flowing into the sector. The insistence of bilaterals, multi-laterals and NGOs on negotiating separate arrangementswith governments, outside of the budgetary processundermines the potential benefits of sectoral and budgetaryreforms and causes fragmentation of national budgets e.g.Tanzania, making it more difficult to prioritise sectorspending in terms of wider poverty reduction objectives.It also contributes to a widespread perception that thewater sector is ‘awash’ with funds.
National budgeting and processes of sectoralresource allocationResearch shows that administration of national budgetaryprocesses (see Figure 2) is generally weak. The studies ofWSS resource flows illustrate the mechanics of resource‘leakages’. In Tanzania for example actual expenditure onWSS over the past three years has been lower thanprojected. In Zambia the highly centralised nature ofresource allocation processes causes a number of‘bottlenecks’ which have resulted in underutilisation andpoor targeting of available resources. National allocationsin recent annual budgets have been greater than theresource envelope referred to in the PRSP, but actual
MTEFMTEFMTEFMTEFMTEFMedium Term RevenueProjections & SectorExpenditure Ceilings &Budget Strategies
BudgetBudgetBudgetBudgetBudgetAnnual Revenues,Agency & LG Allocations& Plans
Line Agency LocalGovernment
Note: arrows indicate intended sequence of decisions and flow. The dottedarrows from ‘Donor Revenue’ point to the funding options open to donors.‘On budget’ aid can be either project or budget support, whilst ‘off budget’aid takes the form of projects, and is usually channelled directly to spendingagencies.
Policy & Strategy
MT Planning &
Budget Formulation
Implem
entation
“Off Budget”
Projects
Donor
Revenue
“On Budget”
Budget Support
Sector Programmes
NGO
6
Implementation of Water Supply & Sanitation programmes under PRSPs
allocations do not match the budget, and funds releasedby the MoF to line agencies are often substantially delayed(even where they have been approved). Furthermore, localauthorities have little say in decision-making on resourceallocation which remains largely centralised.
In Malawi also attempts at budgetary reform have hadlimited success. The allocation of WSS in the nationalbudget is significantly less than the amount for WSS referredto in the PRSP, and the budgeting and allocation processfor the WSS sector is poorly defined. The position of localgovernment in the process of allocation and release of WSSfunds seems to be especially weak. Whilst in administrativeterms decentralisation may have to some extent taken place,in fiscal (and political) terms it is undeveloped. Insteadline ministries competing for funds at the centre arechannelling them to their own district agencies.
In all cases allocations are predominantly for physicalinfrastructure (e.g. number of boreholes). Important issuessurround ensuring an appropriate balance betweenallocations for ‘hardware’ and provision for ‘software’ suchas sectoral consultations, systems for monitoring andevaluation (M&E) and building capacity for operation andmaintenance (O&M). In both Zambia and Malawirecurrent costs tend to be drawn from government revenuewhereas development costs depend heavily on donorrevenue. This separation, combined with the current donoremphasis on MDG targets, means that budgetary processeshave become increasingly skewed towards capitaldevelopment. Sanitation is also consistently underprovided(see later). This is due partly to the fact that provision ofeffective sanitation services requires coordinated activitiesacross a number of different sectors (health, education,water).
PlanningResponsibility for lack of coordinated planning does notjust rest with national government. In Malawi the existenceof multiple projects with different objectives and planninghorizons makes it very difficult to develop a coherentsector-wide programme. It arguably weakens the MoWD’sability for planning and implementation and paradoxicallydiscourages donors from engaging in programmaticsupport.
Analysis of case study districts in Malawi highlights theabsence of coordination around district-level plans andreveals that NGO providers and donor projects often failto share information with national and local authoritieson the criteria used to prioritise and target investments.Despite recent efforts to improve coordination among thenumerous different agencies operating in the sector,adherence to official government policy, for example onestablished principles of user financing and management,remains weak.
EquityIn the study countries, funds available for spending onwater-related PRSP objectives are presently not well-targeted to the poor. The study in Tanzania shows a markeddisparity in terms of improved water supply coveragebetween regions, e.g. Lindi (11%) and Kilimanjaro (74%)
(de Waal, 2004). In Uganda, safe water coverage data iscurrently only calculated at the district level. This disguisessignificant sub-district variations (at sub-county, parish andvillage level). Using aggregate coverage figures as a proxyfor equity is deceptive insofar as it does not guard againstinequity in the location of water points within localities.The total number of people being served by new waterpoints is frequently emphasized in government reports,but research shows that while reported percentage coveragemay be high, inequity within each geographical area isoften pronounced.
The Uganda study also shows that certain areas continueto benefit from investment in new water points year afteryear, while neighbouring areas are persistently neglected.The level of inequity in distribution within districts actuallyappears to be increasing rather than decreasing and someareas have, over a long period of time, cumulatively receivedwater facilities more than others, without this informationbeing picked up by the district, or indeed the centre. InMalawi, the study shows that there are currently nosystematic sector criteria to guide WSS investments. Thesector is fragmented, with planning conducted on a project-by-project basis. In these circumstances, it is very difficultto develop a coherent sector plan.
Many of these problems can be overcome by relativelystraightforward improvements to planning and M&E, inlocal government. A simple technique developed byWaterAid can be used for comparing the equity of serviceprovision at those different levels (see Box 3).
Box 3: Equity assessment toolPilot work carried out by WaterAid Malawi, and tested alsoin Uganda and Zambia, has illustrated the utility of “water-point density mapping” to show the distribution of resourceswithin districts and sub-districts. This mapping uses an equityof distribution indicator which measures the variation in thedistribution of resources for water development in a givenarea, as a way of measuring the difference in access to waterservices between poor and non-poor areas (e.g. in a wholedistrict). Steps for use of the equity of distribution indicator are:(i) a water point survey using a global positioning system(GPS) noting location and condition of improved communitywater points, including information on age and provider; (ii)translation of GPS data into maps; (iii) calculation of “improvedcommunity water point density” (ICWP), based on censusstatistics of population distribution in area; (iv) GPS/GISdatabase created on number & location of water points; (v)calculation of average density in the area, deviation frommean – and equity of distribution.
In Malawi, application of this tool shows that insufficientattention is being paid to the location of new water points.New boreholes (on which the WSS section of the PRSPplaces great emphasis) are being drilled alongside existingones, and villages in the rural ‘hinterland’, which currentlyhave no improved water access, are not benefiting.
The Zambia study shows that WSS financial resourcescan easily be redirected away from unserved into better-served areas. Urban centres are receiving the lion’s share ofallocations, sometimes for projects which were not
7
Key themes
identified as PRSP priorities.
SustainabilitySustainability of facilities is also reported to be a majorconcern in the study countr ies. Performance onsustainability is often gauged by looking at the numberand proportion of functioning and non-functioningfacilities. Functionality is, however, only one dimension ofsustainability. Keeping a water point operational for a longperiod requires attention to a range of managerial, social,financial, institutional – and technical issues. For example,the preparedness of water user committees in rural areasto contribute to the establishment, management andmaintenance of water points, are elements which are keyto sustainability.
Another technique developed by WaterAid is the‘Sustainability Snapshot’ for rapid assessment of thelikelihood that a water supply system will remainfunctioning in the future, by looking at the combinationof relevant factors (‘snapshot’ since sustainability is adynamic process). It can be used in relation to existingwater installations or to assess communities’ preparednessbefore they receive a new water point. Stakeholders ateach level are asked to rate their confidence in performanceaccording to three thematic areas (finance, technical skills,spare parts & equipment). They choose from sets of threestatements in relation to each area (see Box 4) and a markof 1-3 is awarded, depending on which statement is selected,to arrive at a ‘score’ for sustainability.
Detailed analysis of sustainability issues in Malawi revealsinconsistent formation, training, support and developmentof water point committees. Many committees were foundto be ill-equipped for the tasks they were assigned: i.e.operation and maintenance of water points and collectionand deployment of maintenance funds. Important issuessurround the legal status and authority of water committees.These need to be addressed in order to formalise newroles and responsibilities and improve accountability inwater service delivery.
AccountabilityThe above tools should facilitate better decision-making,and enable different levels of local government to make
decisions which are in line with the achievement of sectorgoals. However, these systems will only be useful if theyresult in better decisions being made by managers andpoliticians, and if accountability to the end-user is improved.The Malawi study highlights the current lack ofgovernment capacity to regulate, monitor and supervisecontractors, resulting in sub-standard work and furthercontributing to inequitable access.
The well-served communities in the sample districts inUganda were found to have higher levels of educationwith influential people in the community – in that theyhad NGOs operating in their areas – were nearer toadministrative centres, had better road networks, and theirability to pay was greater than the least well-served areas.The latter were characterised by a comparatively poor roadnetwork, low levels of education, political marginalisation,weak leadership and lack of influential communitymembers.
The equity tool itself does not directly address thepolitical incentives which motivate powerful politiciansto ensure their voters are served. This can never be totallyovercome, but the incentives for districts to adhere to thenational policies, pr ior ities and guidelines can bestrengthened. To achieve this, structured performanceassessment and benchmarking of districts’ implementationof national water and sanitation plans is needed.
SanitationSanitation has for long been the poor cousin of water supply.The findings of the preliminary phase of this project werethat in all five African PRSPs reviewed, sanitation wasaccorded a significantly lower priority than water supply,despite even lower rates of access to sanitation and greatneed for increasing support to sanitation programmes.
The study commissioned by WSP-Africa in Uganda, forexample, notes that work is required to establish clearbudget mechanisms for sanitation which effectively allocateresources to all levels of government and thereby to end-users (Thompson, 2004). Responsibilities for sanitationactivities are divided amongst no less than five differentgovernment ministries at national level. Whilst this reflectsthe wide relevance of sanitation (e.g. to health, education,water, housing, environment) the study points to the needto clarify respective institutional roles. The diminishingcentral control over the allocation and use of resources hassimply conveyed the issue of prioritisation to another level;whilst districts now have more autonomy on how to spendmoney, sanitation is often not seen as a priority. Further,there are some indications that the move from projects toprogramme-based aid may have actually caused a reductionin sanitation financing (at least in the short term).
Monitoring and EvaluationEstablishment and strengthening of a system of M&E ofresource flows under PRSPs/budgets will be an essentialstep in arriving at allocation of funds which is moretransparent and less subject to political capture.
M&E should involve the establishment of objectives, andsystems for the measurement of performance against those
Box 4: Sustainability snapshotFinancialFinancialFinancialFinancialFinancial1 No funds available for maintenance when needed2 Fund available but not sufficient for the most expensive
maintenance3 Fund available and sufficient for the most expensive
maintenanceTTTTTechnical skillsechnical skillsechnical skillsechnical skillsechnical skills1 Technical skills not available for maintenance when
needed2 Some technical skills for maintenance, but not for all3 Technical skills for all maintenance processes availableEquipment and spare parEquipment and spare parEquipment and spare parEquipment and spare parEquipment and spare partststststs1 Not available when needed2 Available but not for all repairs3 Available for all repairs
8
Implementation of Water Supply & Sanitation programmes under PRSPs
objectives. Systems for measuring performance, especiallyat sub-national levels, were found to be weak in all thestudy countries. Local governments often lack the capacityand resources for even basic routine data collection. This isconsistent with the findings of recent ODI research onresults-orientated expenditure management (ODI, 2004b).Even more important is that the institutional mechanismsfor using that performance data are evaluated and decisionsmade to improve the application of resources to thoseobjectives, and for improved accountability and ultimatelyperformance. In the case of PRSPs, these include equityof resource use – as well as its sustainability – and value formoney. In implementation of M&E, identification of alimited and manageable number of indicators, understoodand recognised as key measures by which sub-nationalentities will report, will help to arrive at a robust system.
The water sector in Uganda is in the process ofdeveloping an overall framework for sector performancemeasurement. A first sector performance report using thenew framework was prepared for the Water Sector Reviewin October 2003. Ten different issues for performancemeasurement have been identified, and a small number ofassociated “golden” indicators is being developed. The ideaproposed is to produce national performance reportscovering one or two of these issues every year. Suchperformance measurement is key to ensure strong linkagebetween policy, planning and implementation.
Performance benchmarkingPerformance benchmarking systems are increasinglycommon in developed countries such as the US and theUK, as a means for encouraging local governments toadhere to sector policies and guidelines, and improve servicedelivery. Under the UK’s comprehensive performanceassessment process, league tables of council performanceare created, and councils are required to publish their ownperformance, and make it available to the public. If councilsperform poorly with respect to national goals and targets,then this is public knowledge.
Such practices are also evolving in Uganda. There is anannual performance assessment of local governments underthe Local Government Development Programme (LGDP),where the centre assesses and scores districts’ adherence todecentralisation laws, policies and guidelines. Under thesystem, districts assess sub-county performance and theseassessments are verified by central government. Districtsand sub-counties are required to fulfil various minimumconditions in order to access a discretionary localdevelopment grant. If a local government performs well,it receives an additional allocation.
An important goal in this context is incentivising districtsto adhere to national sectoral priorities and objectives, inaddition to governance criteria. Measuring distr ictperformance relative to the achievement of national WSSgoals, and publishing it nationally, could help buildincentives for politicians and administrations to adhere tosector policies, priorities and guidelines, includingconsideration of factors such as equity and sustainabilityin the planning process. Ultimately, incentives should begeared towards encouraging local governments to makedecisions based on evidence and which result in theefficient and effective delivery of sector goals.
9
Zambia case study
III. Zambia case study
IntroductionThe study in Zambia focused on the process by whichresources are allocated to WSS priorities under theZambian PRSP. It examined the composition of allocationsunder different budgets and compared these with actualexpenditure. The study highlights a number of issues andproblems surrounding decision-making, allocation andrelease of budgetary funds resulting in underutilisation andpoor targeting of available resources. It identifies a numberof bottlenecks relating to the highly centralised nature ofresource allocation processes and notes that these problemsare further compounded by the existence of multiple donorprojects which are poorly integrated with national anddistrict level plans. The study goes on to examine possiblemeans of improving efficiency and effectiveness throughdecentralised planning and resource allocation.
Budgetary problemsIt is often assumed that a lack of financial resources is theprincipal obstacle to providing water supply and sanitationto the world’s poorest people. However research in Zambiashowed that simply increasing funding is unlikely to benefitthe poor unless problems and inefficiencies in resourceallocation and targeting are first addressed. The PRSP isessentially a planning tool while the main mechanism forimplementing programmes remains the annual budget.Zambia’s annual budget processes could be described as‘labyrinthine’ and to get an overall picture of financial flowsentering the water sector detailed analysis of budgetaryprocesses is required. Using the national Financial Reportsof the Ministry of Finance and Economic Development(MOFED), the annual budget process was reviewed from1999-2003 to identify items relating to WSS (not recurringexpenses, but capital items including grants, loans and otherinvestments). This review has revealed a number of issuesand problems.
The study contrasted the theory and practice of annualbudget processes and highlighted a number of failures someof which have become institutionalised. In particular therehas been a persistent failure to plan and approve the budgetbefore the financial year begins in January. For several yearsnow Zambia has started its financial year without anyAppropriation Act, and hence has had to rely on presidentialprovisional warrants in order to authorise release of fundsfrom the treasury. Whereas an Appropriation Act outlinesa range of different items of expenditure for approval,presidential warrants are typically used for additionalmiscellaneous items. Although designed as an ad hoc,emergency solution to problems in budgeting they are usedroutinely in Zambia at the beginning of the financial yearas the budget is often not approved for several months.
Planning is also affected. The budget is not balanced byassessment of proposed government expenditure againstsavings and revenues and the President’s office tends touse previous year’s spending as the basis for releases of fundsunder the warrants. Items with an established ‘paper-trail’are easily approved, but it is much more difficult for MOFEDto obtain authorisation for new capital investments.
Severe bureaucratic bottlenecks have resulted asindividual funding releases must be approved on a case-by-case basis. This has led to a situation of ‘perennial crisisbudget management’ with major delays in the release offunds. These delays have a major impact on capitalexpenditures and affect both new and existing projects.Delays of several months are common during which timeproject machinery and personnel often lie idle, increasingproject costs and reducing overall efficiency. In the watersector timely release of funds is particularly important asborehole drilling activities are only possible in the dryseason. If funds are released late and the window ofopportunity (e.g. before the rainy season) is missed, projectsmay be delayed until the following year. Instead of returningunspent funds many sectoral agencies have found unofficialways of using them to ‘smooth’ cash flow problems leadingto reduced transparency and accountability in the watersector (e.g. opening side accounts to place funds out ofview).
It is important to note that in this situation Parliamenthas relatively little engagement in scrutinising the budget,or in ensuring that line items or allocations requested byspending bodies are justified and can be balanced againstgovernment savings and revenue. Instead expendituresapproved by the executive are simply rubber-stamped byParliament in the form of an ‘ex-post supplementary’estimate i.e. the approval comes only after the expenditurehas been made.
These generic problems with the budgetary process affectall sectors but the study suggests that they tend to be‘institutionalised’ in different ways in each. The success orotherwise of general budgetary reforms therefore dependspartly on the extent to which they are aligned with sectoralreform processes.
Progress in budgetary reformsAs well as the PRSP, important planning tools include theTransitional National Development Plan (TNDP) 2002-2005 and the Public Investment Programme (PIP). Themost recent tool developed is the Medium TermExpenditure Framework (MTEF) or ‘Green Paper’, whichis described by the Finance Ministry as a tool forbroadening consultations during budget preparation inorder to demystify the budget process and clar ifygovernment-proposed medium-term macroeconomicmeasures and fiscal frameworks for medium term plans.Under the MTEF, the various line ministries are requiredto set up Sectoral Advisory Groups, composed of differentstakeholders, to provide advice to each ministry on theirmedium term spending priorities.
These are positive developments but challenges remainin ensuring convergence of these various budget tools witheach other and linking them effectively with the annualbudget process. The PRSP, TNDP, MTEF and PIP aremulti-year instruments, while the budget follows an annualcycle.
10
Implementation of Water Supply & Sanitation programmes under PRSPs
As shown in Table 1 below, the PRSP sets out a totalbudget for poverty reduction objectives of US$1.2 billionover 3 financial years. This figure is technically an estimate,and not an authorised release, based upon the projected‘resource envelope’. To that extent, the PRSP is a ‘wishlist’ in terms of the amount of resources desired formobilisation towards poverty reduction. The Ministry ofFinance and National Planning (MFNP) is responsible forproducing these expenditure estimates based uponsubmissions made by all spending bodies attached to thenational government and taking into account projectedtax and non-tax revenues, grants and loans.
Table 2 shows that in recent years expenditure hasexceeded revenue leading to a budget deficit or ‘financinggap’. It is on this basis that the government of Zambia hasappealed for increases in external support. The PRSP isdesigned to strengthen the case for increased externalsupport by demonstrating that it will be directly used tofund poverty reduction programmes and not anything else.However in order to convince donors, government mustalso demonstrate fiscal discipline and improved transparencyand accountability in resource allocation.
RoadsHealthAgricultureEducationEnergyHIV/AIDSTourismWWWWWater and Sanitationater and Sanitationater and Sanitationater and Sanitationater and SanitationMacroeconomic ReformsGovernanceMiningTransportIndustrySocial Safety NetEnvironmentMonitoring/Eval. StatsGenderTTTTTotalotalotalotalotal
The Zambia PRSP allocates (in principle) some US$42.4million for investment in water and sanitation. A keyconsideration for donor agencies is how far resourcesallocated to WSS priorities in PRSPs, are actuallyconverting into expenditure on WSS-related outputs andthereby contributing to poverty reduction objectives?
• Firstly, is the ‘resource envelope’ sufficient in practiceto support projected WSS allocations?
• Secondly, is the WSS allocation in the PRSP reflected/translated faithfully into annual budgets?; and
• Thirdly, how far are sectoral allocations appropriatelytargeted?
Factors affecting sectoral investment underPRSPsThe Zambia case study involved a detailed analysis ofbudgetary allocations and expenditure on water andsanitation since 1999. The following six key findingsemerge, as summarised below:
1.WSS allocations are typically greater than theresource envelope defined in the PRSP
The PRSP is supposed to provide an overall frameworkfor all activities in the sector but WSS allocations inZambia’s annual ‘estimates of expenditures and revenues’are often greater than the resource envelope defined inthe PRSP.
The original PRSP allocated a total of US$42.4 millionover three years (2002/04) which equates to approximatelyKwacha 212 billion. However, as Table 3 shows, this figurehas already been exceeded in authorisations for the firsttwo years. The combined allocation for 2002 and 2003 isalready over K287 billion.
This shows that the budget ceilings laid out in the PRSPare being disregarded and raises questions as to the capacityof the Finance Ministry to exert control over differentgovernment spending bodies and effectively monitorexpenditure across different sectors.
Donor behaviour is a key factor in determining theavailability and flexibility of the resource envelope. Zambiais heavily dependent on external support which accountsfor about two-thirds of the US$1.2 billion budgeted inthe PRSP (PRSP, 2000: 120-126). However a largeproportion of donor money is already ‘earmarked’ forspecific sectors or projects making it difficult for thegovernment to ensure budgetary allocations match thoselaid out in the PRSP.
2. There is great variance between authorisedprovision and actual spending
The study also undertook detailed analysis of financialreports available for 1999-2001. Perhaps the most strikingfeature of these reports is the huge variation betweenauthorised provision and actual spending (see Table 3). In1999 the entire government machinery spent only 9.65%of the total approved and authorised by Parliament forwater and sanitation line items. In 2000 the figure rose to15.65% and in 2001 it climbed sharply (in relative terms)to 24.70% but even this figure remains extremely low.
11
Zambia case study
There are a number of possible explanations for thesefigures. At first glance one might assume there is insufficientcapacity to spend the money allocated. However interviewswith officials at national and local government level suggestthat existing capacity is in fact largely underutilised. Forexample a senior hydrologist at the department of WaterAffairs in Western Province pointed out that for 2003, theiroffice had requested approx K3 billion for water andsanitation investments but had received just K140 million(4.6%). He pointed out that about K50 million of this wasspent on the construction of two boreholes which werefinished in just five days. ‘We have the personnel and themachinery’, he said, ‘but they just lie idle for most of theyear’, paralysed by a lack of funds.
National level planners at the Ministry of Energy andWater Development (MEWD) who consolidate budgetsubmissions also expressed frustration over the disparitybetween authorised provision and actual releases, arguingthat the MEWD has the capacity to implement thebudgets/plans it submits to the Finance Ministry and doesnot understand why releases are not approved. This reflectsa lack of accountability in decision-making at this level.
It is interesting to note a nearly six-fold increase in capitalexpenditure on water and sanitation in 2001, compared to2000 and 1999 (Table 4). 2001 was an election year andthe bulk of the releases for that year occurred in Septemberprior to the election in December. Finance Ministryofficials suggest that this increased release of funds wasdue to the need to comply with donor conditionalities tomeet HIPC completion point but, while official figuresfor 2002 and 2003 are not yet available, research suggeststhat this level of expenditure has not been maintained.The disparity between authorised provision and actualexpenditure for water and sanitation means there is a certainelement of political discretion in actual funding releases.While overall levels of expenditure on WSS remain low,
Source: Consolidated from 1999, 2000 and 2001 Financial Reports and 2002, 2003 Estimates of Expenditures.
19992000200120022003
3,473,308,922 ($0.64 m)8,176,348,001 ($1.63 m)
48,787,082,154 ($9.76 m)No figures availableNo figures available
YYYYYearearearearear Authorised provision for waterAuthorised provision for waterAuthorised provision for waterAuthorised provision for waterAuthorised provision for waterand sanitationand sanitationand sanitationand sanitationand sanitation
Actual expenditure for waterctual expenditure for waterctual expenditure for waterctual expenditure for waterctual expenditure for waterand sanitationand sanitationand sanitationand sanitationand sanitation
Percentage of expenditure asPercentage of expenditure asPercentage of expenditure asPercentage of expenditure asPercentage of expenditure ascompared with the provisioncompared with the provisioncompared with the provisioncompared with the provisioncompared with the provision
9.65%15.65%24.70%
--
Table 3: Comparison of total authorised provision and actual spending on water and sanitation line items inthe Zambia national budget, 1999-2003 (in kwachas)
35,974,179,010 ($7.19 m)52,213,690,709 ($10.4 m)
197,587,575,884 (($39.5 m)135,696,095,997 ($27.1 m)152,107,960,726 ($30.4 m)
some agencies and certain districts were found to fare betterthan others.
Other line agencies have a much better track record interms of the ratio of actual expenditure to authorisedprovision, according to statements of persons interviewed,e.g. the Zambia Security Intelligence Service and otheragencies attached to the Office of the President. Goodrelations with the Finance Ministry & National Planningwas found to be an important factor in the context ofuncertain revenue flows. Another key factor is whetherthe allocation in question comes from central governmentrevenues or is dependent on external donor sources. Thisis particularly true of the water sector in Zambia which isdominated by donor projects. Levels of external fundingare difficult to predict and big donor projects are subjectto delays and even cancellation, especially if the recipientgovernment fails to meet the conditions of the grant orloan. Furthermore donor disbursements are not always‘captured’ in government accounting systems which makesit difficult to plan and monitor overall expenditure. Sectoralagencies funded from government revenues, e.g. provincialdepartments of water affairs, appear to achieve higher thanaverage levels of expenditure (see Table 5) although thismay partly reflect improved financial reporting.
3. Local authorities have little or no power indecision-making on resource allocation
Decentralisation, according to the PRSP, ‘will be developedand implemented as a matter of top priority’ in order to‘empower local government systems and traditionalauthorities to assume their rightful place in the country’spoverty reduction agenda.’ (PRSP, 2002: 36) A furtherPRSP objective is accountable and efficient publicexpenditure management; participatory budgeting; greatertransparency in public procurement and contractingprocedures; strengthening of government finance,accounting and internal audit systems; and improved
Source: Consolidated from 1999, 2000 and 2001 Financial Reports and 2002, 2003 Estimates of Expenditures.
Table 4: Levels of capital investment: Comparison of authorised provision and actual spending on WSScapital expenditures, Ministry of Energy and Water Development, 1999-2001 (in kwachas)
19992000200120022003
YYYYYearearearearear Authorised provision for waterAuthorised provision for waterAuthorised provision for waterAuthorised provision for waterAuthorised provision for waterand sanitationand sanitationand sanitationand sanitationand sanitation
Actual expenditure for waterctual expenditure for waterctual expenditure for waterctual expenditure for waterctual expenditure for waterand sanitationand sanitationand sanitationand sanitationand sanitation
Percentage of expenditure asPercentage of expenditure asPercentage of expenditure asPercentage of expenditure asPercentage of expenditure ascompared with the provisioncompared with the provisioncompared with the provisioncompared with the provisioncompared with the provision
11,196,941,273No figures availableNo figures available
1.96% 4.52%32.74%
--
12
Implementation of Water Supply & Sanitation programmes under PRSPs
199920002001
YYYYYearearearearear Authorised provision for waterAuthorised provision for waterAuthorised provision for waterAuthorised provision for waterAuthorised provision for waterand sanitationand sanitationand sanitationand sanitationand sanitation
Actual expenditure for waterctual expenditure for waterctual expenditure for waterctual expenditure for waterctual expenditure for waterand sanitationand sanitationand sanitationand sanitationand sanitation
Percentage of expenditure asPercentage of expenditure asPercentage of expenditure asPercentage of expenditure asPercentage of expenditure ascompared with the provisioncompared with the provisioncompared with the provisioncompared with the provisioncompared with the provision
1,642,325,010 ($0.328 m)1,403,634,499 ($0.280 m)
2,117,657,884 ($0.423 m)
1,073,743,419 ($0.214 m)1,076,252,150 ($0.215 m) 1,775,969,919 ($0.355 m)
65.38%76.68%83.86%
Table 5: Comparison of authorised provision and actual spending of provincial departments of wateraffairs, 1999-2001 (in kwachas)
financial management and expenditure tracking efforts(ibid.). These objectives are laudable, but in the case ofwater and sanitation, there appears little progress made inmeeting them, at least to-date.
For example, according to officials interviewed in Kaoma,Mongu, and Kafue Districts, neither the district nor theprovincial government have significant involvement indecision-making on financial allocations to water andsanitation projects. ‘We don’t even know the water projectsapproved for our district,’ said the District Administratorof Kaoma. ‘Sometimes, we learn about projects only whenwe hear the rigs starting to dig for boreholes.’
Officials in Kafue were similarly unable to answerquestions on levels of investment in water and sanitationwithin their jurisdiction because they have no direct rolesin decision-making and allocation. Officials in Kafuereceive a lump-sum from the MFNP, but they are not surehow what they have received is linked either to the budgetor the wider PRSP. None of the district governmentdepartments interviewed were found to have informationon the financial allocations made by governments to theirdistrict for operations. They further suggested that suchinformation is kept only at provincial and national level.
However, interviews with provincial government officialsrevealed limited involvement at this level too, ‘We onlyknow about the projects if the national line agencies copyus their reports. At the moment, that is very rare. In theorythe line agencies should report to us, but that has not reallyhappened’(interviews in Western Province).
Ultimately, it is the national line agencies that bid forbudgetary allocations, and hence receive those allocationsif these are approved. The provincial government is allowedsome capital expenditure (i.e. power to spend forinvestments such as boreholes), but this is severely limited.For example, Western Province prepared a budgetsubmission for 2004 requesting funding for a total of 28boreholes (4 for each of the seven districts), 21 wells, andthe rehabilitation of earth dams. Submissions are based onconsolidated submissions made by the District Councilsand are submitted by the Provinces to the MFNP. However,despite providing budget inputs, officials complain that theyreceive little or no feedback on the outcomes even onapproved items. This contributes to a situation describedas ‘one way planning’.
It is not only government allocations that are not knownat district or provincial level – information on donorallocations as well is at best sketchy. In Kafue and Kaoma,district departments were able to provide only minimal
information on the extent of donor involvement in WSSactivities in their districts.
The Zambia study highlighted a number of problemsrelating to the lack of genuine decentralisation. Mostmoney and projects are channelled through the lineministries, which would rather deal with their field offices,and not the local authorities. The big donors also operatethrough the central government bodies, and generally donot track whether resources are administered throughdistrict offices as intended. For example, contracts for theconstruction of water points are tendered and authorisedfrom central offices. Central government has introducedtender thresholds that disallow local government fromtaking charge of tendering for projects above a certainamount. At the same time the National Tender Board andministerial headquarters do not have the capacity to managethe demand for tenders creating a bottleneck. As a resultonly a limited number of contracts are actually awardedeach year, when in theory the funding already exists.
Thus, in tracking financial flows to water and sanitationin Zambia, the road leads back to Lusaka, and not to theoffices of the district or provincial governments. A reviewof the water and sanitation budget items in the annualbudget from 1999 to 2003 reveals that the two main lineagencies involved in water and sanitation provision are theMinistry of Energy and Water Development and theMinistry of Local Government and Housing (MLGH).Table 6 presents a comparison of capital expenditures forWSS allocated to these ministries, the Office of thePresident - Department of Water Affairs (OP-DWA) andMFNP. Spending is again seen to be low in comparisonwith the authorised allocation.
4. Lack of transparency and accountability inbudgetary processes, with the result that financialallocations are frequently reprioritised and re-directed away from unserved into better-servedareas
If allocations for poverty reduction objectives set out inthe PRSP are to be translated into budget allocations andactual spending, public expenditure management needsto be systematised and made more transparent, withopportunities for scrutiny by outside bodies. While thelegal framework is clear, Zambian budget processes remainunsystematic and untransparent.
For example, it is not clear whether the tools for theannual budget – i.e. the PRSP, the TNDP, the PIP and theMTEF – are actually being used as tools. For instance, inthe annual budget of 2001 (the annual budget passed afterthe PRSP was approved) there were sharp increases in
13
Zambia case study
capital expenditure provisions. MEWD got an additionalK17 billion for a Poverty Reduction Programme(boreholes, wells), which is more than half of the ministry’sapproved allocations for the year. The allocation for MLGHjumped from K30 billion in the previous year to K78 billionin 2001. Even the Ministry of Finance was identified inthe 2001 annual budget as a spending body, allocated withmore funds (K82 billion) than MLGH or MEWD. Thetwo biggest allocations for the funds allocated throughMFNP were K31.8 billion for Lusaka Water SupplyRehabilitation Project, and K31.2 billion for Kitwe WaterSupply Rehabilitation and K20.3 billion for theLivingstone Water Supply service, making a total of K83.3billion which is nearly twice as much as the 42.4 billion‘allocated’ in the PRSP.
It appears therefore that after the PRSP was approved,major new line items for WSS appeared in the annualbudget. These do not seem to conform to the projects andprogrammes identified in the original PRSP (see Table7). Research revealed that recent increases in allocationsfor WSS under the PRSP have been mainly diverted tothese new projects and programmes.
Much of the increases in authorised allocations from2000 came from donor grants. Table 8 shows that a largeproportion of this money earmarked for WSS actuallyfound its way to MFNP (K 24 billion approximately)which is identified as a spending body in annual budgets.The MFNP reported that these amounts were sent to theMEWD, MLGH or relevant agency, but it is not clearwhether it is the receiving agency or the Finance Ministrythat should principally account for them.
Table 6: Comparison of capital expenditure allocations between four key spending bodies onwater and sanitation (in million kwachas, rounded off)
Source: Consolidated from 1999, 2000 and 2001 Financial Reports and 2002, 2003 Estimates of Expenditures.
13,592615
30,3205,8561,4041,0765,416
620
Part of the problem is the lack of clarity in the definitionof functions between the MEWD and the MLGH. UnderZambia’s water policies, water resources management isthe responsibility of the MEWD, while water supply isthe responsibility of MLGH. However, there appears tobe some confusion in relation to these functions. Forexample, the budget item ‘Water Resources ActionProgram’ had an allocation under the MEWD, and alsoanother under the MFNP, but it proved difficult to get anofficial explanation as to the rationale behind this allocationof resources.
These findings reinforce the observations of Mwanawinaet al (2003) on the Zambian annual budget process. Giventhe budgeting by presidential warrant that has become soprevalent, and provisions under legislation (namely theFinance Control and Management Act), a great deal oflatitude and discretion on the budget lies with the FinanceMinistry and the President. This leads to weak legislativeoversight, and a general lack of legislative scrutiny of publicexpenditure management by the executive. The authorityof the Ministry of Finance is well-defined, but lacks checksand balances to ensure transparent as well as efficient budgetdecisions. The roles and the responsibilities of the legislatureare relatively well-assigned in principle, but Parliament doesnot engage critically with the budget in practice. The expost facto rubber-stamping of deviations from the budgetundermines the legitimacy of Parliament’s decision-makingand oversight role (Mwanawina, et al., ibid). These widerissues surely need to be resolved first, if Zambia is toimprove transparency and accountability in sectoralspending, including in the water sector.
Source: Govt. of Zambia, 2002, p. 169
Table 7: How WSS Budget is allocated in the PRSPPolicy Actions/ Action PlansPolicy Actions/ Action PlansPolicy Actions/ Action PlansPolicy Actions/ Action PlansPolicy Actions/ Action Plans Responsible Institution/sResponsible Institution/sResponsible Institution/sResponsible Institution/sResponsible Institution/s
Water Resources Action ProgrammeIntegrated Kafue Basin Water Resources Management ProgramDam Construction and Rehabilitation ProgramGround Water Exploration & Mapping ProgramCapacity Building ProgramComplementary D-WASHE Support ProgramRural Water Supply and Sanitation ProgramTTTTTotalotalotalotalotal
14
Implementation of Water Supply & Sanitation programmes under PRSPs
The lack of transparency and accountability leads to adifficult problem – that the underserved areas will loseout in allocations to areas that are already better served,including those areas which know how to assert or enjoybetter political access in the scramble for resources. Forexample, the three biggest allocations authorised in the2001 budget were for the three biggest cities of Zambia.As noted above, K83.3 billion was authorised for three bigurban water supply project in Kitwe, Lusak and Livingstone.These projects were not identified as priorities forexpenditure under the PRSP, yet they received large chunksof resources.
There are a total of seven WSS items given allocationsin the PRSP budget, as shown above in Table 7.Comparison of the PRSP budget to the annual budgetallocations in relation to these seven items reveals thefollowing, as set out on Table 8.
For one item Water Resources Action Program, therewas over-budgeting. For the five next items, there wasunder-budgeting. For one item (the biggest) the sumsallocated appear to be in line with the PRSP plan, butwhat stands out is that Central and Eastern Provincereceived a huge share of the allocations. In other words,the PRSP plan for WSS budgeting has not been followedin the Annual Budgets.
Overall, the allocations given to these seven PRSP-identified WSS items in the annual budgets are much lessin comparison to what was allocated to the Lusaka, Kitweand Livingstone water supply rehabilitation. In 2002 and2003, Livingstone WSS Services received another K19.18billion and K20.38 billion, respectively. Kitwe, got K8.52billion for 2002 and K14 billion for 2003. Lusaka gotanother K17.604 billion in 2002, and none for 2003.
Table 8: Comparison of allocations under the PRSP and the budget
Policy Actions/ Action
Plans
Water Resources Action
Programme
Integrated Kafue Basin
Water Resources
Management Program
Dam Construction and
Rehabilitation Program
Ground Water Exploration
& Mapping Program
Capacity Building Program
Complementary D-WASHE
Support Program
Rural Water Supply and
Sanitation Program
Responsible
Institution/s
MEWD
MEWD
MEWD, MAC
MEWD
MEWD, MLGH,
NWASCO
MEWD, NWASCO
MEWD, MLGH
Costs (in USD)
2002-2004
1,400,000
2,000,000
24,101,505
900,000
500,000
3,500,000
10,000,000
Actual Allocation
In 2002 and 2003 is K2.8 billion and K4.9 billion, respectively. The annual
budget has already over shot the PRSP budget on this item.
There is no similar item seen in the annual budgets. An allocation of K200
million in 2001, none in 2002, and K60 million in 2003 was given to the
closest item: a “Kafue Basin Action Plan”.
K7 billion in 2002, and K5 billion in 2003.
In the annual budgets of 2002 and 2003, this item was given a total
allocation of K840 million.
No similar item exists in the annual budgets, although there are some
capacity building items that are provided for under the budgets of the
Provincial Departments of Water Affairs.
In 2002 and 2003, the annual budgets provided an allocation of K1.57
billion and K1.7 billion, respectively, for the item “Support to D-WASHE
Program”. This is roughly the same amount provided in the previous years.
In 2000, the annual budget provided K1 billion for D-WASHE support, and
in 2001, K1.5 billion was provided.
For this, the biggest WSS budget item in the PRSP, the 2002 annual
budget provided for:
K4 billion to the MEWD for Rural Water (Boreholes and Wells)
K3.1 billion to the MLGH for the Central Province Rural Water Supply
Program
K6.086 billion to the MFNP for the Eastern Province Rural Water Supply
Program
K7.225 billion to the MFNP for the Central Province Rural Water Supply
Program
K1 billion to the MLGH for the Rural Water Programme.
The 2003 budget provided for:
K5.4 billion to the MEWD for Rural Water (boreholes and wells)
K1 billion to the MLGH for the National Water Rural Supply
K3.19 billion to the MLGH for the Central Province Rural Water Supply
Program
K10.64 billion to the MFNP for the Eastern Province Rural Water Supply
An important concern therefore is whether fundssupposedly earmarked for poverty reduction purposes arebeing re-directed elsewhere or are ‘captured’ for otherobjectives that are not necessarily directly pro-poor.Another question that can be raised is whether donorsthemselves are supporting PRSP objectives or whether inpractice they are in fact by-passing and undermining them.
5. Donor activities are poorly aligned with PRSPobjectives and poorly integrated with governmentbudgeting and accounting systems
As shown in Table 3, there were sharp increases in budgetallocation from 1999 to 2001. There was about a 45%increase in authorised provisions from 1999 to 2000, whilefrom 2000 to 2001, the increase was nearly 300%. Theincreases are attributable mainly to donor grants and loansthat started in 2000 and appear to have peaked in 2001.These represent only the on-budget donor contributions,and do not include off-budget spending on water andsanitation.
It appears that the low-level of actual expenditure (ascompared to authorised provisions) shown in the FinancialReports is attributable to two main factors. Firstly, donorsthemselves do not always deliver funds to match theirpledges for one reason or another. Secondly, they may havemade appropriate releases of funds, but these are notchannelled through the Government of Zambia’saccounting system and are therefore not captured in theFinancial Reports.
For the donor-funded items in the 2001 budget – totalactual expenditure is only 10.63% of authorised provision.It can also be seen that in the 2001 Zambian nationalbudget, the donor supported items make up 70.84% ofthe total authorised provisions for water and sanitation.There are some water specialists in Lusaka who estimatethat if off-budget donor provisions are included, total donorinvolvement in financing for WSS could reach up to 85%of total spending on the sector in Zambia.
A key issue that emerges is whether donor decisions onthe allocation of finances (for poverty reduction in generaland WSS in particular) are in fact supporting PRSPobjectives. As can be seen from Table 8, there seems to bepreference towards allocation of loans and grants to WSSprograms in the big cities and much less attention is givento the poorer rural areas. Donors naturally tend only toapprove loans for those areas that can repay, which are not
necessarily the poorest. Thus, while donors make generalcommitments of support to poverty reduction, it is notclear whether their individual decisions on the grants orrelease of finances to specific programs are still guided bythe same commitment. For example, the PRSP identifiedrural water supply as a priority for expenditure. But moremoney is actually being given to urban water supplyrehabilitation, and it is not clear whether this is targeted tothe poorest e.g. peri-urban areas.
The MFNP reports that the national governmentaccounting systems are not able to account for donorreleases for specific projects in its Financial Reports, becausefunds spent do not pass through the system. While releasesand expenditure may in this way be made more efficient,a consequence is that accountability of donors to nationalgovernment is limited. While donor support is often subjectto conditionalities, donors themselves appear to give lowpriority to reporting back to national government abouthow money is actually being spent.
6. The approach to WSS provision is largely an‘engineering solution’ with inadequate attentionsustainability and sanitation
The approach of both government and donors to WSSprovision in Zambia can be characterised as dominated bypursuit of an ‘engineering solution’, ie: measuring progressin WSS provision mainly in terms of constructedinfrastructure. For example, in rural water supply andsanitation, success is measured in terms of the number ofboreholes or latrines built. There are items in the annualbudget that are for ‘software’ components, but the majorspend is on physical infrastructure.
In financial year 2001, not only did on-budget donorcontributions peak, it was also the year when a new itementitled ‘Poverty Reduction Programme – Rural Water(Boreholes, Wells)’ was introduced in the budget of theMOWED1. However analysis of the supplementary budgetfor 2001 reveals that the K17,873404,000 allocated wasnot additional money for pro-poor infrastructure butactually came from reallocations within the existingMEWD budget i.e. by shaving off varying amounts from15 other items of approved expenditure. This could simplybe viewed as government reordering its priorities withinthe sector, but the result has been a greater bias towardsinfrastructure development.
The study found no clear links between investments ininfrastructure and expenditure on ‘software’ componentssuch as capacity building and training. The low levels ofadministrative capacity at district level mean that D-WASHE committees are frequently bypassed by contractorshired by central government. The problems of capacity atthe district level are evident:
• There was very little evidence of proper bookkeepingin the two districts studied for this research. Thequestions on financial flows for WSS remain largelyunanswered by district officials, including accountantsand treasurers. Offers by the researchers to look intothe books themselves were turned down. In WesternProvince, the researchers were shown the folder of
Implementation of Water Supply & Sanitation programmes under PRSPs
budget submissions from current and previous financialyears. There was no consolidated report; only rawunprocessed data was available at provincial level, hencethe difficulty of providing answers to the questions.
• Communication between the district authorities andcentral government water agencies is poor. Even at theprovincial level, communication is not effective. Localgovernments have little knowledge of what goes on inthe capital, particularly in the allocation of financialresources.
A key indicator of a commitment beyond theengineering solution is support for so-called ‘software’ (i.e.non-infrastructure) components of WSS provision. Onesuch software component is support to the D-WASHEcommittees, for which the PRSP ‘allocates’ US$ 3.5 millionor 8.25% of the total WSS funds. (PRSP, May 2002: 171).However there is conflicting information as to whichagency this amount will be channelled through:
• MEWD and NWASCO were identified as thespending bodies for this allocation in the PRSPdocument;
• MEWD, the MLGH, the Ministry of Health, andNWASCO which have been identified as theresponsible institutions in the TNDP;
• In the annual budgets, funding support to the D-WASHE programme is channelled through the MLGH.In 2000, an allocation of K1 billion was authorised,but none was actually spent. In election year 2001, K1.5
Table 9: Comparison of D-WASHE allocations in the annual budgets, TNDP, and PIP (in kwachas)
Allocations in the Annual EstimatesAllocations in the TNDPAllocations indicated in the PIPEstimated Annual Allocations in PRSP1
NB: The PRSP document gives an overall figure of $3,500,000 for the three year period 2002-2004. Translating this over an average of three years gives anannual figure of approx. $1,166,666(or K5,483,333,333 in Kwacha as indicated above) per year for the D-WASHE Programme
billion was allotted, and K 844.6 million was spent. In2002, K 1.569 billion was allocated, while in 2003, itwas K1.7 billion. These do not tally with the TNDP, norwith the Public Investment Programme (see Table 9)
Table 9 shows that the amount allotted for D-WASHEsupport in the TNDP was different from what actuallycame out in the Annual Estimates, and is different againfrom the amounts indicated in the Public InvestmentProgram. Thus, while Support to D-WASHE has been given8.5% of the PRSP allocations for WSS, it has actually beengiven less than this in the annual budgets. The differences infigures in the TNDP and PIP, as well the differences in theidentified spending bodies with responsibility for thesefunds, create more confusion. It thus becomes difficult toaccept that government and donors are serious in pushingthis software-related budget item to completion.
Finally, despite the repeated reference to ‘water andsanitation’ in the PRSP there is little actual provision forsanitation. One item in the PRSP, for which US$10 millionis provided, is the Rural Water Supply and SanitationProgramme (page 171). The objective is ‘to provide watersupply and sanitation to population in rural parts of thecountry.’ A review of the annual budget shows that this hasbeen interpreted only to mean the construction ofboreholes. There are no items in the annual budgetspecifically for the construction or delivery of sanitationfacilities in the rural areas, or for the construction of drainagetoilets or other sanitation facilities in peri-urban areas.
17
Malawi case study
IV. Malawi case study
IntroductionThe Malawi case study examines the flow of sectoralresources from national to district level and from districtauthorities to beneficiary communities. It focuses inparticular on factors affecting the equity of water pointdistribution, constraints to sustainability, and the linkagesbetween access to water and poverty – in selected ruralareas. The study shows that the process of budgetary reformstill has some considerable way to go in Malawi andhighlights shortcomings in sectoral planning and budgetingprocesses, which remain poorly defined. This is partlybecause a large proportion of funds flowing into the sectorremain off-budget and beyond the control of government.Malawi is in principle undergoing currently a process ofadministrative and fiscal decentralisation, but in practicethe MoWD remains highly centralised. District authoritieshave little control over the allocation of sectoral resources
and serious concerns surround the equity and sustainabilityof current investments. The study proposes a number ofways in which these issues can start to be addressed.
Water Sector Status under the MPRSPIn Malawi the PRSP (MPRSP) is designed to provide asingle and comprehensive national strategy for povertyreduction and forms the basic framework for all stakeholderactivities in the water sector. The MPRSP integrates thepr inciples of the MTEF to determine pro-poorexpenditure and established ceilings for governmentexpenditure over a three-year planning horizon. Table 10summarises indicative resource requirements for the majorMPRSP pillars. This forms the basis for annual resourceprogramming in the medium term i.e. until MPRSP targetsare reviewed after 2005.
Source of growth
Enabling environment
Rural feeder roads
Access to good WSSAccess to good WSSAccess to good WSSAccess to good WSSAccess to good WSS
Rural energy
Rural telecommunication
Total Pillar 1
Total MPRSP
Total governement expenditureRate of exchangeTotal governement expenditure
Implementation of Water Supply & Sanitation programmes under PRSPs
There is a budget line under pillar 1 ‘Sustainable pro-poor growth’ entitled ‘access to good WSS’. During thethree year planning period 20% of total governmentexpenditure (or 28% of pro-poor expenditure) will beallocated to the activities under pillar 1. Table 11 showsthe activity categories under this Pillar, including ‘Accessto good WSS’. The planned allocation for rural watersupply and sanitation services has been set at 12% or Pillar1 expenditure (approximately 3% of total governmentexpenditure).
The MPRSP states that provision of and equitable accessto potable water supplies and reasonable sanitation facilitiesare central to poverty reduction. Furthermore, improvedaccess to safe drinking water had been identified as a highpriority during district consultations on the content ofthe PRSP. However, despite its prominence in the PRSPdocument, it is apparent that the sector has not been giventhe same priority in subsequent budget allocations. A keyaim of this study therefore is to understand why, and toidentify ways in which MPRSP planning processes mightbe strengthened.
Government revenueFirstly, it is important to note that the resources availableto the government of Malawi are limited. Analysis ofgovernment revenue shows that, as a percentage of GDP,Malawi’s dependence on revenue and grants has steadilyincreased from 23% of GDP in fiscal year 2001/02 to 25%of GDP in 2002/03 and is expected to increase to 35% ofGDP by the end of 2003/04. Domestic tax and non-taxrevenue account for nearly 60% of the Governments totalrevenue for 2003/04 see Figure 3.
Under the HIPC initiative, countries that are between‘decision point’ (when commitments to debt relief aremade) and ‘completion point’ (when the final debt write-off occurs) are entitled to some interim relief on theirdebt service payments. In the Malawi case, this, however,depends on the country being ‘on-track’ with its IMFPoverty Reduction and Growth Facility (PRGF). The clearindications are that it is not on-track. Persistent pooreconomic performance, fiscal imbalances, expenditure onnon-priority areas and governance problems have meantthat the full grant has not been disbursed. Consequentlydonors have virtually cut off non-project aid to Malawipending a decision by the IMF to review the PRGF. Withthe PRGF to-date off-track, HIPC debt relief has
Figure 3: Share of revenue 2003/04
Total tax and non taxrevenue (60%)
Projectsupport (8%)
HIPC debtrelief (9%)
Japanese debtrelief (4%)
Generalelections (1%) Balance of
payment support(18%)
Source: MoF, Monitoring Unit
Table 12: Share of HIPC to total revenue
Source: Ministry of Finance
HIPC (MK million)HIPC (MK million)HIPC (MK million)HIPC (MK million)HIPC (MK million)HIPC (US$ million)HIPC (US$ million)HIPC (US$ million)HIPC (US$ million)HIPC (US$ million)HIPC share to tot. RevenueHIPC share to tot. RevenueHIPC share to tot. RevenueHIPC share to tot. RevenueHIPC share to tot. RevenueHIPC share to tot. Govt Exp.HIPC share to tot. Govt Exp.HIPC share to tot. Govt Exp.HIPC share to tot. Govt Exp.HIPC share to tot. Govt Exp.
2001/22001/22001/22001/22001/21,986
306%7%
2002/32002/32002/32002/32002/33,232
438%5%
2003/42003/42003/42003/42003/45,300
539%9%
Table 13: Projected Pro-Poor Expenditure allocation (million of MK)
AgricultureWWWWWater and Sanitationater and Sanitationater and Sanitationater and Sanitationater and SanitationRural Feeder roadsSmall-scale EnterprisesTechnical and Vocational TrainingTourismEducationHealthCommunity ServicesSafety netsGood governanceTTTTTotal PPEotal PPEotal PPEotal PPEotal PPE
Given the limited resources and the massive demand forpublic investment, the government must prioritise theallocation of resources based on clear criteria, systematicallyapplied. The introduction of MTEF and the PRSPrepresents an attempt to link annual, medium and long-term plans through the preparation of output-basedbudgets. The idea is that under the MTEF a subset of keypriorities, programmes and projects are identified by theline ministries themselves and the expected benefits,impacts and outcomes set out in a definitive manner. Inaddition mechanisms have been developed to ring-fencePro-Poor Expenditure (PPE) against budget cuts in orderto avoid fluctuations in the flow of resources to priorityareas. Table 13 sets out the Projected PPE Allocation overthe period from 2001/04.
19
Malawi case study
Pro-poor expenditurePro-poor expenditure is not additional money, but rathera way of redefining parts of the existing budget. MK12.15bn was allocated to PPE in 2003/04 representing21% of total government expenditure. It is important tonote that although the total of MK 12.15bn is protected,the allocation of funds within the PPE is subject to changeand money can be moved from one sector to another. Thestudy’s investigations show that this worked to thedetriment of the water sector in FY 2002/03.
Table 14 shows that MK275 million was allocated tothe water supply and sanitation sector in FY 2002/03 ofwhich the MoWD received MK216.1 million (or 79% oforiginal allocation). In contrast, in the same year, theeducation sector increased its share of the PPE budgetfrom 49.7% to 53% and the agriculture sector increasedits share from 6.3% to 8%. This trend for WSS seems set tocontinue in FY 2003/04. Data from the MoF shows thatthe MoWD had received just MK81.1 million or 47% ofits expected annual allocation by January 2004 (see Table15). In total the Ministry of Finance has disbursed K6.9billion of the K12.15 billion to the relevant ministries forPPE. This represents 56% of the total budget. Howeverthe MoWD has received 9% below this average and is infact the ministry that has received the lowest proportionof its annual allocation.
The annual PPE budgeting process is another area wherethe water and sanitation sector seems to be losing groundin comparison to other sectors. Table 16 illustrates howdifferent sectors fared in the FY 2003/04 allocation process.
Table 14: Actual PPE Expenditure in 2002/03 Illustrating the changes within the year against the originallybudgeted amounts
AgricultureWaterEducationHealthGender youth and community servicesPoliceOther ie feeder roads, mining, tourismTTTTTotalotalotalotalotal
Change as %Change as %Change as %Change as %Change as %of originalof originalof originalof originalof original
Source: MoF, Monitoring Unit
AgricultureWWWWWaterateraterateraterEducationHealthGender youth and community servicesPoliceOther ie feeder roads, mining, tourismTTTTTotalotalotalotalotal
56%47%47%47%47%47%64%51%58%60%50%56%56%56%56%56%
Source: MoF, Monitoring Unit
Table 15: Comparison with other sectors fordisbursement of PPE Funds to date in 2003/04
AgricultureWaterEducationHealthGender youth and community servicesPoliceOther ie feeder roads, mining, tourismTotal
It can be seen that the water sector effectively underwenta 20% cut, whilst predicted expenditure on Agriculturerose by 78%. This is in a year when 15% more in Kwachaterms was allocated to PPE. The net effect of losing out inthe annual resource allocation and the budgeting processesis that the water sector will be receiving 37% less in 2003/04 than it did in 2001/02. In the same period spending onAgriculture in the PPE rose 146%, whilst the total PPEbudget rose 29% in Kwacha terms (see Table 17).
There are a number of reasons why this should be thecase. The strategies of major donors are a key factor. TheEuropean Union, for example, has recently become a majorinvestor in the agriculture sector. Other factors include ageneral lack of confidence and support for the MoWD(vis-a-vis alternative channels such as the Presidents Officeor NGOs), a lack of major donor interest in water, andineffectual lobbying by the ministry (see below).
20
Implementation of Water Supply & Sanitation programmes under PRSPs
Under the or ig inal PRSP indicative resourcerequirements for WSS were equivalent to about 1.7% ofGDP. However, the annual budget allocation for WSS hasin fact fallen by 37% during the past two years, from 1.4%of GDP in 2001/02 to 0.87% in 2003/04.
MoWD expenditureOut of a total government expenditure of MK 56,737million in 2003/04, 3% or MK 1,771 million was allocatedto water supply and sanitation.
Of this MK 1,771 million:• MK1,623 million was allocated to development
expenditure;• MK148 million was allocated to recurrent expenditures
(Table 18).
MoWD budget sourcesTraditionally, the water and sanitation sector has beendonor-led. Domestic, bilateral and multilateral financialsupport was mobilised from institutions such as KFW, JICA,CIDA, NORAD, the World Bank, UNICEF, EU, ADB,and NGOs towards investment in water and sanitation. Awide range of grants and loans have been directed towardswater resources development, urban and rural water supplyand sanitation, capacity building at all levels, water qualitymonitoring and environmental protection.
Currently funding is mainly from government, donors,and from NGOs (both local and international). There area number of organisations that are funding the sectorthrough multilateral and bilateral agreements with thegovernment. Government funding comes through theMoWD, the District Assemblies and also via other sectorssuch as health and agriculture although the latter are mainly
Source: MoF, Monitoring Unit
AgricultureWaterEducationHealthGender youth and community servicesPoliceOther ie feeder roads, mining, tourismTotal
for institutions. Several NGOs also provide resourcesdirectly to districts. But at present there is very littlecoordination of resources flowing into the sector.
Domestic, bilateral and multilateral financial support in2003/04 amounted to MK1,613 million. This can beroughly categorised into who provided the funds, thedegree of MoWD control and how it was used.
Where the provider is for example, World Bank/Malawigovernment, as referred to in Table 19, this implies thatthe costs are shared. A normal ratio of funding would be90:10. The level of control (fiscal and administrative) theMoWD has over the development budget expenditurevaries from total control to virtually no control. Althoughassessing the degree of control is a subjective process basedon the knowledge of the programme, it is interesting toreflect on where the control lies.
3%15%10%5%4%7%3%
53%100%100%100%100%100%
21
Malawi case study
With projects funded by the Japanese government forexample, the MoWD suggests and agrees areas where waterpoints are needed, but then have virtually no control overany other aspect of the project (an example of the thirdcategory in Table 20). With the World Bank fundedNational Water Development Project (NWDP) a separateunit was set up with staff seconded from the Ministry, butwhich fell outside the management structure of theMinistry. This is an example of expenditure where theMoWD has partial control and some influence over thework – the second category in Table 20. The MoWDdoes implement some projects directly, such as the drillingof boreholes and rehabilitation of rural piped schemes usingHIPC money. This would be the type of project wherethe MoWD was considered to have full control (the firstcategory in Table 20).
Performance assessmentIn theory the introduction of MTEF in Malawi representsan attempt to move towards performance-relatedbudgeting. There is general agreement that the governmentneeds to do a better job of measuring its performance.Performance information can take several forms including:• The outputs, or activities, of government programs or
services;• The outcomes, or effectiveness of government programs
or policies;• The efficiency (cost per output) or cost-effectiveness
(cost per outcome) of government programs or services.
Capital investment decisions are typically based on arange of criteria including local need, financial (coveringincome and expenditure associated with the project),economic (wider economic framework in terms of costsand benefits and affordability), technical criteria (relatingto the nature of investment and how and when it can beimplemented), environmental (including the relevantenvironmental cost and benefits of a project), commercialand institutional (that impact on the timing or the successof the project), social and political criteria (especially insupport of or opposition to a proposed project).
The MPRSP currently includes a number of rather crudetargets relating to ‘Good access to WSS’. These include:increasing the proportion of households with access topotable water to 84% (from 65.6% in 2001); increasingborehole functionality to 100% (from 60% in 2001); andincreasing coverage for sanitary excreta disposal from 81.4%to 100%.
To achieve poverty reduction through the PRSP processit is essential that a system exists whereby those not servedcan be targeted. This Malawi study has observed that nosystematic criteria are used to guide investment planningand finance. The sector currently does not have such asystem or in fact any system to monitor the progress thesector is making in achieving such targets as the MPRSPand the MDGs.
With this in mind WaterAid Malawi, together with theSalima District Assembly, has developed a GIS/GPS basedmapping process which enables the unserved to beidentified. This offers a real opportunity for improvingarea, district and national planning and monitoringprocedures. It has also led to a process that quantifies theequity with which water points have been distributed (seeSugden & Stoupy, 2003). The findings show that there arehuge disparities in service levels with many communities
The MoWD has full control of only around 12% oftheir development budget, with the rest of their budgetbeing restricted in some way. The existence of multipleprojects with different objectives and planning horizonsmakes it very difficult to develop a coherent sector-wideplan or programme. The recent creation of a Ministry ofEconomic Planning and National Economic Council anduncoupling of national policy and investment co-ordination from the sectoral ministries has resulted in adecline in the planning capacity of sectoral ministries(including MoWD) and water sector investments in Malawiremain highly fragmented. Theoretically the MoWD isresponsible for the coordination of sectoral investments,but this analysis highlights significant constraints to effectivecoordination.
The lack of a coherent sectoral investment plan,combined with weak capacity within the MoWD forplanning and implementation, discourages donors fromengaging in programmatic support. This is not in line withthe new scheme of aid delivery under PRSPs. This viciouscycle is difficult to break, based as it seems to be on a lowdegree of trust which donors have in the capacity of theMinistry to effectively manage its responsibilities.
Theoretically the MoWD is responsible for thecoordination of sectoral investments but this analysishighlights significant constraints to effective coordination.With knowledge of each project it was possible to dividethe budget into different uses. The results are given below.
2.5% of the budget shown in Table 21 is a reflection ofthe low level of importance that the ministry and donorsgive sanitation.
Table 20: The MoWD development budget bydegree of control
MoWD in full controlMoWD with partial controland some influenceMoWD with little control orinfluenceTTTTTotalotalotalotalotal
greater than 1,000100 to 25050 to 10050 to 100less than 50less than 50
less than 50
Outside MoWD budgetOutside MoWD budgetOutside MoWD budgetOutside MoWD budgetOutside MoWD budget 75%75%75%75%75%MASAFMMMVarious other organisationsEUWorld VisionInter AideGTZPrivateAction AidPlan InternationalVarious Religious InstitutionsProscarpUNICEFAfricareCPARSCF_UKOXFAMADRACCFEkwendeni MissionOPCWaterAidPresidential Fund
greater than 1,000greater than 1,000
500 to 1,000250 to 500250 to 500100 to 250100 to 250100 to 250100 to 25050 to 10050 to 10050 to 10050 to 10050 to 10050 to 10050 to 10050 to 10050 to 10050 to 10050 to 100less than 50less than 50
less than 50
Table 23: Number of new water points installedvia MoWD budget
WVI (UNICEF)GitecCPAR (UNICEF)JICACOMWASH - from CIDAMoWD (HIPC)Unicef district supportTTTTTotalotalotalotalotalPercent of target
continuing to remain unserved despite substantial overallinvestment in the sector in recent years. Other communitiesin contrast have received additional water facilities despitealready being served at a level above those recommendedby the MoWD (see later).
Off-budget and off-plan workingOne of the spin-offs of the mapping process is a databasecontaining up-to-date information about whichorganisations are installing water points and where theyare placing them. With this data it has been possible tobreakdown the water points into providers. In total 33%of the country has been surveyed and it has been foundthat 25% of water points come from sources which areincluded in the MoWD budget (i.e. from the MK1,613million) and 75% were provide outside the MoWD budget.Table 22 below provides a breakdown of the organisationsworking within the sector and indicates by range theapproximate number of water points they have installedin the last five years.
In addition to ‘off-budget’ expenditure being significantlygreater than the money coming through the MoWD, it is
Figure 4: Number of new water points installedbetween 1998 and 2002 within the survey area(18% of Malawi)
20021999 2000 20011998
634704
828
648 681
0
100
200
300
400
500
600
700
800
900
also ‘off-plan’ i.e. outside of MoWD planning procedures– and beyond its control. This further compounds theproblem of coordination within the sector and indicatesthat reforms to the budgetary process instituted under theMPRSP are likely to have only limited impact on sectoralperformance.
Tracing off-budget expenditure is difficult but by usinginformation on the dates new water points were installedit is possible to show overall trends in sectoral investment.Figure 4 shows trends in the total number of water pointsinstalled in the survey area (18% of Malawi) between 1998and 2002.
In the survey area a total of 3,495 water points wereinstalled over the five-year period between 1998 and 2002.If this is extrapolated, based on population figures, to anational level, then a total of approximately 19,225 newwater point would have been installed within Malawi, atan average of around 3,835 per year. This investment isfrom all sources, not just the money in the MoWD budget.As the target within the MPRSP document is for ‘2,500boreholes’ per year it can be seen that the investment overthe preceding 5 years may be as much as one and halftimes (153%) indicative MPRSP requirements.
This clearly shows that the main problem in Malawi isnot the level of investment, but rather the location of thenew water points and the methods used for targeting thepoor. Stoupy & Sugden (2003) show that, if new waterpoints are better targeted to unserved areas, the water supplyMDG could be achieved in Malawi without requiringmajor increases in capital investment.
23
Malawi case study
Non-state actorsTo gain an estimate of past and expected Non-State Actor(NSA) activity within the sector, a brief survey wasundertaken by the researchers looking at new andrehabilitated water points and piped water schemes. Theresults of the survey show that progress towards the MPRSPtargets would be poor if only those water points installedthrough the MoWD are considered (Table 23). However,the picture is greatly improved if the work of all NSAswithin the sector is also considered (Table 24).
Table 24: Number of new water points installedby NSAs
Malawi Red CrossEU Micro ProjectsWaterAidCCAPCCAPWVI (own funds)WVI (UNICEF)GTZ (Machinga)InterAideGitecEmmanual InternationalCPARPlan InternationalJICACOMWASHOxfamMoWD (HIPC)Unicef district supportTotalPercentage of target
To allow for transparent use of funds, each district has aDistrict Development Fund (DDF) into which money canbe deposited and is governed by strict rules laid down bythe Ministry of Local Government. The key to thedecentralisation process is therefore the development ofthree-year District Development Plans (DDPs), includingthe identification of key areas for resource allocation withindistricts, and receiving money to carry out the plans andthe transparent – as well as accountable use of the money.In theory, the consolidated output of all district plans shouldform the basis of the national level plan.
The Malawi case study examined the actual processes ofDDP development and implementation from a water sectorperspective in two case study districts of Salima andMachinga. The decentralisation process is still relativelynew; although the act was passed in parliament in 1998 ithas yet to be fully recognised by either the donors or thecentral ministries. The MoWD was supposed to be one ofthe first ministries to devolve its functions, but to-date itremains highly centralised. The official date forcommencement of fiscal decentralisation was July 2003,but at the time of writing in January 2004 no money hadbeen released from the MoWD to District Assemblies. Thissituation has resulted in considerable uncertainty. Districtstaff generally still feel more accountable to ministry HQin Lilongwe than they do to the District Assemblies whoseauthority is undermined by lack of resources. This situationhas arguably led to a (at least temporary) decrease ineffectiveness while uncertainty persists.
The study found that procedures for development ofthe DDPs are well-defined but that links downwards torespective village level and upwards to national levelplanning processes remain weak. For example, expenditurecategories defined at different levels are not always thesame, which makes allocation of funds unsystematic.Analysis of DDP plans in case study districts shows thatalthough water is given high priority in DDP plans, itoften receives much lower priority in terms of actual levelsof expenditure. The main reason for this is that prioritiesof donors, NGOs and government departments are oftendifferent and tend to override priorities identified in theDDP. District Assemblies currently lack both the capacityand authority to coordinate and regulate sectoralinvestments.
In the districts sampled, accounting for funds depositedin the DDF leaves much to be desired. Tracking where themoney has actually been spent is difficult due to the lackof any systematic monitoring procedures. As a result, donorconfidence in DDFs remains understandably low. In bothcase study districts, HIPC funds earmarked for the watersector are currently not directed through the DDF andDistrict Assemblies are not involved in the process ofselection, implementation or monitoring these investments.Instead, HIPC funds are controlled and disbursed byMoWD central and regional offices. WaterAid Malawi isone of only a few agencies actively working with andthrough district assemblies in an attempt to strengthen thecapacity of district authorities for decentralised planningand management of water sector developments.
The current predominance of NSAs undermines the
This shows that overall levels of sectoral investment arein fact substantial, despite the current economic climatein Malawi. Actual figures may be even higher as some ofthe organisations known to be active in the sector did notrespond to the questionnaire in time to be included in thestudy.
The Malawi case study investigated the underlyingreasons for current trends and disparities in allocation ofsectoral resources, and identified existing constraints(technical, financial, institutional) to more equitableresource distribution. WaterAid Malawi aims to use thesefindings to bring sector stakeholders together and developa more transparent and equitable systems for resourceallocation and strengthen the district capacity to co-ordinate, plan and monitor the water sector in their districts.
Management of the water sector within districtsAs referred to above, Malawi is in principle going througha decentralisation process aimed at devolving administrationand political authority from the central ministries inLilongwe to the districts. Structures at district level havebeen established and Councillor elections have taken place.In theory, the structure is in place for ministries todecentralise their responsibilities to the district and forthem to take a more guiding and regulatory role. In practice,the pace of change has been slow and the flow of resourcesfrom central government and donor agencies to districtscurrently amounts to little more than a trickle.
24
Implementation of Water Supply & Sanitation programmes under PRSPs
Table 25: Sources of funding for 1,218 waterpoints installed in Salima and Machinga between1998 and 2002
GovernmentGITECJapan cooperationNational WDPSub Total
On MoWD BudgetOff MoWD BudgetDistrict TDistrict TDistrict TDistrict TDistrict Totalotalotalotalotal
221 (31%)486 (69%)
707707707707707
54 (11%)457 (89%)
551551551551551
Table 26: Functionality of water points bytechnology at district level
effectiveness of budgetary and policy reforms at districtlevel. The relationships between different NSAs active inthe water sector and District Assemblies (DA) can becategorised as follows:
1. Completely isolated – no formal or informalrelationship with the DA or MoWD at local or nationallevel, operates entirely independently e.g. the MuslimAgency
2. General national agreement with MoWD – agencyestablishes a formal agreement with the MoWD atnational level which enables them to work in anydistrict. District authority involvement varies but istypically limited e.g. MASAF
3. Local agreement for a multi-sectoral approach –agencies establish formal agreements with DA to workin a number of different sectors but the agreements
generally do not specify how resources will be allocatede.g. ActionAid
4. Local agreement specifically on water supply provision– district wide or area specific agreements with districtauthorities to provide water supplies but agenciesmaintain autonomy in resource allocation decisions e.g.SC UK
5. Local agreement specifically on water supply provision,resources allocated by the DA – formal agreementswith DAs to support district authorities (technical andfinancial) to plan and implement water supplies withina specified area e.g. WaterAid.
One consequence of the fragmented nature of watersector developments in the case study districts is a lack ofpolicy coherence. Different agencies adopt differentinterpretations of the national water policy leading toconsiderable confusion among beneficiary populationsregarding institutional roles and responsibilities inmanagement and financing of water sector developments.It is also a key factor underlying poor coordination ofsectoral investment leading, in some cases, to duplicationof effort, and in others to neglect of those most in need. Afurther significant issue highlighted by the mapping exerciseis the use of technology inappropriate to hydro geologicalconditions, which further undermines scheme sustainabilityand sectoral efficiency.
The Malawi case study also explored the impact atcommunity level of these problems of poor coordinationand planning and weak monitoring and regulation of policy.These are analysed firstly in terms of equity and secondlyin terms of sustainability.
SurSurSurSurSurveyed areaveyed areaveyed areaveyed areaveyed area
78.463.277.981.576.877.080.481.5
75.975.975.975.975.9
57.157.373.558.766.056.077.176.6
64.664.664.664.664.6
16.073.332.833.472.563.762.223.9
41.641.641.641.641.6
25
Malawi case study
EquityEquity of distribution is an important consideration ifpoverty reduction objectives impacts are to be achieved.WaterAid Malawi has developed a systematic approach tomapping the distribution of Improved Community WaterPoints (ICWP) using a combination of basic GPS unitsand a simple questionnaire (Stoupy & Sugden, 2003). Thisinformation is combined with population data anddisaggregated by Enumeration Area (EA). The resultingICWP density maps show the number of water points per1000 people. The MoWD target is 4 per 1000 or onewater point for every 250 people. However the resultingmaps reveal highly uneven distributions with investmentstypically clustered in specific areas, typically close to roads(see Figure 5).
Such maps provide a very powerful visual representationof inequities in resource allocation and form a valuable
295 enumeration areas
ICPW Density ICPW Density ICPW Density ICPW Density ICPW Density (No of water points per 1,000 hab)
Figure 6: Equity of investment
current WP density
WP density before 1998
Figure 5: Map showing distribution of water points in Mwanza and Pemba
tool for improved decision making. WaterAid is currentlyworking with local authorities to produce district levelmaps (see Annex).
Figure 6, for example, illustrates equity of investmentin Mulanje South District. It shows 295 EAs arranged inorder of increasing water point density. As such it indicatespatterns of investment in different areas. It shows clearlythat some EAs are well below the recommended densityof 4 per 1000, whilst others are well above. It shows that67% of the population in Mulanje live in areas with awater point density of less than 4 per 1000, while 8% livein areas with a water point density of greater than 8 per1000.
The darker upper shaded area shows how the density ineach enumeration area has changed over the last five years.Closer examination reveals that recent investment in alreadywell served areas (B) has been significantly greater than inpoorly served areas (A). It also shows that a number ofEAs have remained unserved despite substantial overallinvestment, while already well-served areas have continuedto benefit from new investment.
SustainabilityThe long-term benefits of improving equity in thedistribution of water sector investments will depend onthose investments being sustainable. Experience shows thatachieving sustainability requires attention to a complexmix of managerial, social, financial and technical issues manyof which are interdependent. Survey data covering eightout of 28 districts (29%) shows the percentage of function-ality of water points by technology type (Table 26).
It appears that boreholes (76%) are generally more‘sustainable’ than gravity fed rural schemes (42%) but itshould be noted that there has been substantial investment
26
Implementation of Water Supply & Sanitation programmes under PRSPs
in boreholes over the past five years. As a result 44% ofboreholes are less than five years old and are yet to reachthe stage of their life when major repairs are required.
It is interesting also to note the rate at which water pointsfall into disrepair or are abandoned. Figure 7 below, derivedfrom the water point survey, shows the functionality ofhand pumps against the date of installation and whetherthe hand pumps have ever been maintained.
With 40% of hand pumps over 6 years old not working(and can be considered to be abandoned) it is clear thatinvestment is not being maximised. Any donor looking atthese figures should think twice before consideringinvesting in the future. The research identified a wide rangeof different factors contributing to non-functionality. Thesecan be broadly grouped into three main areas i.e. availabilityof funds to maintain the water point, access to technicalskills and access to spare parts. WaterAid Malawi have thusdevised a ‘sustainability snapshot’ as a tool for rapid appraisalof investment priorities (see Box 5).
The snapshot was applied in 39 communities in twocase study districts of Salima and Machinga. The issueswere explored in some depth in each location but theresults can be summarised as follows:• Only one community (4%) was thought to have ready
access to enough money to fund the most expensivemaintenance processes.
• 70% of communities in Salima and 50% in Machingawere thought to have enough ready money availableto fund simple repairs, but not enough for the mostexpensive maintenance processes.
• 26% of communities in Salima and 50% in Machingahad no ready money available to fund even simplerepairs.
There is a dilemma over ensuring adequate communityfunds are available to cover water point maintenance costs.People are often reluctant to contribute towards themaintenance of a water point when it is working (‘Whydo you need maintenance money, its working alright?’)but are equally reluctant to contribute towards themaintenance of a water point when its broken (‘Why
Box 5 Sustainability snapshotThe surveyors are asked to decide which of the statementsis most applicable to the community being assessed withregard to the availability of sufficient fund skill and spareparts to keep their water point working.FinancialFinancialFinancialFinancialFinancial1. No funds available for maintenance when needed2. Funds available but not sufficient for the most expensive
maintenance3. Funds available and sufficient for the most expensive
maintenanceTTTTTechnical skillsechnical skillsechnical skillsechnical skillsechnical skills1. Technical skills not available for maintenance when
needed2. Some technical skills for maintenance, but not for all.3. Technical skills for all maintenance processes availableEquipment and spare parEquipment and spare parEquipment and spare parEquipment and spare parEquipment and spare partststststs1. Not available when needed2. Available but not for all repairs3. Available for all repairsFrom this the areas of strengths and weakness with regardto community management can be derived and aprogramme or project can be guided as to the area’s onwhich it needs to concentrate its efforts.
Figure 7: Percentage of functional handpumpsdependent on installation date and maintenance
1996and before
1997 1998 1999 2000 2001 2002
functional overallfunctional never mantained
%
%
%
%
%
%
%
%
%
%
%0
10
20
30
40
50
60
70
80
90
100
should I pay, its not working and I’m not getting anywater?’). Cash flow is a major problem in many rural areas,especially if the water point breaks at the wrong time (‘Itsthe hungry season, can I pay in June?’). There is also oftena lack of trust within the communities over any form ofmoney collection. No matter how well intentioned, if aperson has access to the community maintenance fundand they need to feed their children, they will take themoney. It’s not always a question of dishonesty; it can bemore one of need. These problems are compounded bythe lack of access to credit or banking facilities.
Study revealed that spare part supply was not as big aproblem as anticipated.
• 31% of communities had access to all type of spareparts;
• 41% of communities knew only where to buy the fastwearing spare parts;
• 28% of communities did not know where to buy spareparts at all.
It is interesting to note that a good reliable source ofspare parts seems to be more important than distance thecommunities have to travel to buy them. However, thelack of methods to communicate information tocommunities on where spare parts can be purchased is amajor problem.
In terms of skills the snapshot showed that:
• 17% of communities have access to the skill necessaryto carry out the most difficult repair processes;
• 82% of communities in Machinga and 74% in Salimawere thought to have to access to the skill necessary tocarry out the simple repairs;
• 19% of communities in Machinga and 9% in Salima
27
Malawi case study
were thought not to have access to the skill necessaryto carry out even the simple repairs.
It should be noted that one CBM training session doesnot result in a sustainable water point. Furthermore thetiming of training is all-important. This usually occurs atthe wrong time, as the critical period is usually 2-3 yearsafter installation. Communities can sometimes hire skilledpeople, but this increases the cost of maintenance andcompounds money collection problems. There areopportunities for private sector organisations to support
communities in handpump maintenance, but these servicesneed to be regulated. One of the biggest problems is thatskilled/trained people may leave the community and aredifficult to replace.
These results show that the key factor affectingsustainability in the communities surveyed is availabilityof maintenance money and not spare parts or technicalskills. Developing a transparent and trusted system forcollection and management of user contributions isparticularly important.
28
Implementation of Water Supply & Sanitation programmes under PRSPs
V. Uganda case study
IntroductionThe Uganda study examines whether PRSP and nationalsector goals are actually being achieved on the groundby Ugandan local governments in rural areas, and howimprovements in planning, monitoring and evaluationin those local governments could potentially improvethe efficiency and effectiveness of service delivery.Despite the sectors slogan of ‘some for all, not all for some’the study presents strong evidence that water servicesare being delivered increasingly inequitably, whistsanitation and sustainability remain secondary concerns.It goes on to examine the underlying factors, includingwhy WSS planning and M&E systems contribute to thisbreakdown between policy objectives, planning andimplementation.
In Uganda, Water and Sanitation has been made a majorpriority of the government and the sector was given ahigh profile within the Poverty Eradication Action Plan,Uganda’s PRSP. The government has set ambitiousobjectives to ensure that the entire population has accessto safe water and sanitation by 2015, in line with theMillennium Development Goals. Since 1997 the sectorhas developed a set of coherent policies and strategies, withrelatively clear institutional responsibilities and financingmechanisms. On the strength of these reforms, and thehigh priority of water needs expressed by the poor inparticipatory poverty assessments, the sector has receivedsubstantial increases in funding from the government whichhas chosen to allocate a significant proportion of its ownrevenues, including HIPC debt relief to water.
Now the challenge is to ensure that the strategies andreforms actually achieve their expressed objectives. Inthe rural sub-sector, there is concern emerging over theperformance of local government in their new role ofplanning for and actually delivering services. Despitereported national increases in safe water coverage from39% in 1996 to 51% in 2003, there are still majorquestions over the value for money, equity andsustainability of water and sanitation services beingdelivered in rural areas.
Progress in budgetary reformsUganda is rightly considered a leader in reform of thewater and sanitation sector in Africa. Coherent legal, policyand financing frameworks have evolved, which haveincluded the development of strong coordinationmechanisms through a sector wide approach, decentralizedservice delivery modalities, and the sector’s full integrationin the PRSP3.
The policy and legal framework for the water andsanitation sector was set out in the 1995 Water Statute,which established the principles of community managedwater and sanitation services, through the formation ofWater User Committees and Associations, as a means ofimproving sustainability of facilities. However, furtherreforms in the water and sanitation sector were necessitatedby Uganda’s decentralisation policy which emerged in the
mid 1990s. A new Water Policy was finalised in 1999,consistent with decentralization policy, and elaborating onthe principles of the demand responsive approach.
A rural water and sanitation reform process was initiatedin 1999 in order to put into operation the principles ofthe new Water Policy. This culminated in the finalisationof the Rural Water and Sanitation Sector Investment Plansetting out the investments required to meet sector goalsby 2015. The objective of the rural water and sanitationsector as stated in this plan is as follows:
‘Sustainable safe water supply and sanitation facilities, basedon management responsibility and ownership by users, withineasy reach of the rural population by the year 2005 with an80% - 90% effective use and functionality of facilities. Theneventually to...100% of the rural population by the year 2015’(MWLE-DWD, 2000)
A five year Rural Water and Sanitation Operational Planwas completed in 2002. This plan sets out in detail boththe operational modalities that guide implementation forthe rural water and sanitation sub sector, investmentsneeded and their cost over the five year period from 2002until 2007. This applies not just to water and sanitationinfrastructure, but covers also institutional requirementsfor delivery at local government levels, and programmesupport at central government level.
Meanwhile, as poverty reduction moved up the politicalagenda, the water and sanitation sector began to emergeas a government priority, starting with the preparation ofthe 1997 Poverty Eradication Action Plan (PEAP). Thisrecognition led to the sector receiving a significant boostin funding from the HIPC debt relief initiative in 1998,and being included in the Poverty Action Fund (PAF), thebudgetary mechanism through which debt relief funds andother earmarked donor budget support were channelled.The national Uganda Participatory Poverty Assessment(PPA) was carried out in the late 1990s, as a precursor torevising the PEAP, and safe water emerged as one of thekey priorities of the poor. This reinforced the priority ofthe WSS in the 2000 PEAP, which also served as Uganda’sPRSP. Subsequently, the Government of Uganda allocatedover $11 million extra to the sector from a second roundof debt relief awarded through the enhanced HIPCinitiative, more than doubling the government own budgetallocations, excluding donor projects. These funds wereexclusively allocated to rural water and sanitation andchannelled directly to local governments via earmarkedconditional grants.
This combined with further budgetary increases insubsequent financial years meant that between 1997 and2002 government budget allocations to the water andsanitation sector rose from just over $3 million to $31million, or from 0.5% to 2.8% of the Government ofUganda Budget over the same period (Figure 8). Muchof this was allocated to rural water and sanitation.
Figure 8: Government of Uganda water andsanitation budget allocations, excluding donorprojects
The combination of a coherent reform process, thedevelopment of sector investment plans, and the evolutionof sector financing provided the key ingredients forimproved coordination in the sector. This provided thebasis for the development of a sector-wide approach bythe government and donor and NGO partner agencies.Government funding is now by far the largest contributionto the rural sub-sector, and this encouraged many donorsto consider moves towards budget support.
Annual joint reviews of progress are now held to reviewand discuss the performance of the sector against establishedoperational and investment plans and strategies. At thesereviews various actions to be undertaken by governmentand donors alike are discussed and agreed. Donorscollectively rather than individually identify and agree issueswhich they wish to raise with government.
Progress in implementationThis section examines the extent to which equitable andsustainable rural water and sanitation services, the basicobjective of the Rural Water and Sanitation OperationalPlan, are being provided in Uganda. The study focused ontwo case study districts of Tororo and Wakiso.
Officially, national safe water and sanitation coverageincreased from 39.4% in 1996 to 51% in 2003. Thesefigures are based on assumed coverage which is measuredby multiplying the number of point sources by therecommended number of people that should be served byeach type of source, and comparing this to the level ofpopulation in the country or district.
For example it is assumed that one borehole serves 300people and a spring 200 people. There is a wide variationin coverage throughout the country from 25% in the leastserved district to 75% in the best served.
Data problemsAnother important problem to note here is that there aresignificant inconsistencies between the data held at thedistrict and that held by central government. In bothdistricts studied, the local administration reported to haveabout twice as many safe water points than were in thecentral government Management Information System, asTable 27 shows. This means that safe water coverage islikely to be underestimated nationally, but it is this datawhich is used to calculate allocation of funds to districts.
Currently safe water coverage data is only calculated atthe district level but there are important levels of localgovernment in Uganda below the district – i.e. sub-county,parish and village, where this is not done. National levelfigures and financial allocations do not therefore takeaccount of equity in service provision within districts.
The Uganda study applied a simple technique developedby Sugden (2003) in Malawi for comparing the equity ofservice provision at different levels (as described in ChapterIV). It simply involves the calculation of the number ofwater points for every thousand people in eachgeographical area (whether district, subcounty or parish).This is called the Improved Community Water Point (ICWP)Density Mapping, referred to henceforth as Water PointDensity (WPD). The national Water Policy states that eachwater point should serve no more than 300 people. Thisequates to a target WPD of 3.3 per 1000 people.
WPD calculations in Tororo and Wakiso distr ictsdemonstrate that otherwise encouraging national anddistrict level coverage figures are masking inequitableservice delivery, and weak operation and maintenancesystems within districts themselves. WPD per district wascalculated using district level data which is considered moreup to date. This stands at 1.40 per 1000 people in Wakiso
DWDDWDDWDDWDDWDDataDataDataDataData121229
-27028152
665665665665665
Table 27: Inconsistent data in local governmentsand DWD
Implementation of Water Supply & Sanitation programmes under PRSPs
Safe
wat
er p
oint
s fo
r 100
0 po
pula
tion
Figure 12: WPD by parish in Masulita, a well servedsubcounty in Wakiso
7
6
5
4
3
2
1
0
Figure 11: WPD by parish in Kasanje, a poorly servedsubcounty in Wakiso
Safe
wat
er p
oint
s fo
r 100
0 po
pula
tion
and 1.57 for Tororo (in both cases less than half the nationalrequirement set out in the National Water Policy). It isimportant to note that this figure excludes subcounties inWakiso which are on the outskirts of Kampala and benefitfrom piped water systems. Outside urban areas, water andsanitation service levels in Wakiso are very low.
The WPD was then calculated for all subcounties ineach district. Figures 9 and 10 show significant variations
in WPD across subcounties in each district4. In Wakisothe values varied between 0.47 in Kira subcounty and 3.59per 1000 people in Masulita subcounty, whilst in Tororothe variation was between 0.67 in Nawanjofu subcountyand 4.09 in Kwapa subcounty. In each case there is onlyone subcounty in each distr ict which meets therecommended service levels set out in the 1999 NationalWater Policy.
Figure 9: Water point density in sub-counties in Tororo districtSa
fe w
ater
poi
nts
per 1
000
popu
latio
nSa
fe w
ater
poi
nts
per 1
000
popu
latio
n
Figure 10: ICWP density in subcounties in Wakiso district
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
UF
OJN
AW
AN
AR
EG
NO
GA
N
AY
AP
EWL
OS
UB
AS
AMI
ZA
M
AW
ERI
K
AB
MU
DU
B
AG
NO
HC
AK
AJEL
AT
UB
UJU
KU
M
AB
AS
UB
UR
UK
US
O
AG
OY
UB
AN
AWL
OYI
AD
NAL
UM
OK
OSI
K
IG
NO
BU
R
OLO
M
AP
AW
K
Subcounties
ARI
K
UR
EW
BA
N
AB
MU
YA
MAN
EJN
AS
AK
IG
NA
SN
OB
AG
NA
N
IB
AT
AK
EB
MO
G
OSI
KA
W
AM
UK
US
UB
IRI
KA
K
ASI
SS
ATIIL
US
AM
Subcounties
0
0.5
1
1.5
2
2.5
3
3.5
4
0
1
2
3
4
5
6
7
OK
AM
ABI
ZZ
AG
NIZ
Z
UB
MUL
UB
IZ
AS
IS
SU
B
EJN
AS
AK
OLO
KO
S
OG
NU JJ
Parish
AD
NE
WK
UM-
B
ELA
BA
K
EZI
ZN
AK
UM
AKI
TA
K
ILA
AB
MU
TT
ED
DN
EW
ME
WL
AZ
EG
NE
YK
ED
DU
GN
UG
UL
EB
UG
NU
KIK
AN
ATIL
US
AM
EZ
NA
M
Parish
31
Uganda case study
Table 28: Increasing inequity the deeper you dig
Subcounties in Wakiso DistrictParishes in Kasanje S/cParishes in Masuliita S/cSubcounties in Tororo DistrictParishes Kwapa S/cParishes in Nawanjofu S/c
Figure 13: WDP for successive years in Kwapasubcounty
AM
ON
I
KWAP
A
MAL
ABA
KALA
IT
MEL
LA
APO
KOR
ICWP pre 1998ICWP 1998-2000
ICWP 2000-2002
Safe
wat
er p
er 1
000
peop
le
Figure 14: WDP in successive years in Nawanjofusubcounty
Safe
wat
er p
er 1
000
peop
le
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
BUGALO BUBBINGE BINGO
Parish
ICWP pre 1998
ICWP 1998-2000
ICWP 2000-2002
For each district one well served and one poorly servedsubcounty were chosen and WPD then calculated at parishlevel. This revealed even greater variation in WPD betweenparishes than between subcounties. For example, in Kasanjesubcounty (a poorly served subcounty in Wakiso Districtwhich had a WPD of 0.69) there were four parishes whichdid not even have a single improved water point betweenthem, whilst the best served parish had an WPD of almost1.5 water points per thousand (see Figure 11). Even inMasulita subcounty, a relatively well served subcounty witha WPD of 3.79, the WPD varied from between 0.77 in theworst served to 6.5 water points per 1000 people in thebest served (see Figure 12). Similar variations wereobserved in the well served and poorly served subcountiesselected in Tororo districts.
Greater variability indicates increasingly inequitableservice provision. By using and adapting another simpletechnique developed by Sugden (2003), a comparison canbe made, and a proxy indicator for the relative equity ofservice provision at those different levels calculated. Thereforefor each of the districts, the average deviation from themean subcounty WPD was calculated for each subcounty.This average deviation is expressed as a percentage of themean subcounty WPD (Box 8). In Tororo and Wakiso theaverage deviations were 33% and 36% of the meansubcounty WPD respectively.
The exercise was then repeated for parishes within eachsubcounty confirming even greater variation in distributionof water points between parishes. Average parish variationsranged from 47% to 101% of the mean parish water pointdensity. This means that the relative inequity in distributionof water points is universally greater between parishes thanbetween subcounties.
There was also evidence that some parishes continuedto benefit from new water sources year after year (see figurebelow for Kwapa and Nawanjufo subcounties), andsimilarly the same villages in those parishes.
All this data point to increasing inequity the lower onedelves into the distribution of water points. This indicatesthat district subcounty planning processes are open topolitical capture, with politicians being able to ensure
certain areas benefit from new water facilities more thanothers.
Therefore despite the reported national increases incoverage, earlier pointed out, equity in distribution is notbeing achieved. Furthermore, the level of inequity indistribution between subcounties and parishes has actuallyincreased over the years. The slogan ‘some for all, not allfor some’ does not therefore appear to be being realized.
A key observation of the Uganda research is that thesocio-economic status of the well-served subcounties wasgenerally better than the least served ones. They appear tohave higher levels of education, have influential people inthe community (in that they had NGOs operating in theirareas), were nearer to administrative centres, had betterroad networks and their ability to pay was greater than theleast well served areas5. The least served areas werecharacterized by comparatively poor road network, lowlevels of education, political marginalisation, weakleadership and lack of influential community members.This suggests that socio economic status of a subcounty isprobably a major factor in both the ability of it to attractnew water and sanitation facilities in the first place, andcapacity to operate and maintain the facilities onceestablished.
In summary, despite the small sample in this study, it ispossible to draw some important conclusions from thisanalysis of Tororo and Wakiso. It appears that the distributionof existing facilities is inequitable, and that the new facilitiesbeing constructed appear to be worsening not improving
32
Implementation of Water Supply & Sanitation programmes under PRSPs
Box 9: Planning & budgeting processThe planning and budget cycle begins in October of eachfinancial year where in a series of regional budget workshopsconvened by the ministry of finance, the planning andbudgeting process is outlined and grant ceilings are presentedto local governments.
The process progresses to the districts which are then expectedto produce a draft Budget Framework Paper by December,outlining the Local Governments medium term budget strategy.It is from the final LGBFP that the districts and sub-countiesproceed to prepare water sector workplans (MWLE- DWD2002). Whereas the planning cycle guidelines proposeconsultative meetings with the community to commence inearly February, the actual consultations start a lot later.
At the district level, the planning process starts with a sub-county consultative meeting involving Sub-county executivecommittee, parish development committee (PDC) membersand subcounty administration. Funding levels for respectivesectors for the fiscal year are communicated and PDCs aregiven planning formats. PDCs then conduct village consultativemeetings with the general community where WSS priorityactivities are set. The PDC committee in a separate meetingintegrates all the village plans into one Parish developmentplan using priority-ranking method. In the same manner, thesub-county executive committee guided by the extension staffuses the parish plans for developing sub-county plans. Theseplans are then submitted to the district for incorporation into adistrict plan.
Other stakeholders in the sector like NGOs and CBOs werealso facilitated with planning formats in Wakiso District.
the situation. The levels of inequity get greater at lowerlevels of local government.
The Uganda report goes on to consider ways in whichtools for quantifying equity and sustainability used in thestudy could form the basis for action by local governments.In particular, ways in which such analysis could help toaddress some of the fundamental weaknesses in planning,monitoring and evaluation of performance in the watersector at subcounty, district and national levels. It suggeststhat they could also help central government to measurethe performance of local governments with respect tonational objectives, and reward good performance.
Towards better targeted planning andbudgeting for WSS interventions within districtsAlthough equity and sustainability are stated objectives innational level policy, local planning and budgetary decisionsmade by districts and subcounties in Wakiso and Tororodo not appear to reflect these goals. This means that thereis a breakdown somewhere between policy, planning,budgeting and implementation. But where, and why?
By examining the strengths and weaknesses of theplanning and budgeting and implementation systems, theUganda study argued that the breakdown is largely downto a combination of poorly coordinated planning processes,and inadequate performance measures and tools that canbe used in these processes. This ultimately means that thereis inadequate incentive for technocrats and politicians toadhere to national sector policy priorities, and improveservice delivery performance.
There are well established systems for planning andbudgeting in Ugandan local government, and these aresupported by specific guidelines prepared for the waterand sanitation sector. The district wide planning andbudgeting process is supposed to be bottom up andparticipatory, although the quality of this participationvaries.
If the planning procedures are followed to the letter, thedistrict guidelines could, and probably should result inequitable distribution of water points. They emphasize abottom-up approach engaging lower levels of localgovernment and communities in planning decisions andrecommend prioritisation of areas with low coverage andunserved communities, incorporation of hygiene educationand community mobilisation and provision whereverpossible of a range of water technology options (MWLE-DWD, 2000).
Planning issuesHowever, the guidelines for planning evidently are notbeing fully followed in the case of Tororo and Wakiso.Although many elements of the process stipulated in theguidelines take place, the criteria for making decisions arenot always followed. This is due to a mix of technical andpolitical issues, which means that ultimately there are fewincentives for districts to implement the guidelinesrigorously. Politicians influence the allocation of watersources in both distr icts, and this has apparentlycompromised equity in distr ibution, with powerful
politicians continuing to extend services in their parisheswhich are often already well-served, leading to furtherconcentration of water facilities. In a decentralized politicalset-up, politicians necessarily should be involved in decisionmaking, however this should be within the bounds ofnational policy.
The key sector planning criteria of geographical coverageis not very useful below district level for ensuring equity,and can be subject to various interpretations. In Wakiso,the district were found to emphasise fair geographicaldistribution so as to spread resources across sub-countiesin the district (regardless of existing coverage figures) andto fulfill political demands (i.e. give equal shares topoliticians). The relative populations of different areas isgenerally not taken into account, although this informationis collected at the grassroots level. Conversely, in the annualreports, the total number of people being served by newwater points is emphasized – in line with political demandsat national level. The above practice satisfies politicaldemands, but has caused persistent inequities in distributionof WSS facilities.
As demonstrated earlier, using coverage figures as a proxyfor equity is deceptive because it masks inequity in thelocation of water points within subcounties. Hence someparishes can over a long period of time cumulatively receivewater facilities more than others, without this informationbeing picked up by the district or indeed the centre. Thus,while percentage coverage figures may be high, inequitieswithin the geographical area might also be high.
Population Population Population Population Population WWWWWater point densityater point densityater point densityater point densityater point densityper 1000 peopleper 1000 peopleper 1000 peopleper 1000 peopleper 1000 people
Explicit tools to promote equity in the planning processneed to be introduced at the district and lower levels. Asillustrated earlier the Water Point Density is very easy tocalculate, and use as a criteria to ensure more equitabledistribution of facilities between subcounties in a district,and parishes within a subcounty.
Typically a subcounty will only be allocated enoughfunds to cater for a handful of new water points eachfinancial year. WPD should therefore be a major criteriafor deciding in which parish water points should beconstructed. If a list of the Water Point Densities in eachof the parishes were presented to the subcounty council(e.g. Kasanje Table 29), it becomes very difficult for apolitician from a well served parish (such as Jjungo in thiscase) to argue for more water points, when threeneighbouring parishes have no improved water points. Thisdata could also be supported by the presentation of GISmaps to subcounty planning committees and councils,showing the actual location of existing water points withinthe parishes.
Budgeting issuesThe study also revealed problems in the selection oftechnology and planning at lower levels. Often lower localgovernments do not have choice over the technology mixwhich is determined by the district. For example, asubcounty may be offered two boreholes by a districtduring the planning process. While the subcounty mayprefer to construct a larger number of shallow wells at thesame cost, under the current system ‘technical choice’amounts to ‘take it or leave it’. Districts tend to have anatural bias towards more costly technologies e.g. boreholes,as these involve greater sums of money and more control.This means that efficiency of service delivery may beundermined.
Another key observation is the dwindling participationof communities in the planning process. The guidelinesrequire communities to participate in planning every yeareven when it is obvious that the chances of an individualcommunity getting a water point in the annual budget areremote. This leads to so-called ‘participation fatigue’.However, consultations for new water points withcommunities need not be an annual event. It is
straightforward to ascertain the communities which needwater points and those that are served and WPD mappinghelps obviate the need for unnecessary and repetitiveconsultation.
The study also showed that while information necessaryfor planning O&M is routinely collected (and clearly showsthe magnitude of the problem), there are no resourcesallocated for O&M activities. Similarly there is little or nobudget for routine software activities, to support andmonitor Water User Communities and ensure that theyare prepared to maintain boreholes. Nor is there adequateprovision for periodic major repairs. In both districts, therepair of boreholes is mentioned in the three-yeardevelopment plan as one of the strategies for increasingaccess to safe water, but O&M is neither itemized in theactivity schedule nor is it budgeted for. Instead all thefinancial allocations are intended for the construction ofnew sources. Resources for ‘borehole rehabilitation’ underconditional grants have been extended to DWD since1999/20006. However evidence of rehabilitation at thedistrict level is minimal. O&M plans, proposed in recentDWD guidelines, were not yet in place in case studydistricts. While there are limited attempts to put intooperation the policy guidelines on O&M, the division ofroles between community, subcounty and district foroperation and maintenance remains unclear, and no fundsare currently allocated to this purpose.
Clearly greater emphasis needs to be placed on the needfor the district and subcounty to plan for operation andmaintenance, including major repairs and rehabilitation ofwater points, instead of the current emphasis on newfacilities. Currently the district water conditional grant isby name a ‘capital development grant’, however it is beingused to finance both recurrent and development activities.Clearer division of the grant into a recurrent anddevelopment component, which could be specified bycentral government, would help ensure there is no unduebias towards new investments.
Sanitation is given similarly low priority in the planningprocess. Although about 20% of the WSS funds are meantto be targeted for hygiene and sanitation activities, this isgenerally only spent on latrine construction in public placeslike markets, schools and health centres. Hygienepromotion is more the remit of the health and educationsectors, and this is given inadequate priority relative todelivery of other aspects of services. Ultimately the softwareaspects of sanitation involve few really tangible results, and
34
Implementation of Water Supply & Sanitation programmes under PRSPs
hence is rarely a politically attractive option when allocatingfunds.
Towards coordinated and focused monitoringand evaluationUltimately it is monitoring and evaluation mechanismswhich ensure that services are being provided and sustainedas planned. This is important for ensuring aspects of equityand sustainability are actually taken care of, measured, andthat managers are able to make better decisions to improveservice delivery during the financial year and in futureplans and budgets.
In Uganda the district water and sanitation grantplanning and operational guidelines stipulate proceduresunder which districts generate and analyze informationfor planning and reporting on progress. Case study researchshows that whilst there is some flow of informationbetween political, administrative and interdepartmentalstructures in evidence, emphasis is put on verticalaccountability to central government. Information is notused effectively for managerial decisions and there is littleproactive feedback to the end users from districts. There isalso limited coordination and information sharing betweenthe district water office, extension workers and otherstakeholders (civil society). Furthermore the relationshipbetween the private sector and the benefiting communitiesis not clearly defined, thus constraining communities’ abilityto monitor especially construction activities. Consequently,the overall quality of information gathered is poor, and itsuse for decision-making weak.
District Water Offices have all been provided withfacilities such as computers, to facilitate data storage andanalysis of data and there have been limited attempts toequip stakeholders with basic skills and tools for M&E,which greatly affects their level of involvement andcontribution in M&E. It was observed that there isadequate staff at the district to carry out the monitoringand evaluation process, but they lack the specific skills andadequate, predictable funding to carry out routinemonitoring activities. Only PAF funds were allocated tomonitoring and evaluation, available without augmentationfrom other sources. This was attributed to the low revenuebase at the district and sub counties. It was also noted thatthe use of PAF monitoring funds was ineffective, in manyinstances monitoring is only done when PAF funds arrive,the timing of which is irregular, otherwise nothing takesplace.
Reporting and management information systems at thedistrict and sub-county level clearly need to be improved,including development and modification of monitoringinstruments, improving the process of collection, recording,analysis and storage, production of information forplanning, reviews and reports. More practical, structuredtools need to be developed for measuring the relative equityand sustainability of investments (including sanitation) andother issues such as value for money. Ultimately in-yearmonitoring and evaluation activities need to serve amanagement purpose and result in improvements in serviceprovision. Mechanisms for critical reflections of WSSperformance by all stakeholders including managers andcivil society need to be created, alongside downwardreporting systems to end users.
Measuring and monitoring equityThe study recommends that districts should carry outequity analysis of parishes on an annual basis, as outlinedearlier. This could involve:
• The calculation of WPD in all subcounties and parishesin a district;
• The calculation of the relative variation in water pointdensity across subcounties in the district, and acrossparishes in each subcounty.
These measures could provide an important input intoplanning, within districts and subcounties by allocatingmore resources to those areas with lower WPDs. Localgovernments could then aim to reduce the relative variancein Water Point Density year on year.
Measuring and monitoring sustainabilityThe sustainability snapshot, developed by Sugden (2003)and used in both the Malawi and Uganda study, provides avaluable tool for rapid appraisal. It could be furtherdeveloped and tested based on functionality, affordability,managerial issues, and include other factors as appropriate.It can be used both as an entry point to in depth discussions,or simply as a mechanism for getting quick informationfrom a member of the Water User Committee at a waterpoint, when carrying out an assessment of the physicalcondition of a water source.
This could be a routine function of subcounty extensionworkers, who would have a target of assessing thefunctionality and sustainability of all the safe water pointsin one financial year. Although the data on individualwater points may not always be reliable, the aggregateddata would provide valuable information on issues andproblems affecting sustainability in different parishes andsubcounties. The district could then verify subcountyinformation and carry out more in depth discussions at afew water points. Such a system would need to be pilotedand tested more widely.
There also needs to be more rigorous implementationof sectoral policy e.g. on capital contributions, andmonitoring of implementation. Subcounties, for example,should be required to produce evidence to the districtthat contributions have been collected and banked, beforeinvestments are carried out in their areas, as was the case
35
Uganda case study
Box 11: Possible tools for monitoring andmeasuring equity, sustainability and sanitationperformance• Relative Variance in Water point density by subcounty
and district• Functionality, finance, skills and spare parts through the
sustainability snapshot• Latrine coverage and use, hand washing, and safe
water chain through the sanitation snapshot
Box 10: Possible indicators in a sanitationsnapshotSanitation FacilitiesSanitation FacilitiesSanitation FacilitiesSanitation FacilitiesSanitation Facilities1 No latrine in existence for the household2 Latrine available, but dirty and poorly maintained3 Clean, well maintained latrine.Use of LatrineUse of LatrineUse of LatrineUse of LatrineUse of Latrine1 Family members rarely use the latrine2 Family Members sometimes use the latrine3 Family Members always use the latrineHand washingHand washingHand washingHand washingHand washing1 Family members rarely wash hands after relieving
themselves2 Family members sometimes wash hands, not always
with soap3 Family members regularly wash their hands with soap
and waterMaintenance of safe water chainMaintenance of safe water chainMaintenance of safe water chainMaintenance of safe water chainMaintenance of safe water chain1 Family members rarely clean containers for collecting
and storing water2 Containers sometimes cleaned, but not always3 Containers always cleaned before collecting water,
usually with soap
under donor funded projects.
Sanitation performanceDistricts and subcounties need instruments where theycan capture sanitation performance regularly and easily,just as with sustainability. However as sanitation is ahousehold responsibility it is very difficult to collectcomprehensive data, which is borne out in the impracticalnature of current data collection formats.
Sampling would be the only way to collect quality dataon a regular basis. One way may be to develop a tool forsanitation along the lines of the sustainability snapshot (seeBox 10) for gaining an impression of the hygiene andsanitation practices of different communities. Such a‘sanitation snapshot’ could be used to assess themes suchas existence of facilities, understanding of hygiene issues,and behavioural change.
A system for measuring sanitation performance is onlyvaluable if it goes hand in hand with planning for andallocating adequate funds towards activities which resultin improvement of sanitation practices. Emphasis shouldbe placed on strengthening the linkages between the districtwater office and other departments e.g. health andcommunity services, for example, thorough jointimplementation and coordination of activities.
A major problem often cited is facilitation. Thefunctioning of proper performance information systemscosts money, and is a routine activity that should be carryingon throughout the financial year. Firstly subcountyextension workers require adequate operational fundingto collect performance information from communities onsustainability and sanitation issues, as well as their routinemobilization activities. Currently this is not occurring onthe ground. The use of community extension workers insuch ways needs prior agreement of sector departments atthe district and national levels.
These systems are ultimately only going to be useful ifthey result in better decisions being made by managersand politicians, and if accountability to the end-user isimproved. Local governments need to be encouraged topublish their performance at subcounty and below, givingregular feedback to beneficiaries. Civil society organizationsand NGOs can be involved in monitoring. Sectoral andcross sectoral managerial decision making forums need tobe made more functional. Even with slightly differentinstitutional arrangements water and sanitation servicesrequire coordination within a district.
Aligning local incentives with central objectivesthrough local government performanceassessmentThe proposals outlined above are largely technical, whichif applied should facilitate better decision making, andenable different levels of local government to makedecisions which are in line with the achievement of sectorgoals. However, they do not directly address the politicalincentives that motivate powerful politicians to ensure theirvoters are served. This can never be totally overcome butthe incentives for districts to adhere to the national policiesand guidelines can be strengthened.
District performance assessmentIn order to align incentives with national policies routinemonitoring and evaluation could be supplemented bystructured performance assessment or benchmarking ofdistricts’ implementation of national water and sanitationpriorities, including measures for equity and sustainability.Performance benchmarking systems are common indeveloped countries such as the US and the UK, as a meansfor encouraging local governments to adhere to sectorpolicies and guidelines. Under the UK’s comprehensiveperformance assessment process league tables of councilperformance are created, and councils are required topublish their own performance, and make it available tothe public. If councils perform poorly with respect tonational goals and targets, then this is public knowledge.
Such practices are not new in Uganda. There is an annualperformance assessment of local governments under theLocal Government Development Programme (LGDP),where the centre assesses and scores districts adherence todecentralisation laws, policies and guidelines. Under thesystem districts assess subcounty performance and theseassessments are verified by central government. Districtsand subcounties are required to fulfil various minimumconditions in order to access a discretionary localdevelopment grant. If a local government performs well
36
Implementation of Water Supply & Sanitation programmes under PRSPs
they get an additional allocation. Box 12 shows that failureto perform with respect to the legal framework bore ahigh cost in Mubende District.
The Ministry of Health in Uganda has also started asystem of measuring and ranking district performance inprimary healthcare, and publishes a district league table inthe Annual Health Sector Performance Report. Thiscombines indicators of processes such as the timeliness ofinternal reporting from health centres, and outputinformation such as immunization rates and outpatientattendance. It also includes one indicator on sanitation –household latrine coverage, however the source for thisdata is unclear. Although there are no financial rewards orpenalties, the league table is made public, and this has
provided an incentive for district directorates of healthservices, and politicians to improve performance.
National performance measurementframeworkThe water sector in Uganda is in the process of developingan overall framework for sector performance measurementnationally, and a first sector performance report7 using thenew framework has been prepared for the Water SectorReview in October 2003. Ten different themes forperformance measurement have been identified, and it isproposed that ‘national performance reports’ cover one ortwo of these themes every year. These are encouragingdevelopments as such performance measurement is key toascertaining whether there is strong linkage between policy,planning and implementation.
Currently, assessment of districts in the sector is mainlythrough review of performance against previous sectorplans and reports which focus on new infrastructure andleaves out other key elements of performance. Districtperformance is only compared in terms of coverage. Inthe 2003 sector performance report there appear to befew specific proposals for a structured system forperiodically assessing the performance of districts relativeto policy objectives. Three to five golden indicators areproposed which could be used in assessing district levelperformance. These indicators may be able to provide someimportant information on the achievement of sector goals,but would not be specific enough to provide concreteincentives to local technocrats and politicians to start
Possible PerPossible PerPossible PerPossible PerPossible Perforforforforformance Measures in WSSmance Measures in WSSmance Measures in WSSmance Measures in WSSmance Measures in WSS
Table 30: Developing a district balanced scorecard in the water and sanitation sector
Achievement of missionAchievement of missionAchievement of missionAchievement of missionAchievement of missionThe extent to which objectivesand goals are being realised
EfEfEfEfEfficiencyficiencyficiencyficiencyficiencyThe value for money of servicesbeing provided
Customer PerspectiveCustomer PerspectiveCustomer PerspectiveCustomer PerspectiveCustomer PerspectiveHow well are customers beingserved?
SerSerSerSerService improvementvice improvementvice improvementvice improvementvice improvementHow has and what is thelikelihood that services willimprove?
Sustainable safe water supply andsanitation facilities, based on managementresponsibility and ownership by users,within easy reach of the rural populationby the year 2005 with an 80% - 90%effective use and functionality of facilities.Then eventually to ..100% of the ruralpopulation by the year 2015.
Water and sanitation services deliveredefficiently to the population, usingappropriate low cost technologies wherepossible
The population actively engaged indecision making over WSS facilities,managing and using high qualitysustainable water and sanitation facilities
Local Governments making improvementsin the deliverable of efficient, equitableand sustainable water and sanitationservices
• District water coverage• District Water Point Density• Average and relative variation in WPD• Functionality of existing water points• Household Latrine Coverage• Institutional Latrine Coverage
• Unit costs of constructing different facilities• Average per capita investment cost• The technology mix (proportion of low cost
technologies)• Collection of capital contributions
• Community engagement in the planningprocess
• Results from the sustainability snapshot• Results from the sanitation snapshot• Water quality & quantity
• Improvements in safe water coverage overthe last two years
• Improvements in equity over the last twoyears
• Improvements in unit costs over the last twoyears
• Quality of work plans
PerPerPerPerPerforforforforformance Areamance Areamance Areamance Areamance Area GoalGoalGoalGoalGoal
Box 12: The political cost of failing to perform inMubende districtSeveral subcounties failed to reach the minimum standards inthe internal assessment of the administrative capacity ofsubcounties, conducted by Mubende District Administration.This means that these subcounties were not able to access thelocal development grant in the following year from the LocalGovernment Development Programme.
This was widely publicised within the district, and the publicdid not like it. In the 2001 local government elections, allthose leaders of subcounty councils who presided over failingsubcounties were voted out of office. Incentive enough toperform?
37
Uganda case study
ensuring that services are delivered more equitably andsustainability.
One assessment method used elsewhere, which couldbe particularly relevant in the water sector, is called thebalanced scorecard technique (Estis, 1998). The technique,which is based on corporate methods which evolved inthe 1980s and 1990s, involves identifying objectives,measuring four aspects of performance and scoringperformance accordingly. See Table 30, with ideas for theUgandan water sector.
Under each performance area there is a goal and set ofperformance measures. In the Ugandan context, districts’performance in each of the performance areas would bedocumented and scored, using the performance measuresidentified. The performance measures would be morecomprehensive than the 5 golden indicators8 and wouldattempt to describe each goal in each performance area.
This technique could be used by district water officesto carry out assessments of subcounty performance. TheDWD, using a combination of staff from regional technicalsupport units and headquarters, could then carry out anassessment of districts using a similar technique, verifyingsome of the subcounty assessments that will have takenplace, along the lines of the existing LGDP assessment.
The resulting scores from districts would help identifygood and bad practice in local governments. It wouldenable district water offices to target technical support toindividual subcounties which are performing poorly.Similarly it would allow the Directorate of WaterDevelopment to differentiate the types of technical supportit provides to districts.
Most importantly it would also address the incentives ofdistricts to adhere to national priorities. Publishingperformance at the district and national levels would helpbuild incentives for politicians to adhere to sector policiesand guidelines, and use factors such as equity and
sustainability in the planning process.
This could also be linked to the size of development grantsa given district can access. For instance a minimum conditionfor accessing capital grant funds could be collection ofcommunity capital contribution, and demonstrating certainlevels of functionality. This would ensure more appropriatebalance between capital and recurrent investment byencouraging politicians and technocrats to ensure that factorssuch as equity, sustainability and sanitation are given priorityby local governments in advance of new investments.Ultimately incentives should be geared towards encouraginglocal governments to make decisions which result in theefficient and effective delivery of sector goals.
Uganda has achieved much in its reform of the waterand sanitation sector, and these must be recognized.National systems for planning and financing the sectorhave been established, alongside modalities fordecentralized service delivery. Coordination has improvedwithin the sector with common systems for programmingand reviewing sector performance through a Sector- WideApproach. The Sector is truly a high national priority andthis is borne out in the large increases in governmentbudget allocation and the integration priority lent to thesector in the PRSP.
However the study highlighted significant problems inthe equity and sustainability of rural water and sanitationservice delivery, which need to be addressed if national PRSPgoals for the sector are to be achieved. It has alsodemonstrated that these problems could be overcome byrelatively straightforward improvements to planning andmonitoring and evaluation in local governments. The useof tools for assessing equity and sustainability, using WPD,GIS mapping and the ‘sustainability snapshot’, could sharpenplanning decisions and focus monitoring and evaluation.This, combined with a district performance assessmentsystem which covers other aspects of performance such asefficiency, should help align political and administrativeincentives behind the achievement of sector and PRSP goals.
38
Implementation of Water Supply & Sanitation programmes under PRSPs
VI. Conclusions andrecommendations
Factors behind poor integration of WSS withinPRSPsThe research highlighted a number of important factorsbehind current relatively poor integration of WSS ascompared with other sectors:
Progress in sector reforms – in general health andeducation sectors are further ahead in the developmentof SWAPs than WSS. Institutional fragmentation whichmanifests itself in plans produced by different ‘sub-sectors’(rural WSS, urban WSS, water resources managementetc.) which are not planned in coordination and notmutually supporting is an obstacle in the water sector.In Uganda, by adopting themes not already ‘owned’ byone particular agency, some of the usual divisionsbetween sub-sectors were avoided. This has enabled someprogress in moving towards a SWAP which has in turnmeant WSS has been better able to engage with PRSPprocesses.
Political commitment to poverty reduction and budgetaryreforms – greater political commitment to povertyreduction in Uganda provided a greater incentive forindividual sectors to align themselves with the PRSP.Experience in Uganda suggests that civil society lobbyingcan contribute to building political preference for WSS.In terms of political ‘clout’, health and education sectorsare generally more powerful with larger budgets and thusin a stronger position when lobbying for budgetaryresources. Whether they have made a better case forinvestment in terms of poverty reduction, or it is simplytaken for granted that such expenditure is povertyreducing, is not clear.
Greater on-budget funding – a greater proportion offunding available for education and health comes fromgovernment revenue, whereas WSS has historicallydepended heavily on project funding, i.e. external tonational revenues. In Uganda there has been a shift infunding available for WSS from projects towardsprogramme-based support. This has resulted in strongerlinks to the MoF and greater incentive to engage in thebudgetary process. In contrast the WSS sector in Zambiaand Malawi remains heavily dependent on multiplefragmented donor projects.
Strengthening sectoral processesThese findings point to the need for ‘systemic’strengthening of sectoral processes if confidence is to becreated in programmes led by government, rather thanindividual projects selected by donors, and there is to beconsolidation of poverty-reduction via PRSPs. Not onlyare sector plans poorly aligned with wider povertyreduction strategies but decision making processessurrounding the allocation of resources within the watersector are currently often weak. There is evidence inMalawi and Zambia that weak sectoral planning is causingdonors to bypass sectoral programmes and opt for the
‘old’ scheme of project-by-project funding. Whilst donorfunding of ‘off-budget’ projects/programmes runs generallycontrary to the philosophy of PRSPs, in the current stateof preparedness of the water and sanitation sector (at leastthree of the four study countries), budget support seemsat present an unattractive proposition – except to the extentit is specifically directed towards remedying the ‘system’failures outlined in this report.
At the same time, in-country, better sector plans andimplementation mechanisms are required for the waterand sanitation sector to make its case for public resourcesmore forcibly, whether from government or donors. InMalawi, the water sector is failing to attract and retainnational, public funds from the Ministry of Finance andthis failure is (in part at least) due to existing water strategiesbeing inadequately planned and costed as compared withthose in other sectors – which are better able to maketheir case for funds.
The problem in Malawi is symptomatic of the failure ofstakeholders to mobilise as a sector in response to the PRSPprocess which was largely separate and de-linked fromsector review processes. As a result, sectoral policies havenot been sufficiently represented and championed andthere is little sense of sectoral ownership over the resultingPRSP. From a WSS perspective, therefore, the process has,currently at least, lost momentum.
Under existing PRSPs WSS is categorised variouslyunder ‘infrastructure’ or ‘social/human’ pillars but itscontribution to wider economic growth and socialdevelopment objectives is generally poorly understood orarticulated. The water and sanitation sector needs to addressitself to mainstream concerns for promoting broad basedeconomic growth. Further research is required to improveour understanding of the impact of WSS investment ongrowth and specifically how it can best be combined withinvestment in other sectors to maximise poverty impact.In Tanzania, for example, methods for analysing the benefitincidence of water sector investments are currently beingdeveloped (de Waal, 2004). Further development ofanalytical tools and support for this kind of detailed sectoranalysis is essential if the poverty-reducing potential offuture WSS investment is to be realised.
Joint Sector ReviewsThe prospects in Uganda look better, thanks to the JointSector Review process which has made considerableadvances over the last three years and has succeeded ingenerating a momentum for sector reforms which issupported by government, donors and other stakeholdersin the water sector.
The original motivator of sector review was the Ministryof Finance. The key to generating and maintaining thedynamic has, it seems, been to focus on relatively few high-level policy objectives (nine, grouped under five themes),
39
Conclusions and recommendations
In Malawi research shows that donor agencies arebypassing central government and funding districts directly,resulting in multiple fragmented projects with diverseobjectives poorly aligned with decentralised planningprocesses and priorities. Such donor behaviour activelyundermines the capacity of local government to coordinateand plan sectoral development and lines of accountabilityare blurred as beneficiary populations are unable to holdeither donors or government to account.
Civil society groups have a potentially important role toplay in monitoring sectoral developments and lobbyinggovernment to give WSS greater priority. They also have arole to play in strengthening co-ordination within the sector.In Zambia, a multi-stakeholder forum for rural water supplyand sanitation, which grew out of the involvement of itsmembers in the PRSP process, has been established by theMinistry of Local Government with the active support ofWaterAid. In Malawi, WaterAid is promoting Water PointDensity mapping as a tool for improved planning andcoordination between different sector agencies at nationallevel. It is also supporting local authorities to map the equityof water point distribution in their areas. The resulting mapsenable district governments to plan and request resourceson the basis of clear evidence, and to resist political patronagein the allocation of their limited resources for water andsanitation provision.
Improving performance through monitoringequity and sustainabilityThe research has demonstrated the utility of simple toolsfor planning and monitoring progress in sectoral develop-ment, notably WPD mapping and the ‘sustainabilitysnapshot’. The Uganda research suggests that a similar‘snapshot’ technique could be used for monitoringsanitation investments. These rapid low-cost techniquescould make a significant contribution towards improvedtransparency and accountability and form a basis forperformance monitoring either within government orby donors or independent civil society groups.
The Uganda study highlighted significant problems inthe equity and sustainability of rural water and sanitationservice delivery, which need to be addressed if nationalPRSP goals for the sector are to be achieved, but it alsodemonstrated that these problems could be overcomeby relatively straightforward improvements to planningand monitoring and evaluation in local governments (eg.against a limited set of standard indicators to be measuredin all districts). The use of tools for assessing equity andsustainability combined with a district performanceassessment system which covers other aspects ofperformance such as efficiency, should help align politicaland administrative incentives behind the achievement ofsector and PRSP goals. However, this is only likely to beeffective in the context of a sustained commitment toaddressing capacity and resource constraints currentlyfacing local governments.
instead of trying to manage a long list of disparate actionpoints at a technical level. Also, to choose cross-cutting or‘intra-sectoral’ themes instead of the habitual sub-sectors(e.g. rural WSS, urban WSS, water resources managementetc.), with the themes being of relevance and interest to allstakeholders, without one sub-sector being able to claimthat they ‘own’ the agenda.
The bi-annual review meetings, in March andSeptember, have now established themselves as key datesin the sectoral calendar, both in-country and for visitingmissions (e.g. World Bank and African DevelopmentBank). Alongside the big sector forum – comprising some200 representatives of all institutions and organisationsinterested in the sector – which deals with the policyissues, there are sub-groups taking forward the technicalagenda in relation to each of the five themes. Thedecision-making body is a sector working groupcomprised of 20-25 members (government, donor andothers) which acts as formal governance body. Themeetings of the large forum have been facilitated byexternal agencies (with careful management of the agenda,with submission of papers in advance and focusedinterventions) and the substantial costs of managing theprocess supported with donor funds. Donor funds havealso been used strategically to fund timely studies designedto improve understanding of the blockages to effectiveand efficient sector performance. This has enabled therespective government decision-making agencies and thesector stakeholders to prioritise and target improvementsin delivery mechanisms or decision-making systems moreeffectively.
Coordinated and predictable donor supportAccording to one active participant, consistent donorparticipation has been an essential ingredient in the UgandaJoint Sector Review with a donor working groupestablished to discuss and agree common positions on keystrategic issues, i.e. far beyond an initial function ofinformation-exchange. If, as in Uganda, the governmenttakes the lead in expressing the desire to develop coherentsector policies, strategies and plans, this provides impetusto donors to coordinate themselves. Once a critical massof donors interested in a sector-wide approach wasestablished, with a rotating lead donor, other donors wereinvited to coordinate, and project-based finance becameless likely to be ‘parachuted’ into the country without anylink to sector plans.
In Zambia, donors provide 70-80% of allocations to WSS,and are therefore in a strong position to drive throughsimilar sectoral and budgetary reforms. According to theproject research, on-budget donor support is up to 70% oftotal spending in the sector, and including off-budgetfinance, it could reach as much as 80%. With this leverage,some of the donors have been pushing for sector reformsbut in the continued absence of genuine politicalcommitment progress is likely to remain slow.
40
Implementation of Water Supply & Sanitation programmes under PRSPs
VII. References
Annamraju S., Calaguas B. and Gutierrez, E. (2001) ‘Financingwater and sanitation: key issues in increasing resources tothe sector’, WaterAid Briefing Paper, London: WaterAid
Booth, D and Lucas, H (2001) ‘Initial review of PRSPdocumentation: Desk study of good practice of PRSPindicators and monitoring systems’, report commissionedby DFID for the Strategic Partnership with Africa, by ODI,and Institute of Development Studies
Booth, D. and Lucas, H. (2002) ‘Good practice in the developmentof PRSP indicators and monitoring systems’, ODI WorkingPaper No. 172, London: ODI
Booth, D. ed. (2003) Fighting poverty in Africa: Are PRSPs makinga difference?, London: ODI
Calaguas, B, Gessesse, H and Hussain, S (2001) ‘Poverty ReductionStrategy Papers Processes: a coordinated learning andadvocacy programme’, London: WaterAid UK
de Waal, D. (2004) ‘Strengthening design, finance and delivery ofWSS programmes under PRSPs’, Presentation at KampalaWorkshop, February 2004, unpublished
DFID (2001) ‘Addressing the water crisis: Healthy and moreproductive lives for poor people; strategies for achieving theinternational development targets’, London: DFID
Estis, A (1998) ‘The balanced scorecard – Applying a privatesector technique to the public sector’, Paper presented atthe 1998 Conference of the Association for Public PolicyAnalysis and Management
Evans, A. (2002) ‘PRSPs – Emerging issues and lessons – thePRSP monitoring and synthesis project’, Presentation toODI, December 2002
Foster, F., Fozzard A., Naschold F. and Conway T. (2002) ‘How,when and why does poverty get budget priority; Povertyreduction strategy and public expenditure in five Africancountries (Uganda, Ghana, Tanzania, Malawi, Mozambique):Synthesis Paper’, ODI Working Paper No. 168, London: ODI
Fozzard A. (2001) ‘The basic budgeting problem: Approaches toresource allocation in the public sector and their implicationsfor pro-poor budgetting’, Centre for Aid and PublicExpenditure, ODI Working Paper No. 147, London: ODI
Government of Malawi (2002) ‘Malawi Poverty ReductionStrategy Paper’, final draft, April 2002, Ministry of Financeand Economic Planning
Government of Zambia (2002) ‘Zambia Poverty ReductionStrategy Paper 2002-2004’ Ministry of Finance and NationalPlanning
Gutierrez, E., Mwumbwa, S. and Wake, W. (2004) ‘The case offinancing rural water supply and sanitation in Zambia’sPoverty Reduction Reduction Strategy’, March 2004WaterAid Zambia Research Report, unpublished (draft atwww.wateraid.org)
Kanyesigye J., Anguria J., Niwagaba E., and Williamson T. (2004)‘Are national water and sanitation objectives being achievedon the ground? A review of service delivery, planningmonitoring & evaluation in Tororo and Wakiso districts’,February 2004 WaterAid Uganda Research Report,unpublished (draft at www.wateraid.org)
Mehta M. (2001) ‘Water supply and sanitation in PRSP initiatives- A desk review of emerging experience in sub-SaharanAfr ica’, World Bank/UNDP Water and SanitationProgramme, Nairobi: WSP-Africa
Mehta M. and Fugelsnes T. (2003) ‘Water supply and sanitationin Poverty Reduction Strategy Papers in sub-Saharan Africa:Developing a benchmarking review and exploring the wayforward’, Nairobi: WSP-Africa
MLWE/DWD (2000) ‘Water sector reform rural water andsanitation component. strategic investment plan 2000-2015’,Ministry of Lands Water & Environment/Directorate of WaterDevelopment, Government of Uganda
Mwanawina, I., Akapelwa, M., Sampa, KJ. and J. Moonga (2003)‘Afr ica fiscal transparency br ief: Zambia’ http://www.internationalbudget.org/resources/ZAMBIA.pdf .Downloaded November 2003
MWLE (2003) ‘Measuring performance for improved servicedelivery’, Government of Uganda, Ministry of Water, Landsand Environment, September 2003
MWLE (2003) ‘Measuring Performance for improved servicedelivery’, Ministry of Lands Water & Environment,Government of Uganda
MWLE/DWD (2002) ‘Guidelines for the planning and theoperation of district water and sanitation development grant’,July 2002, Ministry of Lands Water & Environment/Directorate of Water Development, Government of Uganda
Newborne P. (2004) ‘Water and poverty reduction’, Report ofreview of Water & PRSPs in ten countries, commissioned byWWF Living Waters, ODI, March 2004 available onwww.odi.org.uk/rpeg/wpp
Nicol, A, (1997) ‘Water projects and livelihoods: Poverty impactin a drought-prone environment’, Workshop Proceedingsorganised by Save the Children-UK, Harare, Oct 1997
Norton, A. and Elson, D. (2002) ‘What’s behind the budget?Politics, rights and accountability in the budget process’, ODIWorking Paper No.148, London: ODI
Nyirenda, N. (2004) ‘WatSan & PRSPs: Flow of Funding fromCentral Government’, Report prepared for WaterAid Malawi,unpublished
ODI (2002) ‘Poverty reduction and water: WatSan & PRSPs insub-Saharan Afr ica’, ODI Water Policy Brief No. 3,www.odi.org.uk/rpeg/wpp, London: ODI
ODI (2004) ‘From plan to action: Water supply and sanitationfor the poor in africa’, ODI Briefing Paper, April 2004, London:ODI
ODI (2004b) ‘Why budgets matter: The agenda of publicexpenditure management’, ODI Briefing Paper, May 2004,London: ODI
ODI and WaterAid (2002) ‘Water & PRSPs: Integrating WatSanactivities within PRSP development and implementation’,Inception Report of DFID-funded project, June 2002, ODIWater Policy Programme, www.odi.org.uk/rpeg/wpp
OPM and ODI (2002) ‘General budget support evaluability Study,Phase I’ Final Synthesis Report, to DFID, 30th December2003
Piron, LH. & Evans, A. (2004) ‘Politics and the PRSP approach –Synthesis paper’, ODI Working Paper No. 237, London: ODI
Piron, LH. & Norton, A. (2004) ‘Politics and the PRSP approach– Uganda case study’, ODI Working Paper No. 240, London:ODI
Sugden et al, (2004) ‘Report of research into the financial flows,resource allocation processes and impact of investments madein the water supply and sanitation sectors in Malawi’,WaterAid Malawi Research Report, unpublished (draft atwww.wateraid.org)
Sugden S. (2003) ‘What is Going Out There?’ Paper developedfor 3rd World Water Forum Kyoto Japan, London: WaterAidUK
Sugden S. and Stoupy, O. (2003) ‘Halving the number of peoplewithout access to safe waters by 2015 – A Malawianperspective’, WaterAid Malawi
41
Conclusions and recommendations
Endnotes
1 There have been budget allocations for rural water supply inprevious years, but it was only in 2001 when it was entitled‘Poverty Reduction Programme’.
2 Note the origin of a large proportion of the water supply is‘unknown’. The information was collected as part of the processof collating a district water point inventory, which involvedasking community members who had installed the water pointfor them. If they were unclear, the answer was recorded‘unknown’. Even with this proviso, it can be clearly seen thatthe vast majority of the water points have been installed outsidethe DDP budget.
3 In a major ‘benchmarking’ review of progress towards
Thompson, M. (2004), ‘Strengthening budget mechanisms forsanitation financing in Uganda’, Presentation to KampalaWorkshop, February 2004
WaterAid Uganda (2003) ‘Poverty reduction and water access insub-Saharan Africa: Uganda case study’, unpublished
Williamson (2003a) ‘Factors behind Poor Integration of the Waterand Sanitation Sector in PRSPs in sub-Saharan Africa’,Forthcoming, WSP-Africa
Williamson, (2003b) ‘Targets and Results in Public SectorManagement – Uganda Case Study’, ODI Working Paper no.205, London: ODI.
World Bank (1998) Public expenditure management handbook,Washington D.C. : World Bank
WSP Africa (2002) ‘Workshop proceedings for the regionalworkshop on WSS in PRSPs’, WSP-Africa, unpublished
incorporation of WSS in PRSPs in twelve African countries,carr ied out by WSP-Afr ica, Uganda showed the bestperformance in ‘poverty diagnostics, sector reform, monitoringand evaluation and sector financing’.
4 In both districts urban subcounties with access to piped waterschemes were excluded.
5 Source: Participatory Rural Appraisal tool to identify socialcharacteristics of an area
6 As described in MWLE-DWD 20007 MWLE 2003, ‘Measuring Performance for improved service
delivery’8 These could be used to assess overall ‘Achievement of Mission’
42
Implementation of Water Supply & Sanitation programmes under PRSPs
Annex
Figure 15: Improved Community Points Density Mapping – Salima District
