Embed Size (px)
Citation preview
Putting Aid On Budget
SyntheSiS RePORt
April 2008
Putting aid on budgetSynthesis report
APRIL
2008
iii
cOntentS
Acknowledgements vii
Acronyms and abbreviations viii
1. Introduction 1
1.1 Terms of reference 1
1.2 Approach and outputs 1
1.3 Structure of the Synthesis Report 2
2. Basic concepts 4
2.1 Introduction 4
2.2 Aid 4
2.3 The different dimensions of on budget 5
2.4 Disbursement channels 6
2.5 Complex patterns of aid capture 6
2.6 Misconceptions to avoid 6
3. Why should aid be on budget? 9
3.1 Introduction 9
3.2 Reasons for capturing aid 9
3.3 Aid and the budget cycle 10
3.4 International standards and targets 11
4. How much aid is on budget? 16
4.1 Introduction 16
4.2 Survey on the implementation of the Paris Declaration 16
4.3 Consolidated data on PFM performance 17
4.4 Conclusions from the data 21
4.5 Recommendations for monitoring aid on budget 25
5. Approach to the case studies 29
5.1 Introduction 29
5.2 Stakeholder perspectives and incentives for aid capture 29
5.3 The case studies 32
iv
6. Lessons from the case studies 35
6.1 Introduction 35
6.2 Putting aid on plan 35
6.3 Putting aid on budget 39
6.4 Putting aid on parliament 42
6.5 Putting aid on treasury 44
6.6 Putting aid on accounting 45
6.7 Putting aid on audit 47
6.8 Putting aid on report 48
6.9 Decentralisation 50
7. Conclusions and recommendations 51
7.1 Introduction 51
7.2 Overall approach to improving aid capture 51
7.3 Recommendations for governments 53
7.4 Recommendations for donors 55
7.5 Recommendations for joint action 55
7.6 Next steps 55
Bibliography 56
Annexure: Summary of Country Study Findings 59
Acronyms and abbreviations 62
Introduction 63
Group A country studies 65
Burkina Faso 65
Ethiopia 67
Kenya 72
South Africa 76
Tanzania 81
Group B country studies 85
Ghana 85
Mali 89
Mozambique 94
Rwanda 100
Uganda 105
v
BOxeS
Box 1: The Putting Aid On Budget Literature Review 2
Box 2: Paris Declaration commitments cover all aid to the government sector 4
Box 3: Different dimensions of on budget/capturing aid 5
Box 4: Complex patterns of aid capture in Tanzania 7
Box 5: Aid capture and the budget cycle 11
Box 6: Overview of the PEFA indicator set 12
Box 7: Paris Declaration commitments to putting aid on budget 14
Box 8: Paris Declaration indicators and targets for putting aid on budget 15
Box 9: Paris Declaration monitoring data for case study countries 18
Box 10: Donor ratios for relevant indicators for putting aid on budget 19
Box 11: Consolidated country ratings 20
Box 12: Paris Declaration monitoring data: Country results for indicators 3 and 5a 21
Box 13: PEFA performance indicator scores across countries 22
Box 14: Paris Declaration survey: Budget support and on budget data 24
Box 15: PFM quality versus the use of PFM systems 25
Box 16: Selected available PEFA scores in rank order 26
Box 17: Sharing information and using government systems 30
Box 18: Interlocking incentives in traditional aid 31
Box 19: Examples of perverse incentives 32
Box 20: ODA: Basic data for the case study countries 32
Box 21: Factors behind quality of aid capture in the budget 34
Box 22: Paradoxical aid projections (Uganda illustration) 37
Box 23: Putting aid on plan: Realistic projections and scenario planning 38
Box 24: Engaging donors in medium-term planning 38
Box 25: Deviations between budgets and actual spending 40
Box 26: Sector budget support in Mali 41
Box 27: Integrating recurrent and development budgets: Rwanda’s approach 41
Box 28: Bringing aid on treasury through the single treasury account in Mozambique 45
Box 29: Tanzania: Aid-in-kind and dummy entries 47
Box 30: Good practice: Follow international accounting standards for aid (IPSAS) 47
Box 31: The ODAMOZ database in Mozambique 49
Box 32: Decentralised services on budget in Uganda 50
Box 33: Ghana: Preparing a comprehensive aid strategy 54
vi
the Putting Aid On Budget Study
The Collaborative Africa Budget Reform Initiative (CABRI) and the Strategic Partnership with Africa (SPA) commissioned a comprehensive study of putting aid on budget. The study has the following outputs:
An • inception report (Putting Aid On Budget Inception Report), which defines the issues and research methodology.
ten country studies from Sub-Saharan Africa.• Of the ten country studies, Ghana, Mali, Mozambique, Rwanda and Uganda were studied in depth, and separate country reports are available. The experiences of Burkina Faso, Ethiopia, Kenya, South Africa and Tanzania were also reviewed and summary information is included in the Synthesis Report annexes. Findings from all ten countries are included in the Synthesis Report.
A • literature review (Putting Aid On Budget Literature Review), which: documents existing good practice guidance that is relevant to the incorporation of aid in recipient country budgets; reviews the policies and guidelines of the major multilateral and bilateral agencies as these affect the incorporation of their aid into government budgets; and documents relevant experiences of efforts to capture aid in government budgets, including desk reviews of some additional countries, including countries from outside Africa.
A • synthesis report (Putting Aid On Budget Synthesis Report), which draws on all the other study components to develop overall findings and recommendations.
A • good practice note (Good Practice Note: Using Country Budget Systems), which distils the lessons of the study and is aimed at donors as well as partner governments.
Full reference details are given in the bibliography.
The reports can be downloaded from the CABRI website at http://www.cabri-sbo.org
vii
AcknOwledgementS
This report was jointly commissioned by the Collaborative Africa Budget Reform Initiative (CABRI) and the Strategic Partnership with Africa (SPA) and financed by the UK Department for International Development (DFID), with additional support from the Japan International Cooperation Agency (JICA) for the Putting Aid On Budget Literature Review. The report was prepared by an independent study team from Mokoro Ltd. The team was led by Stephen Lister and included: Mary Betley, Rupert Bladon, Rebecca Carter, Mailan Chiche, Alta Fölscher, Piet Lanser, Alex Warren-Rodriguez and Tim Williamson. Support to the study was provided by Philip Lister and Daniel Ross.
CABRI, SPA and the study team would like to express their gratitude to the senior government budget officials of Burkina Faso, Ethiopia, Kenya, Ghana, Mali, Mozambique, Rwanda, South Africa, Tanzania and Uganda for their valuable input into the study. CABRI, SPA and the study team would also particularly like to thank Aarti Shah and Neil Cole of CABRI for their valuable inputs, and Peter Dearden of DFID and Jiro Otsuka of JICA for their support during the study.
The report was prepared in consultation with aid agencies and other organisations, including: the African Development Bank (AfDB), the Asian Development Bank (AsDB), the German Federal Ministry for Economic Cooperation and Development (BMZ), the Canadian International Development Agency (CIDA), the Danish International Development Agency (Danida), DFID, the European Commission (EC), the German Agency for Technical Cooperation (GTZ), the International Monetary Fund (IMF), the Inter-American Development Bank (IDB), Irish Aid, JICA, the Netherlands, the Norwegian Agency for Development Cooperation (Norad), the Organisation for Economic Co-operation and Development’s Development Assistance Committee (OECD-DAC), the Public Expenditure and Financial Accountability (PEFA) Secretariat, the Swedish International Development Cooperation Agency (Sida), the United Nations Development Programme (UNDP) and the United States Agency for International Development (USAID). The cooperation and support of staff of these institutions is gratefully acknowledged.
viii
liSt OF AcROnymS
AcROnym deScRiPtiOn
AAP assessment and action plans
AfDB African Development Bank
AsDB Asian Development Bank
AIMS aid information management system
AMP Aid Management Platform (Ethiopia)
BF basket funding
BS budget support
BMZ German Federal Ministry for Economic Cooperation and Development
CABRI Collaborative Africa Budget Reform Initiative
CIDA Canadian International Development Agency
CIPFA Chartered Institute of Public Finance and Accountancy
CPIA country policy and institutional assessment
DAC Development Assistance Committee (of the OECD)
DFID Department for International Development (UK)
EC European Commission
GBS general budget support
GNI gross national income
GTZ German Agency for Technical Cooperation
HIPC Heavily Indebted Poor Countries
IDA International Development Association
IDB Inter-American Development Bank
IFAC International Federation of Accountants
IFMIS integrated financial management information system
IMF International Monetary Fund
IPSAS International Public Sector Accounting Standards
IPSASB International Public Sector Accounting Standards Board
JICA Japan International Cooperation Agency
LGDP Local Government Development Programme
MCC Millennium Challenge Corporation
MDG Millennium Development Goals
ix
MDAs ministries, departments and agencies
MDBS multi-donor budget support
MTEF medium-term expenditure framework
Norad Norwegian Agency for Development Cooperation
OECD Organisation for Economic Co-operation and Development
ODA official development assistance
ODI Overseas Development Institute
PBS Protection of Basic Services (Ethiopia)
PEFA Public Expenditure and Financial Accountability
PER public expenditure review
PFM public finance management
PIU project implementation unit
PRS poverty reduction strategy
PRSP poverty reduction strategy paper
RDP Reconstruction and Development Programme (South Africa)
SBS sector budget support
SBAS Strategic Budget Allocation System (Tanzania)
Sida Swedish International Cooperation Agency
SPA Strategic Partnership with Africa
SSA Sub-Saharan Africa
STA single treasury account
SWAp sector wide approach
TA technical assistance
TOR terms of reference
UN United Nations
UNDP United Nations Development Programme
USAID United States Agency for International Development
x
1
1
intROductiOn
This chapter explains the objectives of this study, describes the study components that have fed into this Synthesis Report, and outlines the structure of the Synthesis Report.
1.1 terms of reference (tOR)
The Paris Declaration has put the use of government systems at the centre of the international aid effectiveness agenda. Several of its indicators relate directly or indirectly to putting “aid on budget”, which has thus gained prominence as a joint concern of governments and donors.
Against this background, the Collaborative Africa Budget Reform Initiative (CABRI) and the Strategic Partnership with Africa (SPA) commissioned the present study on putting aid on budget. In essence, the study’s aim is:
to produce outputs which will better equip governments in Sub-Saharan Africa to lead country-level processes to ensure external development assistance (aid) flows are properly reflected in national budget documents, ex ante (budget presented to legislature) and ex post (out-turn accounts).
While the study is particularly intended to help Sub-Saharan African (SSA) countries that receive substantial aid flows, its scope is broader. It is one of the first attempts to examine the issue of putting aid on budget in a more systematic way, drawing on evidence from a wide range of countries. The TOR required a literature review of existing good practice, a study of country practices in at least ten African countries, a more in-depth investigation of what works and what does not in some of the countries that were subject to case studies, and the preparation of a synthesis report and a good practice note based on the research results.
1.2 Approach and outputs
The study methodology is set out in the Putting Aid On Budget Inception Report and is strongly reflected in Chapter 2 of this report (“Basic concepts”).
Case studies of ten SSA countries formed the core of the study. The Annex to this report (Summary of Country Study Findings) assembles the main findings from all ten countries, and is a separate publication. A standard factual analysis was conducted in all countries, with more in-depth studies carried out in five (those listed as Group B below). Free-standing reports on the Group B countries are also being published. These look more closely at the reasons behind the levels of aid capture that were found. (See the Bibliography1 for details of all the study outputs.)
1 At the end of this report.
2
cASe StudieS
Group A Group B
Burkina Faso Ghana
Ethiopia Mali
Kenya Mozambique
South Africa Rwanda
Tanzania Uganda
The case studies did not involve a new quantitative survey. In each country, researchers assembled the best available information from existing sources2 and focused especially on trying to explain the patterns of aid capture that were found. All the researchers were already familiar with the countries concerned, and they were assisted by the involvement of the senior budget officers linked to CABRI and key donor representatives. Drafts of the country findings were circulated to government and aid agency representatives for comment.
The research was supported by the Putting Aid On Budget Literature Review, which has been published separately. Box 1 shows the main requirements for the review and how they were addressed.
Box 1: The Putting Aid On Budget Literature Review
The Putting Aid On Budget Literature Review was required to:
• documentexistinggoodpracticeguidancethat isrelevanttotheincorporationofaid inrecipientcountry budgets
Section A looks at the international good practice guidance for putting aid on budget and documents the literature on: sound budgeting and financial management and the implications for aid management; the aid effectiveness consensus and the good practice principles for putting aid on budget; and how aid on budget is monitored.
• reviewtheapplicablepoliciesandguidelinesofthemajormultilateralandbilateralagenciesastheseaffect the incorporation of their aid into government budgets
Section B provides illustrations of different donor approaches to putting aid on budget. The bilateral agencies covered are: Canada, Denmark, France, Germany, Ireland, Japan, the Netherlands, Norway, Sweden, the United Kingdom and the United States. The multilateral organisations covered are: the African Development Bank (AfDB), the Asian Development Bank (AsDB) and the Inter-American Development Bank (IDB), the World Bank, the International Monetary Fund (IMF), the European Commission (EC) and the United Nations (UN).
• seekanddocumentrelevantexperiencesofeffortstocaptureaidingovernmentbudgets,focusingmainly on countries that were not case study countries in the main putting aid on budget consultancy, including countries from outside Africa
Section C supplements the core case studies by profiling six other countries: the African region: Botswana and Senegal; the Asian region: Bangladesh and Vietnam; and the Latin American region: Bolivia and Nicaragua.
1.3 Structure of the Synthesis Report
An essential starting point is for there to be clarity about concepts and definitions. Chapter 2 (“Basic concepts”) therefore highlights the multiple dimensions of “aid on budget”, and the need for precision in the use of this and other terminology. This is closely linked to Chapter 3’s (“Why should aid be on budget?”) review of the rationale for putting aid on budget. This provides the context and criteria for assessing the quality of aid capture. Chapter 3 also notes the most relevant donor diagnostics, good practice guidance and targets.
2 See Chapter 4 for a discussion on available quantitative and standardised assessments.
3
As noted earlier, this study did not undertake a country-by-country quantitative survey. Nevertheless, Chapter 4 (“How much aid is on budget?”) reviews the available evidence for the extent of aid capture at various stages in the budget process, and for the quality of the public finance management systems involved. This provides an essential background for the case studies.
Chapter 5 (“Approach to the case studies”) explains the research methodology, including the choice of case study countries and the importance of understanding the perspectives and interests, and hence the incentives, of various stakeholders involved with aid and budgeting. The effective capture of aid on budget is more than a technical matter: it requires a good understanding of the institutional setting in which it takes place.
Chapter 6 (“Lessons from the case studies”) highlights the main findings and conclusions from the country case studies. These are systematically presented, country by country, in the Annex (Summary of Country Study Findings), and Chapter 6 therefore focuses on the main factors that help or hinder efforts to ensure that aid is usefully captured during the budget cycle.
Finally, based on these findings, Chapter 7 (“Conclusions and recommendations”) offers good practice recommendations for governments and aid agencies.
4
2
BASic cOncePtS
This chapter explains the concepts and terminology used throughout the report, identifies the different dimensions of “on budget”, and draws attention to some common areas of confusion.
2.1 introduction
Discussions about putting “aid on budget” are easily undermined by inconsistencies in the use of terms and concepts. The Putting Aid On Budget Literature Review highlights different interpretations among donors.3 (See also the discussion in Chapter 5 of this report.) The review also notes that the lack of consensus on practical definitions was one of the main challenges facing the first survey on the implementation of the Paris Declaration. In this chapter, therefore, the principal concepts and definitions used in this study are explained.
2.2 Aid
For the purposes of this study, “aid” means official development assistance (ODA), as defined by the OECD’s Development Assistance Committee (DAC).4 ODA includes both grants and (soft) loans, as well as contributions in cash and in kind. Distinguishing between these different types of aid is important, as the use of particular forms of ODA affects the likelihood of aid being captured in budgets.
The relevant Paris Declaration indicators and targets cover all aid to the government sector. (See Box 2.)
Box 2: Paris Declaration commitments cover all aid to the government sector
The Paris Declaration commitment to putting aid on budget covers all aid that is disbursed for the government sector, regardless of how the aid is executed. The OECD provided the following definition of “disbursements for the government sector” that were monitored for the aid on budget targets for the 2006 Survey on Monitoring the Paris Declaration:
The disbursement of ODA in the context of an agreement with the government sector including works, goods or services delegated or subcontracted by government to other entities (e.g. NGOs, private companies).The government sector is defined as: administrations – ministries, departments, agencies or municipalities – authorised to receive revenue or undertake expenditures on behalf of central government.
Source: OECD Development Assistance Committee, 2006b.
3 For the sake of brevity the term “donor” is used to cover all aid agencies providing ODA, including ODA loans/credits as well as grants.4 “Grants or Loans to countries and territories on Part I of the DAC List of Aid Recipients (developing countries) which are: (a) undertaken by the official sector;
(b) with promotion of economic development and welfare as the main objective; (c) at concessional financial terms [if a loan, having a Grant Element (q.v.) of at least 25 per cent]. In addition to financial flows, Technical Co-operation (q.v.) is included in aid. Grants, Loans and credits for military purposes are excluded.” (Development Co-operation Directorate, DAC’s Glossary, www.oecd.org.)
5
2.3 the different dimensions of “on budget”
The study is concerned with official government budgets. While at one level, it should be apparent from the budget documents whether aid has been incorporated or not, this is not entirely straightforward:
Budgets may or may not show aid as a • source of budget finance in general, and/or track the use of aid to particular expenditure lines.
In either case, presentations may be highly aggregated or quite detailed.•
There may also be differences between the presentation of a budget ex ante (before expenditure takes •place) and various ex post presentations. The final version of the budget typically takes account of adjustments during the year, and the record of actual expenditures against the budget may differ again.
Furthermore, there is rarely only one relevant budget:
Budgets frequently appear in draft, final, and amended versions.•
There may be a multiplicity of government budgets. (This is typically the case in federal or other •decentralised systems.)
Even for a central government budget, there are numerous budget holders (and time lags before •information reaches them).
In addition, budget documents need to be seen as punctuation marks in a continuing budget process. Aid may feature more or less effectively at different stages of the budget process. “Aid on budget” correspondingly has different dimensions. Box 3 defines the dimensions on which this study focuses.
Box 3: Different dimensions of on budget/capturing aid
term definition
On plan Programme and project aid spending is integrated into spending agencies’ strategic planning and supporting documentation for policy intentions behind the budget submissions.
On budgeta External financing, including programme and project financing, and its intended use are reported in the budget documentation.
On parliament (or “through budget”)
External financing is included in the revenue and appropriations approved by parliament.
On treasury External financing is disbursed into the main revenue funds of government and managed through government’s systems.
On accounting External financing is recorded and accounted for in government’s accounting system, in line with government’s classification system.
On audit External financing is audited by government’s auditing system.
On report External financing is included in ex post reports by government.Source: Putting Aid On Budget Inception Report.
a It should be clear from the context if “on budget” is meant in a more general sense than this precise definition. The term “on system” is also used for this wider sense.
This framework was applied in the country studies, and in the analysis of issues generally. It proved to be very helpful in clarifying the ambiguities which bedevil this subject, and it also provides the structure of country-by-country findings presented in the Annex to this report (Summary of Country Study Findings).
Box 3 is not exhaustive: there are other dimensions of government systems that may also be relevant. “On procurement” is an example. Externally financed expenditures may or may not follow the government’s standard
6
procurement procedures, and several of the country studies noted that the perceived quality of government procurement systems was also an important influence on the capture of aid in other dimensions.
2.4 disbursement channels
When discussing the “on treasury” dimension, it is important to be precise about different possible disbursement channels for aid. Three main channels have been identified:
Channel 1 is the normal channel used for government’s own-funded expenditures. Aid is disbursed to the •government’s finance ministry (or “treasury”), from where it goes, via regular government procedures, to the ministries, departments or agencies (MDAs) responsible for budget execution. Note that Channel 1 funds may or may not be earmarked for specific expenditures.
In Channel 2, external funds are provided directly to a particular MDA – most often a sector ministry,• 5 and managed through special accounts outside of the regular government system. Thus, Channel 2 funds, although held by a government body, do not follow the normal government procedures.
In Channel 3, expenditure is undertaken by the donor agency itself or by non-government agents on its •behalf. Assets or services are delivered to the government in kind, but government does not handle the funds itself.
It is important to note that:
Funds that are disbursed via government do not necessarily follow regular government channels, and •may or may not be “on budget”. This applies especially to Channel 2 funds, but can apply also to Channel 1.6
Funds that are not disbursed via government (Channel 3) are often completely off budget, but this need •not be the case. It is possible for these funds to be anticipated in government plans and to be recorded in government budgets and accounts.
2.5 complex patterns of aid capture
Aid may be captured on some of the dimensions defined in Box 3 above, and not on others. There are many possible permutations, and the study’s findings at country level reflect this complexity. Box 4 illustrates the pattern found in Tanzania.
2.6 misconceptions to avoid
The discussion of bringing aid on budget is often confused by a misunderstanding of the issues. Five common areas of confusion are discussed below:
“• Aid on budget” is not a synonym for budget support. To quote the definition used in the OECD DAC evaluation of general budget support:
The general characteristics of budget support are that it is channelled directly to partner governments using their own allocation, procurement and accounting systems, and that it is not linked to specific project activities. All types of budget support include a lump sum transfer of foreign exchange; differences [e.g. between general and sector budget support] then arise on the extent of earmarking and on the levels and focus of the policy dialogue and conditionality. (IDD & Associates, 2006)
Thus, budget support is, by definition, on budget.7 However, other aid modalities, including project aid, technical assistance and other forms of aid in kind can also be on budget.
5 Channel 2 funds may be administered by a project implementation unit (PIU) under the control of government.6 For example, UN system projects that follow “national execution” procedures are typically off treasury, and often also off budget in other relevant dimensions.7 However, this study found examples where budget support was not made clearly visible in budget documents as a source of finance.
7
Sector budget support, similarly, should be on budget by definition. But • the term “sector budget support” (SBS) is often used inaccurately:
Part of the standard definition of budget support (above) is that it is not linked to specific o
project activities. However, in some country cases, the term sector budget support is used to describe funding that is very tightly earmarked ex ante to specific budget lines.8
Equally, sector budget support, by definition, uses government’s own allocation, procurement o
and accounting systems. However, pooled funding arrangements that do not use regular government systems are often wrongly described as sector budget support.
• the term sector wide approach (SwAp) is not synonymous with pooled funding (or basket funds). The discussion about whether aid is on budget needs to consider specific aid instruments. However, a SWAp is not itself a specific aid instrument but a way of working at sector level. A SWAp may be supported by a variety of aid instruments: a pooled donor fund is one possibility, but a SWAp may also
8 See Putting Aid On Budget: Mali Case Study and Putting Aid On Budget: Burkina Faso Case Study.
Box 4: Complex patterns of aid capture in Tanzania
Source: Putting Aid On Budget: Tanzania Case Study.
PlAn
Support that donors already know about during budget preparation
ON: Budget support (BS) and basket funding (BF). Some projects and technical assistance (TA). OFF: Some projects, TA
ON: What was on plan (PERs, MTEF, strat plans, budget submissions) for all modalities.Budget OFF: Projects, TA that were not
on plan
PARliAment ON: What was on budget submissions for all modalities.
OFF: What was not on budget submissions for all modalities.
tReASuRyON: Channel 1:
BS, some BF and some OnBudget Projects (Cash loans and grants)
Disbursed through government systems throughaccount at BoT, then transferred to recipient agency.
AccOunt ON: Channel 1 (all BS, some BF) and someof Channel 3 (some On Budget Projects)
Audit
RePORt
ON: Some of Channel 1 (all BS, BF) andsome Channel 3 (some On Budget Projects)
OFF: Some Channel 3 (Some On and AllOff Budget Projects), some BF
ON: Channel 1 (all BS, BF) and someChannel 3 (projects)
OFF: Some Channel 3 (Some On and AllOff Budget Projects), some BF
Supportdonors
addduringyear,
allmodalities
OFF: Channel 3: Cash Grants for On and
Off Budget Projects, some BF and TADisbursed to third party (e.g. PIUs) or managed by donor
OFF: Channel 3 (On and Off BudgetProjects, TA) and some of Channel 1 (BF)
8
(and at the same time) be supported by other instruments, including projects and both general and sector budget support (OECD Development Assistance Committee, 2006b).
• Projects are not necessarily off budget. The use of parallel systems is a very common feature of externally funded projects, and bypassing country systems has often been part of the rationale for using the project modality. However, it is quite possible for project aid to be on budget at some or all of the stages of the budget cycle. It is therefore not at all helpful to imply that the use of parallel systems is an automatic feature of project aid.9 Using project terminology in this way can easily reinforce the misconception that shifting to budget support is the only way to bring more aid on budget.
• Putting aid on budget is not just a matter of donors informing governments about their aid disbursements. There is a fundamental difference between reporting aid on budget (providing information that can be included in budget documents) and integrating aid on budget (actually using government systems at the different stages of the budget cycle). Reporting and integration are not mutually exclusive activities, but reporting will not necessarily result in the same benefits as can be achieved from integration. (See additional discussion in Chapter 5, especially Box 17.)
9 Some of the country case studies provide examples of such usage. Thus the Ghana country study noted: “In Ghana, a project is considered to be a set of activities with a common objective managed by a project implementation unit and financed by resources which are kept in separate bank accounts (usually government-designated) not linked to the Consolidated Fund” (Putting Aid On Budget: Ghana Case Study). A similar usage was found in Mali. (Putting Aid On Budget: Mali Case Study).
9
3
why ShOuld Aid Be On Budget?
This chapter explains the rationale for putting aid on budget and notes the main international commitments that support putting aid on budget.
3.1 introduction
Judging whether aid is usefully captured in the budget process requires an understanding of why capturing it is considered desirable. This chapter summarises the reasons given for putting aid on budget. These are rooted in internationally accepted standards of public finance management (PFM) that apply equally to rich and poor countries and to both domestic and external resources. They are reflected in many governments’ PFM reform agendas, in criteria for assessing PFM systems, especially the Public Expenditure and Financial Accountability (PEFA) diagnostic and in international aid effectiveness commitments. (These are encapsulated in the Paris Declaration, to which developing country governments as well as their aid partners are signatories.)
3.2 Reasons for capturing aid10
The conventional wisdom about putting aid on budget draws on two interacting streams of thought – on public finance management, and on aid effectiveness. The underlying PFM arguments relate to the comprehensiveness, transparency and accountability of government financial activities. These are reinforced by aid effectiveness arguments for strengthening government systems.
Sound budgeting and financial management
There is general consensus that sound budgeting and financial management are based on the following principles:
Budgets need to be • comprehensive – include all revenues and expenditures – to enable governments to follow good macroeconomic management and promote allocative efficiency.
Fiscal • transparency – decision-makers have all relevant information – is a driving force for improving fiscal management.
Transparency is also important to ensure • accountability – decisions and their basis, results and costs are accessible, clear and communicated to the wider community – so that decision-makers are held responsible.
Aid constitutes a substantial share of public resources for many developing countries. How aid is managed is therefore an important determinant of overall public finance management. These good practice principles therefore imply that aid should be on budget.11 If aid is not on budget then the budget cannot be comprehensive,
10 See the Putting Aid On Budget Literature Review for a fuller discussion.11 Putting aid on budget as good practice was identified in the 1990s in public finance management seminal texts by Schick (1998), the World Bank (1998) and the
International Monetary Fund (1998). See the Putting Aid On Budget Literature Review for further detail.
10
there cannot be full transparency of information and decision-makers cannot be held to account. Failure to capture aid has repercussions for the effectiveness of domestic resources (their allocation is less efficient if based only on a partial view) and often leads to unsustainable patterns of expenditure.
Aid effectiveness
It is in recipients’ own interests to maximise aid effectiveness by combining aid effectively with domestic resources. And for their part, donors have both a developmental and a fiduciary interest in the quality of partner countries’ public finance management. The OECD declared in 2006:
Effective public finance management is crucial to countries making progress in reducing poverty. It is fundamental to both government performance and successful aid delivery. (OECD Development Assistance Committee, 2006b)
Both the Heavily Indebted Poor Countries (HIPC) initiative, focusing on the adequacy of government systems to allocate resources to pro-poor activities, and the poverty reduction strategy (PRS) approach, working to link priorities with domestic and external resources, have helped to focus attention on government strategic budgeting and allocative efficiency. As a result, donors have been paying more attention to the comprehensiveness and transparency of government budgets, and, as part of that, to what extent aid is and is not on budget.
One important aim of the PRS approach was to provide a framework for facilitating donor alignment with country priorities and facilitating donor use of country systems to deliver aid. Revised assessments of the effectiveness of aid in the 1990s concluded that inappropriate aid modalities had been part of the problem. Sometimes the non-capture of aid by the government accounting system was a direct consequence of donor financing arrangements when accountability to donors was ensured at the expense of transparency and accountability in the recipient country (International Monetary Fund, 1998). Donor coordination was also an issue, with the many donors and their numerous programmes and projects making it difficult to keep track of how much was being spent and on what. For governments, this was a distraction from management of their own resources and tended to undermine their capacity to formulate and implement coherent policies.
Hence, there is a general consensus today that effective aid relies on country ownership of strategy and country leadership of aid management. Good practice principles adopted at the Rome High Level Forum in 2003 and further developed at the Paris High Level Forum in 2005 are based on the principle of donor alignment with partner countries’ strategies and systems in order to ensure this country ownership. Using country systems is seen as an important way of strengthening them. It is also seen as a way of seeking sustainability, as parallel systems often wither when external funding ceases.
3.3 Aid and the budget cycle
As implied by the dimensions shown in Box 3, the alignment of aid with government systems should be seen in the context of the budget cycle. Aid may be (more or less) aligned with the government system during some stages of the budget cycle and not during others – as Box 4 illustrates in the case of Tanzania. And the rationale for aligning aid with the government systems is somewhat different for different parts of the cycle. (See Box 5.)
Whether capturing aid at any of these points will have the desired effects depends on:
the quality of the information captured, in terms of its completeness, credibility (predictability), level of •disaggregation and detail, and so on
the ability to make use of the information, which will depend on its timeliness and accessibility (who has •the information, and at what stage in the relevant process)
the quality of the information and processes relating to non-aid resources. (The benefits of capturing aid •are likely to be limited if the processes for managing domestic resources are dysfunctional.)
11
Box 5: Aid capture and the budget cycle
Stage Relevance of capturing aid
1. Strategic planning Resource mobilisation; macroeconomic discipline; integration of aid financed and government financed programmes; policy coherence; allocative efficiency.
2. Budget preparation Resource mobilisation; fiscal discipline; integration of aid financed and government financed programmes; allocative efficiency; policy and spending coherence; spending capacity assessment; financial sustainability, including links between capital and forward recurrent spending; clarity on government’s co-financing commitments; executive authorities’ decisions on trade-offs informed by view of aid; scrutiny by parliament.
3. Budget execution Timely aid disbursement, integrated cash flow projections; strengthening of operational planning; strengthening of procurement, commitment and payment systems, lower transaction costs; improved efficiency through coherent programme implementation.
4. Accounting and monitoring Strengthened accounting systems; integrated financial information; monitoring of and accountability for aid disbursement; monitoring of aid transfers to lower level agencies; comprehensive and integrated monitoring of spending programmes to identify problems early.
5. Reporting and audit Accountability for budget execution and achievement of policy results.
6. Policy review Understanding aid effectiveness, and the contribution of aid to overall results (to feed into strategic planning).
3.4 international standards and targetsStandards for PFM
Under the PEFA initiative, a standardised PFM performance measurement framework was launched in June 2005.12
The framework includes a set of high-level indicators (28 partner performance indicators and 3 donor performance indicators) which draw on a number of earlier assessment exercises and international standards. (See the Putting Aid On Budget Literature Review, Annex B, for further details.) The PEFA framework measures the operational performance of the PFM systems, processes and institutions against six core dimensions of PFM performance: credibility; comprehensiveness and transparency; policy-based budgeting; predictability and control in budget execution; accounting, recording and reporting; and external scrutiny and audit.
12 “The Framework was developed between 2003 and 2005 through extensive consultations, including with the DAC Joint Venture on PFM, a group of African PFM experts, and government representatives from Eastern Europe and Central Asia. Comments were also received from practitioners within the World Bank, IMF, other PEFA partners, government agencies and professional organizations.” (PEFA Secretariat, 2006a)
1StRAtegic PlAnning
2Budget PRePARAtiOn
3Budget executiOn
4AccOunting & mOnitORing
6POlicy ReView
5RePORting & Audit
12
The full set of PEFA indicators is shown in Box 6, with the indicators of most direct relevance to putting aid on budget highlighted. However, none of these indicators addresses one dimension of putting aid on budget exclusively or precisely. They therefore need to be interpreted with care. (See further discussion in Chapter 4 below.)
Box 6: Overview of the PEFA indicator set
A. PFm-outturns: credibility of the budget
PI-1 Aggregate expenditure out-turn compared to original approved budget
PI-2 Composition of expenditure out-turn compared to original approved budget
PI-3 Aggregate revenue out-turn compared to original approved budget
PI-4 Stock and monitoring of expenditure payment arrears
B. key cross-cutting issues: comprehensiveness and transparency
PI-5 Classification of the budget
PI-6Comprehensiveness of information included in budget documentation * NB: Coverage of aid funds or not does not explicitly enter this indicator.
PI-7Extent of unreported government operations* NB: This is the key indicator of “capture” of donor funds in budgetary reports.
PI-8 Transparency of inter-governmental fiscal relations
PI-9 Oversight of aggregate fiscal risk from other public sector entities.
PI-10 Public access to key fiscal information
c. Budget cycle
c(i) Policy-based budgeting
PI-11 Orderliness and participation in the annual budget process
PI-12 Multi-year perspective in fiscal planning, expenditure policy and budgeting
c(ii) Predictability and control in budget execution
PI-13 Transparency of taxpayer obligations and liabilities
PI-14 Effectiveness of measures for taxpayer registration and tax assessment
PI-15 Effectiveness in collection of tax payments
PI-16 Predictability in the availability of funds for commitment of expenditures
PI-17 Recording and management of cash balances, debt and guarantees
PI-18 Effectiveness of payroll controls
PI-19 Competition, value for money and controls in procurement
PI-20 Effectiveness of internal controls for non-salary expenditure and assets management
PI-21 Effectiveness of internal audit
c(iii) Accounting, recording and reporting
PI-22 Timeliness and regularity of accounts reconciliation
PI-23 Availability of information on resources received by service delivery units
PI-24 Quality and timeliness of in-year budget reports
PI-25 Quality and timeliness of annual financial statements
c(iv) external scrutiny and audit
PI-26 Scope, nature and follow-up of external audit
PI-27 Legislative scrutiny of the annual budget law
PI-28 Legislative scrutiny of external audit reports
d. donor practices
D-1 Predictability of direct budget support
D-2 Financial information provided by donors for budgeting and reporting on project and programme aid
D-3 Proportion of aid that is managed by use of national proceduresSource: PEFA Secretariat, 2006a.
* These are the authors’ comments.
13
Various PEFA indicators are directly relevant to the extent that they provide information about the degree to which aid is captured in the budget system; indicators may also be indirectly relevant for the light they shed on aspects of the underlying PFM system that affect the scope for bringing aid on budget, or the quality of its capture in the budget system. The particular relevance of each of the highlighted indicators is as follows:
indicator Relevance
PI-5 classification of the budget on budget, on account
PI-6 comprehensiveness of information included in the budget documentation
on plan, on budget, on parliament
PI-7 capture of income and expenditure information for donor funded projects, even if they are not technically on budget or are managed outside of government’s budget management and accounting system.
on budget, on account, on report
PI-27 legislative scrutiny of the annual budget law on parliament
PI-28 legislative scrutiny of external audit reports on parliament, on audit
D-1 predictability of direct budget support on plan, on budget
D-2 degree to which donors provide financial information for budgeting and reporting on project and programme aid
on plan, on budget, on parliament, on audit, on report
D-3 proportion of aid funds that are managed by the use of national procedures
on budget, on treasury, on audit
Several of the other PEFA indicators also influence the potential benefits of putting aid on budget and donors’ willingness to put aid on budget. (For example, PI-12 – multi-year perspective in fiscal planning, expenditure policy and budgeting – is highly relevant to the “on plan” dimension.) Moreover, the score for most indicators draws on several different sub-components. In some cases it may be worth highlighting specific sub-indicators, for example, component (i) of Indicator PI-7 does not relate to aid, but component (ii) focuses explicitly on the capture of aid in fiscal reports.
Relevant Paris Declaration commitments and targets
The Paris Declaration includes a wide range of reciprocal commitments by aid agencies and partner governments. These are relevant to putting aid on budget at all points of the budget cycle. (See Box 7.)
The Paris Declaration is supported by a set of 12 indicators with targets for 2010, developed to track and encourage progress against the broader set of partnership commitments. Three of these indicators are directly concerned with putting aid on budget. (See Box 8.)
Other international standards
There are other international standards and processes that are also focusing attention on putting aid on budget. The IMF Code and Manual of Good Practice on fiscal transparency (first introduced in 1998, with a revised version in April 2007) has new recommendations to strengthen the standard for reporting on external financing, calling for separate identification of aid in the budget.13 Also relevant has been the work on the 2003 International Public Sector Accounting Standard (IPSAS) for cash-based financial reporting to develop mandatory and optional (“encouraged”) disclosures on official development assistance. The cash-based standards are especially relevant to developing countries as the vast majority of their governments do not have accrual-based systems. (See also Box 30.)
13 “Receipts from all major revenue sources, including resource-related activities and foreign assistance, should be separately identified in the annual budget presentation” (International Monetary Fund, 2007b).
14
Chapter 4 provides a review of what is known about international performance against the key aid-on-budget indicators and targets described above.
Box 7: Paris Declaration commitments to putting aid on budget
Ownership
(14) Partner countries commit to:Exercise leadership in developing and implementing their national development strategies through •broad consultative processes.Take the lead in coordinating aid at all levels in conjunction with other development resources in •dialogue with donors and encouraging the participation of civil society and the private sector.
(15) Donors commit to: Respect partner country leadership and help strengthen their capacity to exercise it.•
Alignment
Donors align with partners’ strategies(16) Donors commit to:
Base their overall support – country strategies, policy dialogues and development co-operation •programmes – on partners’ national development strategies and periodic reviews of progress in implementing these strategies (Indicator 3).
Donors use strengthened country systems (21) Donors commit to:
Use country systems and procedures to the maximum extent possible. Where use of country systems •is not feasible, establish additional safeguards and measures in ways that strengthen rather than undermine country systems and procedures.Avoid, to the maximum extent possible, creating dedicated structures for day-to-day management and •implementation of aid-financed projects and programmes.
Strengthen public financial management capacity(25) Partners commit to:
Publish timely, transparent and reliable reporting on budget execution.•(26) Donors commit to:
Provide reliable indicative commitments of aid over a multi-year framework and disburse aid in a timely •and predictable fashion according to agreed schedules (Indicator 7).Rely, to the maximum extent possible, on transparent partner government budget and accounting •mechanisms (Indicator 5).
harmonisation
Donors implement common arrangements and simplify procedures(32) Donors commit to:
Implement, where feasible, common arrangements at country level for planning, funding (e.g. joint •financial arrangements), disbursement, monitoring, evaluating and reporting to government on donor activities and aid flows. Increased use of programme-based aid modalities can contribute to this effect (Indicator 9).
Deliver effective aid in fragile states(39) Donors commit to:
Avoid activities that undermine national institution building, such as bypassing national budget •processes or setting high salaries for local staff.
managing for results
(44) Partner countries commit to:Strengthen the linkages between national development strategies and annual and multi-annual budget •process.
mutual accountability
(48) Partner countries commit to:Strengthen as appropriate the parliamentary role in national development strategies and/or budgets.•
(49) Donors commit to:Provide timely, transparent and comprehensive information on aid flows so as to enable partner •authorities to present comprehensive budget reports to their legislatures and citizens.
Source: Paris High Level Forum, 2005
15
Box 8: Paris Declaration indicators and targets for putting aid on budget
indicator monitoring target
Indicator 3: Aid flows are aligned with national priorities
The gap between what was disbursed by donors for the government sector and what was actually recorded in the annual budget by government.
Halve the proportion of aid flows to government sector not reported on government’s budget(s) (with at least 85% reported on budget)
Indicator 5a: Use of country PFM systems
The volume of aid and the percentage of donors that uses a partner country PFM system as a percent of total aid provided to the government sector.
Looks at the three components of a country’s PFM procedures:i. national budget execution proceduresii. national financial reporting proceduresiii. national auditing procedures.
Score 5+: all donors use partner countries PFM systems and a two-thirds reduction in the % of aid to the public sector not using partner countries PFM systems;
Score 3.5–4.5: 90% all donors, and a one-third reductiona
Indicator 7: Aid is more predictable
The gap between aid scheduled and aid effectively disbursed and recorded in countries accounting systems.
(Focuses specifically on in-year predictability of aid flows to the government sector.)
Halve the proportion of aid not disbursed within the fiscal year for which it was scheduled
Source: Paris High Level Forum, 2005; OECD Development Assistance Committee, 2006a.
a These are World Bank country policy and institutional assessment (CPIA) scores. (See Box 11.)
16
4
hOw much Aid iS On Budget?
This chapter provides an overview of international efforts to monitor the proportion of aid that is on budget, and draws some tentative conclusions.
4.1 introduction
This chapter assembles international data from various diagnostic instruments and surveys. It compares data for the ten main case study countries, plus some additional comparators. Some initial warnings are necessary: in most cases the data were not designed to permit international comparisons, and therefore need to be treated very cautiously; equally, there is a lack of reliable time series within countries; and the aid on budget case studies, and some observers, have highlighted further weaknesses in the main indicator set, from the 2006 Survey on Monitoring the Paris Declaration.
4.2 Survey on the implementation of the Paris declarationThe 2006 Survey on Monitoring the Paris Declaration
The 2006 Survey on Monitoring the Paris Declaration reports on aid flows in 2005. Its data were collected from 34 self-selected countries and a comprehensive list of donor organisations covering 37 percent of the global aid programme (OECD, 2007a). Of the data available on country experiences of putting aid on budget, the Paris Declaration monitoring is the only quantified cross-country (and cross-donor) survey.
However, difficulties were experienced in relation to consistency in the application of definitions. The leeway that it allowed on the local application of definitions makes its data unreliable for comparisons across countries. The survey report recommends a more rigorous standardisation of measures for the next survey round (which is in 2008).14 These problems have already undermined the possibility of reliable comparisons, even at the individual country level, between the 2006 and 2008 surveys on the indicators most relevant to putting aid on budget.
Country performance
Box 9 shows the monitoring data for the ten main case study countries. The survey was designed to set a baseline for measurement of progress on the Paris Declaration indicators; its reference year is 2005 (or the 2004/05 financial year, when the fiscal year does not match the calendar year). The survey’s categories correspond only approximately to some of the on budget dimensions identified for the present study. Moreover, the way the baseline is assessed includes some upward biases:
The indicator does not distinguish between over-recording and under-recording of aid in the budget. •Moreover, the aid recorded as spent may not be the same (in its source or in its use) as the aid anticipated in the budget. This means that any similarity between ex ante and ex post totals may be coincidence.
14 See the Putting Aid On Budget Literature Review (especially its Box 8) for a full discussion.
17
As the Putting Aid On Budget Literature Review describes, countries were allowed some latitude in •interpreting definitions, and the survey’s compilers themselves point out that this is likely to have biased scores upwards (notably in the criteria applied for the use of country systems).
Doubts about the reliability of the Paris Declaration data were reinforced by observations from several of the putting aid on budget country case studies:
The reports for Ghana and Mali noted the lack of transparency in how the reported Paris Declaration •scores were calculated; this makes it difficult to verify the scores, and will cause problems in ensuring within-country comparability between this survey and subsequent ones.
For Uganda, it was reported by some donors, in particular the World Bank, that project aid used •government budget execution systems. But this is not the case, and the survey generally seems to provide an over-optimistic assessment of aid capture. (See Putting Aid On Budget: Uganda Case Study, Box 7.)
The Rwanda report highlights reservations (on the part of the Rwanda government as well as some •donors) about how much aid is properly described as “aid to the government sector”. This may make a large difference to the denominator in calculating the proportion of aid on budget, and could make a huge difference to some donors’ scores.
The South Africa study finds that aid is not systematically incorporated in the budget. However, the Paris •Declaration survey’s score for Indicator 3 is based on the fact that some aggregate information on past, current and expected forward flows into the Reconstruction and Development Programme (RDP) fund (a fund established at national level to receive aid flows) appears in a few tables in the Budget Review.
Donor performance
Box 10 shows selected donors’ performance against Paris Declaration indicators 3, 5a and 7. These data are subject to the same caveats as the country performance data. Moreover, they are based on a (self-selected) subset of the countries to which these agencies provide aid, and therefore cannot be assumed to be representative of each agency’s entire portfolio.
4.3 consolidated data on PFm performance
Box 11 consolidates Paris Declaration monitoring scores with other PFM diagnostic scores for indicators that measure the extent of aid capture on budget. The data set includes Heavily Indebted Poor Countries (HIPC) assessment and action plans (AAPs), World Bank country policy and institutional assessments (CPIAs), Public Expenditure and Financial Accountability (PEFA) assessments, and the 2006 Survey on Monitoring the Paris Declaration.
Once again, caveats are required. The HIPC AAP scores provide comparisons between 2001 and 2004 assessments, but the indicators are much cruder and less transparent than the subsequent PEFA diagnostic. On the other hand, none of the PEFA indicators addresses the dimensions of putting aid on budget exclusively or precisely. Therefore they need to be interpreted with care, as detailed further in Annex B of the Putting Aid On Budget Literature Review. It is also important to note that cross-country comparison of PEFA indicators was never an objective of the framework.15
15 The PEFA Report on Early Experience details why such a comparison is difficult:
• “Comparisonbetweentwocountriesonlymakessenseonindicatorsforwhichthecompliancewithscoringmethodologyishighinordertoensureconsistency and adequate documentation for differences; on the current stock of reports that limits the countries for which such a comparison can be meaningfully taken.
• Comparisonofacountry’sperformanceagainstrelevantglobalandregionalaveragesrequiresthatassessmentshavetakenplaceinasignificantnumber of countries with comparable characteristics, which is not yet the case” (PEFA Secretariat, 2006b).
18
Box
9: P
aris
Dec
lara
tion
mon
itor
ing
dat
a fo
r ca
se s
tud
y co
untr
ies
dat
a fo
r 20
05
Bur
kina
Fa
soe
thio
pia
gha
nak
enya
mal
im
oz.
Rw
and
aSo
uth
Afr
ica
tanz
ania
ug
and
a
Co
re a
ida r
epo
rted
to
the
DA
CU
S$m
688
1,33
31,
147
793
714
1,32
654
575
91,
543
1,12
6
Aid
rep
ort
ed in
200
6 su
rvey
b
US$
m59
31,
288
1,04
766
762
51,
267
571
583
1,43
31,
088
Tota
l bud
get
sup
po
rt d
isb
urse
dU
S$m
168
356
296
6512
433
719
80
573
391
% o
f to
tal a
id28
%28
%28
%10
%20
%27
%35
%0%
40%
36%
Tota
l aid
dis
bur
sed
to
g
ove
rnm
ent
sect
or
US$
m53
11,
048
946
456
557
1,13
355
435
11,
294
854
On
pla
n %
to
tal d
isb
urse
men
ts–
––
––
––
––
–
On
bud
get
%
of
do
nor
com
mit
-m
ents
rep
ort
ed in
b
udg
et e
stim
ate
75%
74%
90%
67%
60%
77%
52%
69%
82%
89%
% o
f d
ono
r d
isb
urse
-m
ents
rep
ort
ed in
b
udg
et e
stim
ate
68%
74%
96%
91%
60%
83%
49%
71%
90%
79%
On
par
liam
entc
% t
ota
l dis
bur
sem
ents
68%
74%
96%
91%
60%
83%
49%
0%90
%79
%
On
trea
sury
%
to
tal d
isb
urse
men
ts44
%58
%69
%47
%32
%40
%37
%36
%76
%66
%
On
acco
unt
% t
ota
l dis
bur
sem
ents
50%
45%
61%
46%
29%
35%
40%
42%
61%
48%
On
aud
it %
to
tal d
isb
urse
men
ts40
%32
%56
%49
%28
%33
%40
%36
%61
%66
%
On
rep
ort
%
to
tal d
isb
urse
men
ts–
––
––
––
––
–
Sour
ce: 2
006
Surv
ey o
n M
oni
torin
g t
he P
aris
Dec
lara
tion
(OE
CD
, 200
7a).
a.
“Co
re a
id”
mat
ches
clo
sely
the
defi
nitio
n us
ed in
the
200
6 Su
rvey
on
Mo
nito
ring
the
Par
is D
ecla
ratio
n; it
exc
lud
es d
ebt
reo
rgan
isat
ion
and
hum
anita
rian
aid
.
b.
For
the
2006
Sur
vey
on
Mo
nito
ring
the
Par
is D
ecla
ratio
n th
e ai
d c
ove
rag
e b
y co
untr
y ra
nged
fro
m 6
2 p
erce
nt t
o 1
18 p
erce
nt (s
urve
y re
po
rt s
ays
diff
eren
ces
coul
d b
e d
ue t
o a
id fr
om
oth
er d
ono
rs, v
aria
tions
in t
imin
g o
f re
cord
ing
the
flo
ws
and
slig
ht d
iffer
ence
s o
f defi
nitio
n) (O
EC
D, 2
007a
).
c.
On
par
liam
ent
sco
res
are
rep
eats
of t
he o
n b
udg
et s
core
s, o
n th
e b
asis
tha
t th
e co
untr
y p
arlia
men
ts a
pp
rove
the
nat
iona
l bud
get
pre
sent
ed b
y th
e g
ove
rnm
ent,
whi
ch in
clud
es t
he a
id o
n b
udg
et. A
n ex
cep
tion
is S
out
h A
fric
a w
here
aid
is n
ot
app
rove
d (a
s p
art
of t
he b
udg
et o
r se
par
atel
y) b
y p
arlia
men
t.
•
Theex
tentofc
aptureofa
idatthedifferen
tdim
ensionsofthe
bud
getcouldbeeitheran“o
ver”or“u
nder”captureofa
id.Th
isisnotshownbytheperce
ntag
esaspresented
above
.
•Onplanan
donreportdim
ensionswereno
tassessed
bythe20
06Surve
yonMonitorin
gthe
ParisD
eclaratio
n.
19
Box 10: Donor ratios for relevant indicators for putting aid on budget
Source: 2006 Survey on Monitoring the Paris Declaration (OECD, 2007a) (From Putting Aid On Budget Literature Review, Table A5).
Notes:
• Generalnote:Theextentofcaptureofaidatthedifferentdimensionsofthebudgetcouldbeeitheran“over”or“under”captureofaid.This is not shown by the percentages as presented above.
• Thebaselineratioisaweightedaverage,basedoneachdonor’sportfoliointhesurveyedcountry.Itisthe2006SurveyonMonitoringthe Paris Declaration results for the base year of 2005. It is the aggregate value of the numerator divided by the aggregate value of the denominator; i.e. each country is weighted by the volume of the activity (OECD, 2007a).
• Thecountryaverageisanunweightedaverage.Itprovidesacomparativemeasureofthebaselineirrespectiveofthevolumeofactivityineach country; i.e., it gives equal weight to each country) (OECD, 2007a).
• Theguidanceforthe2006SurveyonMonitoringtheParisDeclarationincludedencouragementtoUNagenciestoreportbothindividuallyand collectively in completing the survey. In the survey report the UN information was reported in total for the UN.
Ag
greg
ate
AfD
B
AsD
B
Canad
a
Denm
ark
EC
France
Germ
any
IDB
Ireland
Japan
Netherland
s
No
rway
Swed
en
UK
UN US
Wo
rld B
ank
3. Aid flows aligned
Government budget estimates of aid flows as % of aid
disbursed for the government sector
baseline ratio
88% 95% 88% 75% 47% 81% 52% 55% 60% 48% 68% 70% 56% 49% 84% 30% 90% 94%
country average
42% 59% 62% 51% 47% 56% 43% 50% 48% 48% 30% 44% 57% 35% 45% 34% 30% 62%
5a. Use of country PFM systems
% aid flows disbursed for the government sector that use
national PFM systems
baseline ratio
40% 33% 69% 42% 29% 40% 28% 35% 45% 90% 29% 71% 61% 47% 75% 18% 10% 42%
country average
33% 28% 56% 35% 27% 38% 28% 28% 29% 90% 16% 60% 56% 40% 53% 15% 15% 36%
7. Aid is more predictable
Government acounts of disbursements as % of aid
scheduled by donors for disbursement
baseline ratio
70% 56% 91% 73% 49% 65% 45% 75% 82% 72% 66% 65% 50% 54% 90% 32% 45% 68%
country average
41% 52% 86% 42% 50% 49% 30% 48% 88% 63% 34% 52% 55% 48% 48% 18% 26% 63%
20
Box
11:
Con
solid
ated
cou
ntry
rat
ing
s
Par
is d
ecla
rati
on
(200
5)P
eFA
Ass
essm
ents
(a
sses
smen
t ca
rrie
d o
ut 2
006
or
2007
)ah
iPc
AA
Pc
PiA
ind
icat
or
3:
% a
id fl
ow
s re
po
rted
on
bud
get
ind
icat
or
5a:
% a
id fl
ow
s us
ing
co
untr
y P
Fm
ind
icat
or
7: %
ai
d d
isb
urse
d
as s
ched
uled
Pi-5
Pi-6
Pi-7
Pi-2
7P
i-28
d-1
d-2
d-3
ind
icat
or
4(2
004
and
sc
ore
s in
b
rack
ets
for
2001
)b
ind
icat
or
13
(200
6)c
Bud
get
Op
er-
atio
nsA
udit
do
nor
pra
ctic
es
Ave
rag
es88
%40
%70
%3.
2
Ban
gla
des
h 88
%53
%91
%C
CD
DC
BC
Bn/
a3.
0
Bo
livia
71%
26%
63%
No
ass
essm
ent
pla
nned
A (A
)3.
5
Bo
tsw
ana
n/a
n/a
n/a
Ass
essm
ent
dra
ft re
po
rt p
lann
ed e
arly
200
8n/
an/
a
Bur
kina
Fas
o68
%45
%92
%A
AB
+B
+C
+A
C+
CC
(C)
4.0
eth
iop
ia74
%45
%96
%A
BD
+D
+C
+C
CU
nab
le
to s
core
B (B
)4.
0
gha
na96
%62
%92
%B
CA
C+
C+
C+
CD
B (B
)4.
0
ken
ya91
%47
%44
%C
BC
D+
D+
DD
Dn/
a3.
5
mal
i60
%29
%71
%C
CC
+B
+D
DD
+D
B (B
)3.
5
mo
zam
biq
ue83
%36
%70
%B
BC
+B
+C
+C
+D
+D
B (B
)3.
5
nic
arag
ua73
%44
%70
%A
sses
smen
t co
mp
lete
d D
ecem
ber
200
6, n
ot
pub
licly
ava
ilab
lesc
ore
s no
t g
iven
4.0
Rw
and
a49
%39
%66
%A
DB
C+
D+
B+
DD
B (B
)4.
0
Sene
gal
89%
23%
69%
Ass
essm
ent
com
men
ced
200
7B
(B)
3.5
Sout
h A
fric
a71
%38
%44
%N
o a
sses
smen
t p
lann
edn/
an/
a
tanz
ania
90%
66%
70%
CA
BC
+C
AC
CB
(B)
4.5
ug
and
a79
%60
%84
%B
BC
C+
D+
C+
D+
CB
(A)
4.0
Vie
tnam
81
%32
%78
%A
sses
smen
t co
mm
ence
d 2
008
n/a
4.0
Sour
ce: P
uttin
g A
id O
n B
udg
et L
itera
ture
Rev
iew
, Tab
le D
3.
a.
The
PEFA
ind
icat
ors
are
: PI-
5: C
lass
ifica
tion
of t
he b
udg
et; P
I-6:
Co
mp
rehe
nsiv
enes
s o
f inf
orm
atio
n in
clud
ed in
bud
get
do
cum
enta
tion;
PI-
7: E
xten
t o
f unr
epo
rted
go
vern
men
t o
per
atio
ns; P
I-28
: Leg
isla
tive
scru
tiny
of t
he
exte
rnal
aud
it re
po
rts;
D-1
: Pre
dic
tab
ility
of D
irect
Bud
get
Sup
po
rt; D
-2: F
inan
cial
info
rmat
ion
pro
vid
ed b
y d
ono
rs fo
r b
udg
etin
g a
nd re
po
rtin
g o
n p
roje
ct a
nd p
rog
ram
me
aid
; and
D-3
: Pro
po
rtio
n o
f aid
tha
t is
man
aged
b
y us
e o
f nat
iona
l pro
ced
ures
.
b.
H
IPC
ass
essm
ent
and
act
ion
pla
ns (A
APs
) inc
lud
e as
sess
men
t o
f the
ext
ent
to w
hich
do
nor
fund
s ar
e in
clud
ed in
cen
tral
, sta
te a
nd lo
cal g
ove
rnm
ents
’ bud
get
(s) a
nd/o
r fis
cal r
epo
rts.
The
defi
nitio
n o
f aid
on
bud
get
us
ed in
the
AA
Ps is
bro
ad (n
o d
isag
gre
gat
ion
of a
id o
n b
udg
et a
t d
iffer
ent
dim
ensi
ons
of t
he b
udg
et c
ycle
and
ex
ante
and
ex
po
st re
po
rtin
g b
y d
ono
rs in
clud
ed in
sam
e in
dic
ato
r) a
nd t
he A
AP
answ
ers
on
the
exte
nt t
o
whi
ch a
id is
on
bud
get
are
no
t p
reci
se (g
ener
al g
rad
es u
sed
of “
all”
, “in
com
ple
te”,
“no
ne”)
. (Se
e Pu
ttin
g A
id O
n B
udg
et L
itera
ture
Rev
iew
, Ann
ex B
, fo
r m
ore
det
ail.)
c.
Thes
e sc
ore
s ar
e ta
ken
fro
m t
he 2
006
resu
lts o
f the
Wo
rld B
ank
annu
al c
oun
try
po
licy
and
inst
itutio
nal a
sses
smen
t (C
PIA
) exe
rcis
e, w
hich
co
vers
the
Inte
rnat
iona
l Dev
elo
pm
ent
Ass
oci
atio
n (ID
A) e
ligib
le c
oun
trie
s.
Ind
icat
or
13 is
the
qua
lity
of b
udg
etar
y an
d fi
nanc
ial m
anag
emen
t.
21
4.4 conclusions from the data
Efforts to assess the extent of aid flow capture and the quality of that capture have become increasingly systematic. As detailed in the paragraphs above, there are major limitations in both the coverage and consistency of the data. However, it supports the following broad conclusions.
Large volumes of aid are (still) off budget
Box 12 shows the 2006 Survey on Monitoring the Paris Declaration findings (by country) for Paris Declaration indicators 3 and 5a. Results for Indicator 3 (“donor disbursements recorded in budget estimates”) are in aggregate high (88 percent) and with a number of countries with over 80 percent of aid recorded on budget (Bangladesh, Ghana, Kenya, Mozambique, Senegal, Tanzania and Vietnam). However, both in aggregate and at country level, the survey results seem likely to overestimate the extent of aid capture. (See discussion under “Country performance“ on page 16.) The survey’s findings show that in fact there is much less aid using country systems than is captured in the budget documents.16
Box 12: Paris Declaration monitoring data: Country results for indicators 3 and 5a
Source: For the full set of detailed figures see the Putting Aid On Budget Literature Review, Annex A.
Notes:
The 2010 targets are set as follows:
• Indicator3target:Halftheproportionofaidflowstogovernmentsectornotreportedongovernment’sbudget(s)(withatleast85%reportedon budget).
• Indicator5atarget:
– World Bank CPIA score 5+: all donors use partner countries’ PFM systems and a two-thirds reduction in the % of aid to the public sector not using partner countries PFM systems;
– CPIA Score 3.5–4.5: 90% all donors, and a one-third reduction.
16 The “using country systems” percentages in Box 12 are the average use of country systems of budget execution, financial reporting and audit; for separate estimates of each element, see Box 9.
% Donor disbursements recorded in budget estimates
88% 88%71% 68% 74%
96% 91%
60%
83%73%
49%
89%71%
90%79% 81%
0%
20%
40%
60%
80%
100%
Ag
gre
gat
e
Ban
gla
des
h
Bo
livia
Bur
kina
Fas
o
Eth
iop
ia
Gha
na
Ken
ya
Mal
i
Mo
zam
biq
ue
Nic
arag
ua
Rw
and
a
Sene
gal
Sout
h A
fric
a
Tanz
ania
Ug
and
a
Vie
tnam
2010 target
2005 baseline
% Donor disbursements using country systems
40%53%
26%
45% 45%62%
47%29% 36%
44% 39%23%
38%
66% 60%
32%
0%
20%
40%
60%
80%
100%
Ag
gre
gat
e
Ban
gla
des
h
Bo
livia
Bur
kina
Fas
o
Eth
iop
ia
Gha
na
Ken
ya
Mal
i
Mo
zam
biq
ue
Nic
arag
ua
Rw
and
a
Sene
gal
Sout
h A
fric
a
Tanz
ania
Ug
and
a
Vie
tnam
2010 target
2005 baseline
% Donor disbursements recorded in budget esimates
% Donor disbursements using country sytems
22
Low scoring is also noticeable for the three donor performance indicators in the PEFA indicator set, which measure the predictability of direct budget support (D-1), the financial information provided by donors for budgeting and reporting on project and programme aid (D-2) and the proportion of aid that is managed by use of national procedures (D-3). D-2 and D-3 are especially relevant. The scores recorded against these criteria are almost without exception consistently low across the countries assessed (and for which the PEFA scores are available), as Box 13 shows. (See also Box 16, which shows the rank order of country ratings against the indicators shown in Box 13.)
Box 13: PEFA performance indicator scores across countries
indicAtOR A B+ B c+ c d+ d
B. key cross-cutting issues: comprehensiveness and transparency
PI-5 Classification of the budget 6% 0% 35% 6% 53% 0% 0%
PI-6Comprehensiveness of information included in budget documentation
16% 0% 47% 5% 26% 0% 5%
PI-7Extent of unreported government operations
12% 6% 18% 18% 18% 24% 6%
c(iii). Accounting, recording and reporting
PI-27Legislative scrutiny of the annual budget law
0% 16% 0% 37% 11% 26% 11%
PI-28Legislative scrutiny of external audit reports
0% 0% 6% 13% 25% 31% 25%
d. donor practices
D-1Predictability of direct budget support
25% 0% 8% 17% 17% 17% 17%
D-2Financial information provided by donors for budgeting and reporting on project and programme aid
14% 0% 0% 7% 21% 36% 21%
D-3Proportion of aid that is managed by use of national procedures
0% 0% 8% 0% 42% 0% 50%
All Average of 31 indicators 10% 3% 14% 14% 22% 21% 17%
Source: PEFA Secretariat, 2006b.
Note:
The country set for this data is Afghanistan, Bangladesh, Congo Brazzaville, Fiji, Ghana, Guatemala, Kyrgyz Republic, Lesotho, Malawi, Moldova, Mozambique, Panama, Papua New Guinea, Syria, Tanzania, Uganda, Zambia. Not all these reports are publicly available.
23
Levels of aid on budget are strongly driven by budget support aid (which, by definition, is on budget). In many cases, off budget proportions for other aid modalities still remain very high.
Budget support is by definition on budget. Therefore, when taking budget support as fully captured on budget, it is possible to infer the (limited) extent of capture of other modalities. Taking the example of Rwanda:
49% of total ODA provided to Rwanda in 2005 was recorded in the national budget. If one takes into account that budget support represents approximately 41%17 of total ODA disbursed to Rwanda in 2005 and that budget support is 100% accounted for in the budget, this would imply that project support is very badly captured in the national budget. (Putting Aid On Budget: Rwanda Case Study)
The 2006 Survey on Monitoring the Paris Declaration captures the amount of aid provided as budget support, which makes it possible to calculate how much of the aid included in government budget estimates is additional to budget support. Box 14 uses the survey data to show what proportion of aid on budget is budget support, and how much other aid (not budget support) to the government sector is on budget. The data has been organised in order of percentage capture of other aid to the government sector on budget.18 This is done in order to see whether the countries that receive a lot of budget support, are also good at putting their other (project) aid on budget. If anything, the opposite seems to be the case. The countries which have the lowest proportion of non-budget support (project) aid on budget include many of the largest recipients of budget support.
The extent of capture of other modalities on treasury is even weaker. As noted in Chapter 2, there are some countries where it is taken as axiomatic that projects are off treasury. (See discussion under “Misconceptions to avoid” on page 6 in Chapter 2.)
Large (and sometimes puzzling) differences in donor scores
The 2006 Survey on Monitoring the Paris Declaration reports that for the extent to which donors’ disbursements are recorded in governments’ budget estimates, the detailed figures reveal quite “striking differences” among donors that are involved in similar numbers and types of countries (OECD, 2007a). However, the limited consistency in the application of definitions, across both donors and countries, means that it is not possible to draw conclusions from this data.
Looking at the issue of disbursement according to schedule, the survey finds that a few of the major donors are substantial under- or over-disbursers, but concludes that it is unclear if there is an explanation related to the subset of surveyed countries where those agencies are involved.
Lack of strong correlation between aid on budget and scores for quality of public finance management
There is considerable variation in the use of country systems across countries and the correlation with the quality of the country PFM system. Also, the use being made of them by donors is rather weak.19 (See Box 15.) Countries such as Bangladesh, Mozambique or Kenya – whose PFM systems are not too good, as captured by the World Bank’s CPIA index – have achieved levels of aid capture on budget comparable to those of countries with better CPIA scorings on the quality of their PFM systems. On the other hand, countries such as Vietnam or Ethiopia, whose PFM systems are rated among the best of this group of countries, are among those with lower proportions of aid making use of national PFM systems.This suggests that factors other than the quality of systems are affecting donors’ willingness to use them. (The country case studies suggest that the other factors involved may include political relationships and perceptions about countries’ willingness to address PFM weaknesses.)
Quality is crucial
For aid effectiveness, it is very important to look behind the percentages and consider the quality of aid capture at different stages in the budget process. This is the main focus of the country case studies reviewed in Chapter 6.
17 Government of Rwanda source for budget support statistics, which records budget support at circa 6 percent more of total aid than as recorded by the 2006 Survey on Monitoring the Paris Declaration.
18 As discussed in the section on country performance in Chapter 4, these data are at the upper limits due to the methodological upward biases.19 The 2006 Survey on Monitoring the Paris Declaration draws attention to this point (OECD, 2007a).
24
Box 14: Paris Declaration survey: Budget support and on budget data
OdA Aid disbursed for government sector
1 2 3 4 5 5(i) 5(ii) 6
Total Budget support
Other aid modalities
Total Share recorded on budget Share non budget
support aid for government
sector that is on budget
Baseline ratio of aid on budget
Budget support as share of aid
for government sector
Proportion of baseline ratio
(col.5) that is not budget support
US$m % % US$m % % % %
Rwanda 571 35% 65% 554 49% 36% 13% 21%
Mali 625 20% 80% 557 60% 22% 38% 48%
Burkina Faso 593 28% 72% 531 68% 32% 36% 53%
Ethiopia 1,288 28% 72% 1,048 74% 34% 40% 61%
Nicaragua 533 17% 83% 418 73% 22% 51% 65%
Bolivia 791 13% 87% 628 71% 16% 55% 67%
Uganda 1,088 36% 64% 854 79% 46% 33% 67%
South Africa 583 0% 100% 351 71% 0% 71% 71%
Mozambique 1,267 27% 73% 1,133 83% 30% 53% 76%
Vietnam 1,956 17% 83% 1,941 81% 17% 64% 76%
Tanzania 1,433 40% 60% 1,294 90% 44% 46% 81%
Bangladesh 1,837 16% 84% 1,414 88% 21% 67% 85%
Senegal 515 12% 88% 453 89% 13% 76% 87%
Kenya 667 10% 90% 456 91% 14% 77% 90%
Ghana 1,047 28% 72% 946 96% 31% 65% 94%
Source: OECD, 2007a; OECD, 2007b
Notes:
• ODA: Official development assistance (ODA) includes all transactions as defined in OECD-DAC Statistical Directives para. 35, including official transactions that: 1) are administered with the promotion of the economic development and welfare of developing countries as main objective; and 2) are concessional in character and convey a grant element of at least 25 percent. Excluded from the scope of the survey are: transactions made to beneficiaries not based in the country receiving ODA or to regional organisations; debt reorganisation/restructuring; and emergency and relief assistance.
• Aid disbursed for government sector: As reported by donors, the disbursement of ODA in the context of an agreement with the government sector administrations – ministries, departments, agencies or municipalities – authorised to receive revenue or undertake expenditures on behalf of central government including works, goods or services delegated or subcontracted by government to other entities (for example, NGOs and private companies).
• Total aid recorded on budget: Included in the budget estimate. As reported by government, ODA recorded in the annual budget as revenue or grants.
The data sources and calculations for each column of the data table above are:
1. Paris Declaration (PD) survey, table A.0: Aid reported in the 2006 survey.
2. PD survey, table A.9: Budget support as percentage of total aid reported.
3. Percentage of total aid reported that is not budget support.
4. PD survey, table A.3: Aid disbursed by donors for government sector.
5. PD survey, table A.3: Baseline ratio of aid disbursed by donors included in government budget estimates.
5(i). Budget support as percentage of aid disbursed for government sector.
5(ii). Proportion of baseline ratio (col.5) that is not budget support.
6. Percentage of non-budget support aid for government sector (calculated from PD survey, table A.3: Aid disbursed, minus PD survey, table A.9: Budget support) that is on budget (calculated from PD survey, table A.3: Aid on budget, minus PD survey, table A 9: Budget support).
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50%
% aid to the government sector that is budget support
% non-budget support aid for the government
sector that is on budget
South Africa
Kenya
Senegal
Ghana
Bangladesh
Vietnam
Bolivia Nicaragua
Mozambique
Tanzania
Ethiopia Uganda
Mali
Burkina Faso
Rwanda
25
Box 15: PFM quality versus the use of PFM systems
Paris declaration monitoring (2005 data) cPiA
indicator 5a: % aid flows using
country PFm
Rank on indicator 5a
indicator 13 (2006) [PFm quality]
Averages 40% 3.20
Tanzania 66% 1 4.50
Ghana 62% 2 4.00
Uganda 60% 3 4.00
Bangladesh 53% 4 3.00
Kenya 47% 5 3.50
Ethiopia 45% 6= 4.00
Burkina Faso 45% 6= 4.00
Nicaragua 44% 8 4.00
Rwanda 39% 9 4.00
Mozambique 36% 10 3.50
Vietnam 32% 11 4.00
Mali 29% 12 3.50
Bolivia 26% 13 3.50
Senegal 23% 14 3.50
Source: Selected data from Box 11 above.
4.5 Recommendations for monitoring aid on budget
Measuring the extent of aid using country systems at the various dimensions of “on budget” is complex. Recommendations for joint action by government and their aid partners are to strengthen the monitoring of the Paris Declaration commitments by:
setting standardised and less flexible definitions at the international level•
monitoring the proportion of project aid that is on budget•
continuing the country-level focus on the use of government systems, and setting country-level targets •for progress.
26
Box 16: Selected available PEFA scores in rank order
The PEFA framework rates PFM performance in participating countries using a total of 28 PFM indicators, Ratings for each of these indicators are made using an A to D scale ranking, where A is the highest score and D is the lowest. The aid on budget case study countries are highlighted, and analysis behind the scores is included in the case study reports.
The PEFA diagnostic is not designed to produce international league tables on PFM. Therefore, it is important not to read too much into these rankings. However, it is interesting that:
• formostoftheindicatorsthereisaspreadofscoresamongthecasestudies,andtherankingsare significantly different for different indicators. (This suggests some useful variety in the case study sample.)
• forindicatorsD-2andD-3(donorperformanceonprovidinginformationandusingnationalprocedures), scores for all the study countries are in the lower half of the range. (The same is true of Indicator PI-28: legislative scrutiny of external audit.)
Pi-5 classification of the budget
Burkina Faso Apr 2007 A
ethiopia 2007 A
Rwanda Sept 2007 A
Ukraine Mar 2007 A
ghana Jun 2006 B
mozambique mar 2006 B
uganda Jul 2006 B
Afghanistan Dec 2005 C+
Zambia Dec 2005 C+
Bangladesh Jul 2007 C
kenya Jul 2006 c
Kyrgyz Republic Jan 2006 C
mali mar 2007 c
Moldova Jun 2006 C
tanzania Jun 2006 c
Pi-6 comprehensiveness of information included in budget documentation
Burkina Faso Apr 2007 A
Moldova Jun 2006 A
tanzania Jun 2006 A
Ukraine Mar 2007 A
ethiopia 2007 B
kenya Jul 2006 B
Kyrgyz Republic Jan 2006 B
mozambique mar 2006 B
uganda Jul 2006 B
Zambia Dec 2005 B
Afghanistan Dec 2005 C
Bangladesh Jul 2007 C
ghana Jun 2006 c
mali mar 2007 c
Rwanda Sep 2007 d
27
Pi-7 extent of unreported government operations
ghana Jun 2006 A
Burkina Faso Apr 2007 B+
Moldova Jun 2006 B+
Afghanistan Dec 2005 B
Rwanda Sep 2007 B
tanzania Jun 2006 B
mali mar 2007 c+
mozambique mar 2006 c+
kenya Jul 2006 c
uganda Jul 2006 c
ethiopia 2007 d+
Ukraine Mar 2007 D+
Zambia Dec 2005 D+
Bangladesh Jul 2007 D
Kyrgyz Republic Jan 2006 no score
Pi-27 legislative scrutiny of the annual budget law
Burkina Faso Apr 2007 B+
mali mar 2007 B+
Moldova Jun2006 B+
mozambique mar 2006 B+
Ukraine Mar 2007 B+
ghana Jun 2006 c+
Rwanda Sep 2007 c+
tanzania Jun 2006 c+
uganda Jul 2006 c+
Zambia Dec 2005 C+
ethiopia 2007 d+
kenya Jul 2006 d+
Kyrgyz Republic Jan 2006 D+
Afghanistan Dec 2005 D
Bangladesh Jul 2007 D
Pi-28 legislative scrutiny of the external audit reports
Burkina Faso Apr 2007 C+
ethiopia 2007 c+
ghana Jun 2006 c+
mozambique mar 2006 c+
Zambia Dec 2005 C+
Bangladesh Jul 2007 C
tanzania Jun 2006 c
kenya Jul 2006 d+
Moldova Jun 2006 D+
Rwanda Sep 2007 d+
uganda Jul 2006 d+
Ukraine Mar 2007 D+
Afghanistan Dec 2005 D
Kyrgyz Republic Jan 2006 D
mali mar 2007 d
d-1 Predictability of direct budget support
Afghanistan Dec 2005 A
Burkina Faso Apr 2007 A
tanzania Jun 2006 A
Rwanda Sep 2007 B+
Bangladesh Jul 2007 B
ghana Jun 2006 c+
Kyrgyz Republic Jan 2006 C+
mozambique mar 2006 c+
uganda Jul 2006 c+
ethiopia 2007 c
mali mar 2007 d
kenya Jul 2006 d
Zambia Dec 2005 D
Moldova Jun 2006 no score
Ukraine Mar 2007 no score
28
d-2 Financial information provided by donors for budgeting and reporting on project and programme aid
Burkina Faso Apr 2007 c+
Bangladesh Jul 2007 C
ethiopia 2007 c
ghana Jun 2006 c
tanzania Jun 2006 c
Afghanistan Dec 2005 D+
mali mar 2007 d+
mozambique mar 2006 d+
uganda Jul 2006 d+
Zambia Dec 2005 D+
kenya Jul 2006 d
Moldova Jun 2006 D
Rwanda Sep 2007 d
Ukraine Mar 2007 D
Kyrgyz Republic Jan 2006 no score
d-3 Proportion of aid that is managed by use of national procedures
Bangladesh Jul 2007 B
Burkina Faso Apr 2007 c
tanzania Jun 2006 c
uganda Jul 2006 c
Afghanistan Dec 2005 D
ghana Jun 2006 d
kenya Jul 2006 d
mali mar 2007 d
Moldova Jun 2006 D
mozambique mar 2006 d
Rwanda Sep 2007 d
Ukraine Mar 2007 D
Zambia Dec 2005 D
ethiopia 2007 no score
Kyrgyz Republic Jan 2006 no score
Source: Individual country PEFA assessments.
Notes:
Afghanistan scores have been translated to the A–D ranking from the original 1–4 ranking used.
29
5
APPROAch tO the cASe StudieS
This chapter highlights the different perspectives of different agencies and stakeholders. It emphasises the importance of incentives to provide and use information, and the likely importance of the country context. The country cases were chosen to illustrate a range of such contexts.
5.1 introduction
This chapter describes the study’s approach to the case studies from two angles. First, it provides a brief overview of the perspectives and incentives on both sides of the aid relationship, as these are likely to influence whether and how country budget systems are used. The case studies are an exploration of how these factors interact in a particular country context. The second part of the chapter explains the rationale for the choice of case study countries, and the study’s emphasis on the quality of aid capture.
5.2 Stakeholder perspectives and incentives for aid capture
The capture of aid in government budgets requires action by both governments and donors. Technical features of the budget process are relevant, but effective aid capture depends fundamentally on the perspectives of different actors and the incentives that motivate them.
Donor perspectives and incentives
Influence of aid instrument characteristicsAid instrument characteristics have a strong influence on whether they are likely to be captured in budget documents and budget processes. Budget support is captured (at least in most dimensions) by definition.20 Loans are more likely to be captured than grants, because of associated legal procedures and the need to anticipate the servicing of debt. Technical assistance and other forms of aid-in-kind are less likely to be captured, especially emergency aid, which is less likely to follow the budget cycle. And aid that is disbursed through non-government partners or to sub-national levels of government is also less likely to be captured than aid to the central government.
Influence of aid agency characteristicsAid agencies have different mandates and report to different constituencies, and they operate with differing degrees of flexibility at country level. All these factors may constrain the aid instruments they use. Further, they have different perspectives on aid and different attitudes to risk, including the fiduciary risks associated with using government systems. Although the whole spectrum of aid agencies (along with partner governments) signed up to the Paris Declaration, there are significant differences in how it is interpreted. These different interpretations are explored in the Putting Aid On Budget Literature Review and briefly illustrated in Box 17 below.
20 However, this study found examples where budget support was not made clearly visible in budget documents as a source of finance. Late notification and unpredictable disbursement may also make it difficult to capture budget support on plan.
30
Box 17: Sharing information and using government systems
Among the donors, the differences in approaches to putting aid on budget can be plotted along a spectrum. At one end, there are donors that advocate supplying information on aid flows to the partner country so that it is “reported on” budget. At the other end there are donors that advocate ensuring aid is “integrated on” budget, with country systems used at each stage of the budget cycle.
An example of the first approach is set out by the Millennium Challenge Corporation (MCC). The MCC guidelines for fiscal transparency lay out requirements for reflection without requiring the use of country systems:
MCC will require that, wherever possible, MCC Program activities are reflected in budget documents of the recipient country. … Reflecting the activity in budget documents does not necessarily mean that MCC resources are managed by existing public expenditure systems. (Millennium Challenge Corporation, 2006: 5)
An example at the other end of the spectrum is the Swedish International Development Cooperation Agency (Sida). Sida’s interpretation of the Paris Declaration commitment to put aid on budget is:
In line with the Paris Declaration, Sida shall align to the maximum extent possible with national systems. (Sida, 2007: 27)
Reporting aid on budget and integrating aid on budget are not mutually exclusive activities but they are not substitutes either. Reporting aid on budget is a different activity from integrating aid on budget and reporting will not necessarily result in the same benefits as can be achieved from integration.
Source: Putting Aid On Budget Literature Review.
Differing interests and incentives within aid agencies and governments
The interests of staff working for aid agencies (as with most organisations) may pull in different directions. Headquarters’ attitudes to risks and the use of country systems may differ from the perspective of staff based in-country. And generalists’ priorities and preferences may differ from those of sector specialists. How these potential tensions affect the agency’s use of country systems will depend on procedures as well as policies, and, not least, on the personal incentives facing their staff. For example, a gap between policy and practice may emerge if there is an inconsistency between the policy and the short-term career incentives that motivate staff.
A similar pattern of conflicting interests and tensions applies on the government side. For example, finance ministries and sector ministries are likely to have a different perspective on the merits of funds channelled directly to sectors; central and local government bodies may have a similar divergence; project management units have an interest in perpetuating parallel systems. Box 18 illustrates the interplay between incentives on both donor and government sides.
Effects on incentives for putting aid on budget
The case studies prepared for this report illustrate some of these problems with bringing aid on budget originating from the conflict of interests and incentives that often exists within aid recipient governments. For example, the Mozambique case study reports how efforts to put aid on budget in that country have been hampered by government agencies’ (line ministries, provincial governments, districts, and so on) fear that reporting on budget all the aid they receive will reduce the amount of funding they obtain from the state budget. In other words, a concern that, as the true amount of total funding they manage becomes apparent, the Ministry of Finance might be tempted to reduce budget funding for these projects in the same amount as the aid they receive from donors. Box 19 describes similar perverse incentives in Ethiopia and Uganda arising from explicit attempts to factor aid into overall resource allocations; it also illustrates perverse incentives that can arise from the way budget deficit targets are defined.
The Mozambique country study also reports on how ongoing tensions between the two main Mozambican political parties, Frelimo and Renamo, hamper the quality and extent of parliamentary debates regarding the budget, including issues regarding aid funding of the budget. As discussed in Chapter 6 below, the political framework was found to be similarly important in many of the countries studied.
31
Box 18: Interlocking incentives in traditional aid
Incentives are to be found within both donor agencies and recipient governments that help to maintain and reproduce the vicious circles associated with limited country ownership of development efforts. Case studies of sectors in Mozambique, Tanzania and Uganda suggest the importance of the following:
• Within recipient governments, project modalities with parallel funding and managementmechanisms generate multiple material and non-material benefits for the ministers and civil servants in whose sectors they are located, including salary top-ups, allowances, vehicles, training and travel opportunities, and prestige. Ministers, parliamentarians and local authorities are interested in the political credit they get for attracting a stand-alone project to a specific sector or area. The resource flows from a free-standing project are visible, reliable and relatively simple to control. The government officials at the sector or local-government level prefer to avoid the unpredictability, rigidities and reporting requirements associated with funding through the national budget, and the reduced control this implies. In addition, dealing with a single donor is simpler than dealing with several through a pooled-funding or budget support arrangement, where donors tend to “gang up” on the ministry in ways that reduce its discretion.
• Donor agencies, on the other hand, benefit from the visibility associated with separatelymanaged and “branded” projects. They assist in defending the aid budget to their own parliamentary committees and audit authorities, and in defending the departmental budget within the agencies. In some agencies, it is still the case that staff promotion prospects are enhanced when a particular large project can be presented as the work of an individual or small team. In contrast, where more programmatic, multi-donor ventures are introduced, visibility is lost and the attribution of development results to the particular donor’s support becomes problematic. This concern is especially pronounced when donor agencies are heavily concentrated in a particular field or when a donor is small relative to others in a field. Donor risk-aversion also favours free-standing projects. A standard assumption, which is not always borne out in practice, is that projects can be tightly managed, whereas using government systems to manage projects or programmes calls for a “leap of faith”.
Source: Williamson and Kizilbash Agha, 2008 (reproduced from Booth et al, 2008).
Conclusion
Technical and legislative solutions will not work if they are not sufficiently aligned with the interests of those involved. The complexity of the systems being investigated – the number of interacting factors and interests across many different phases of the planning and budget cycle – makes it likely that general principles will need to be customised to particular country contexts.
5.3 the case studiesChoice of case study countries
Case studies were undertaken in ten Sub-Saharan African countries: Burkina Faso, Ethiopia, Ghana, Kenya, Mali, Mozambique, Rwanda, South Africa, Tanzania and Uganda. A factual analysis was conducted in all of them, with more in-depth studies in five of the countries: Ghana, Mali, Mozambique, Rwanda and Uganda.
This represents a good range of countries, both overall, and among the countries where more in-depth studies were carried out. The countries are from different geographical locations: Central Africa (Rwanda), East Africa (Ethiopia, Kenya, Tanzania and Uganda), Southern Africa (Mozambique and South Africa) and West Africa (Burkina Faso, Ghana and Mali). The countries have inherited distinct administrative and linguistic features, with the group including francophone and lusophone countries in addition to the anglophone ones. There is also a range of experiences of federalism and decentralisation. At one end of the spectrum, Ethiopia is a federal state with devolved regional and district governments. The other countries covered include unitary states that have implemented decentralisation to varying degrees; in several, further decentralisation is under way.
32
Box 19: Examples of perverse incentives
ceilings and offsets: Both ministries, departments and agencies (MDAs) and aid agencies have an incentive not to report aid (fully) if they fear reporting will lead to a reduction in the allocation of domestic finances. This problem has been experienced in Ethiopia where grants to sub-national governments are adjusted to offset aid that flows directly to the sub-national level. More recently, Uganda decided that off budget aid should be included in medium-term expenditure framework (MTEF) ceilings, and experienced a similar problem of under-reporting. (The extent of under-reporting was revealed by a separate stock-taking of donor aid programmes. This highlights the importance of triangulating information when there may be perverse incentives.)
Previously the incentive in Uganda worked in the opposite direction: MDAs would exaggerate likely aid flows in order to maximise the allocation of counterpart funds.
deficit targets: The (unintended) distortions generated by the domestic primary deficit target previously included in the IMF’s macroeconomic and fiscal support programme to Mozambique is an interesting example of how differences in development agencies’ mandates can undermine the process of bringing aid on budget. (See the Mozambique country study.) This indicator, which is typically used to monitor fiscal discipline and sustainability, in the past hindered efforts to increase aid capture on budget by creating incentives to under-report and misclassify aid to the government. There seem to have been two main dynamics:
• First, increases in aid tended to increase the valueof thedomesticprimary fiscal balance,unless these funds were entirely devoted to investment. This gave the government a strong incentive to keep off budget any unexpected surge in externally funded spending that was not already factored into the IMF’s domestic primary fiscal deficit target.
• Second,theprimarydeficittargetcreatedstrongincentivestomisclassifybudgetspending,reporting as investment expenditure the full value of donor-funded projects, even when a significant proportion of these funds was actually funding recurrent spending.
The IMF has since revised its approach to macroeconomic targets to avoid these particular perverse incentives.
Source: Putting aid on budget case studies.
Box 20: ODA: Basic data for the case study countries
Average 2003–05
OdAa
(uS$ m)
Aid / gni (%)
Aid per capita (uS$)
South Africa 694.7 0.3 15.2
Kenya 812.7 4.0 24.3
Burkina Faso 653.0 12.6 51.0
ghana 1,696.2 13.0 78.3
mali 715.0 13.1 54.6
Tanzania 1,017.1 14.9 27.1
uganda 1,256.3 15.9 45.1
Ethiopia 1,890.5 18.8 27.0
mozambique 1,251.8 21.9 64.4
Rwanda 502.6 25.9 56.5Source: OECD DAC international development assistance statistics (DAC Online, OECD.Stat.).
Note: Countries in bold are those for which free-standing country studies have been prepared.
a. ODA = total grants disbursed and loans extended
33
Importantly for this aid on budget study, the countries provide a spectrum of experience across the following dimensions in the area of development assistance:
• Ratio of aid to gross national income: South Africa has the lowest ratio of aid to gross national income (GNI) (0.3 percent). The other countries are all significantly aid dependent countries with eight of the ten countries averaging aid over 10 percent of GNI (only South Africa and Kenya record ratios below 10 percent). Mozambique and Rwanda are the most aid dependent, with ratios of 22 percent and 26 percent respectively. (See Box 20.)
configurations of donors• : Different donors are active in different countries, although the World Bank and the European Commission (EC) tend to be among the large disbursers for most of the countries. All countries are recipients of aid from both bilateral and multilateral agencies. Some of the countries have very congested aid environments (for example, around 40 donors are active in both Mozambique and Uganda).
types of aid modalities in use• : Countries have individual experiences of what share of ODA is provided through different aid modalities. At the same time there are some shared trends. In most of the countries most ODA is provided in the form of grants (for example, 84 percent of ODA in Rwanda, 75 percent in Kenya and 73 percent in Tanzania) while in South Africa, as a matter of government policy, all ODA to the general government sector is in the form of cash grants and technical assistance support. Another trend is the increasing use of budget support. Again with South Africa the exception, in all the other case study countries budget support has become an important aid modality in recent years. For example, general budget support was 26 percent of ODA in 2005 in Burkina Faso, 33 percent in Mali and 34 percent in Ghana.
level of progress on aid management institutions and alignment and harmonisation agendas• : There are some (general) similarities among the countries in the embracing of the Heavily Indebted Poor Country (HIPC) and poverty reduction strategy paper (PRSP) approaches. All have reached HIPC completion point and most are on at least their second generation PRSP, with the exceptions of: Kenya, which did not meet the HIPC indebtedness threshold but has introduced a PRSP as a condition for access to new concessional lending from the IMF and the World Bank; and South Africa, which is not a highly indebted country and has not prepared a PRSP. Similarly, all the countries are involved in progressing alignment and harmonisation agendas, and as part of this all ten countries are taking part in the monitoring of progress on the Paris Declaration indicators. (See the 2006 Survey on Monitoring the Paris Declaration [OECD, 2007a] and the forthcoming survey, 2008.) The countries where detailed case studies were conducted also include several where concerted efforts to bring aid on budget offer scope for learning. For example, in Mozambique in 2005, several donors created a specific on-budget taskforce, and in Uganda, in 2006, a division of labour exercise was launched with one of its aims being to improve the alignment of donor engagement with the implementation of the Poverty Eradication Action Plan.
• PFm reform programmes: All the countries have ongoing PFM reform programmes, but with a range of different emphases and starting points.
Case study approach
The case studies did not involve a new quantitative survey. In each country, researchers assembled the best available information from existing sources and focused especially on trying to explain the patterns of aid capture that were found. The researchers were all already familiar with the countries concerned, and they were assisted by the involvement of the senior budget officers linked to CABRI and of key donor representatives.
Researchers investigated the technical factors involved (notably the characteristics and the quality of public finance management systems) but also explored the interests and incentives involved on both sides of the aid relationship (in line with the discussion at the beginning of this chapter). Researchers considered the different dimensions or phases of aid on budget (Box 3), and for each phase tried to assess the quality of aid capture. They adopted the perspective that “putting aid on budget is the joint effect of donor and government practices, and where there is poor quality of aid capture, this is the compound effect of both donor and government shortcomings”(Putting Aid On Budget Literature Review: para. 51). Moreover, whether information is successfully captured depends on the demand for information as well as on the technical capacity to supply it.
34
The simultaneous attention to demand and supply, quality and incentives is illustrated in Box 21. It was also reflected in the reporting format adopted. For all ten countries, researchers prepared summary reports in a matrix format which, for each dimension of aid capture, considered:
the evidence of capture (what is captured?)•
the quality of capture (how well? how useful?)•
explanations (why/why not?).•
The resulting case study matrices are presented in full as the Annex to this report (Summary of Country Study Findings).
Box 21: Factors behind quality of aid capture in the budget
Source: Mokoro Ltd, based on OECD, 2007a.
As criteria of quality, researchers were asked to consider:
the technical quality of information provided (its completeness, credibility, disaggregation and detail) •
timeliness and accessibility (Who has the information, and at what stage in the relevant process?)•
whether and how information was actually used (Does the information influence decisions and choices? •Is there feedback to providers and does this influence future budget rounds?).
It was expected that the ability to make good use of information about aid would largely depend on the quality of the information and processes relating to non-aid resources.
Weak supplyof information by donors
Donors are not always attentive to getting information on intended
disbursements to the budget authorities in good time and in a usable
form, resulting in systematic under-inclusion of aid in budget.
Lack of realismin information provided by donors
When donors do provide such information, they are not always realistic about their
ability to disburse on schedule, resulting in a tendency to over-estimate some kinds of
aid flows and under-estimate others.
Weak demandfor information by governments
Budget authorities are often not strongly motivated or equipped to capture information on donor dibursement
intentions, or to make realistic estimates of shortfalls, resulting in both over- and under-
counting on quite a large scale.
35
6
leSSOnS FROm the cASe StudieS
This chapter highlights findings from the case studies researched for this report. The principal findings from all ten African case studies are summarised in the Annex to this report (Summary of Country Study Findings), and separate reports have been published on five of the case studies. This chapter therefore focuses on the key findings that have wider relevance.
6.1 introduction
Experience of putting aid on budget in ten African countries was specially researched for this report. This chapter highlights key findings that seemed especially relevant to the identification of future good practice. Findings are discussed under the headings of each dimension of aid capture. However, this is simply a convenience for the purposes of presentation. As the findings themselves will illustrate, there are many overlaps and interactions between the different dimensions. The practical relevance of each dimension is considered first, the main factors that influence the capture of aid are then noted, and finally the good (and bad) practices are identified. This is followed by a cross-cutting discussion of decentralisation.
6.2 Putting aid on planThe relevance of putting aid on plan
Good practice calls for annual budgets to be guided by medium-term strategies. Where aid is a significant proportion of resources, it clearly needs to be factored into the government’s medium-term plans. Being able to anticipate the aid that will augment or substitute for domestic budgetary resources in the coming budget year is a minimum requirement. Planners also need to take a view on the availability of aid and other resources over a much more extended horizon. The information requirements for the short and the medium term, and for different modalities, are different. Aid needs to be included (broadly) on the resource side, even if it can’t be allocated (specifically) on the expenditure side. As the Uganda country report notes:
… in the [PRSP] and strategic plans, the capture of aid in very general terms is largely appropriate. Project aid in the overall MTEF, which is presented by sector, is helpful in facilitating high level inter- and intra-sectoral decision making, whilst budget support is appropriately captured as a revenue only. (Putting Aid On Budget: Uganda Case Study)
Factoring aid into the planning process should influence the choices of both governments and donors. The Rwanda country report highlights the difference between a data collection exercise and a true planning exercise. The South Africa researcher noted a tendency for information about government plans to inform aid choices “rather than the reciprocal using of aid information to inform budgeting choices”.
36
Factors influencing aid on plan
Quality of government planning processTo bring aid on plan in a meaningful way depends on the government itself having a meaningful planning system. Key elements of effective planning include a transparent planning process with a clear calendar, realistic costing of alternatives, and, ideally, a medium-term expenditure framework (MTEF) that sets annual budgets within a longer perspective. If there is not a credible government planning process, it is difficult to make systematic use of information that donors provide. For donors, it is easy to “align” with objectives that lack the discipline of prioritisation. In practice, the case study countries exhibit varying degrees of sophistication and discipline in their planning; MTEFs are at varying stages of development, and the quality of the plan–budget link may vary a lot among sectors.
Quality of capture also depends on the capacity to use the available information. This is a matter of systems as well as the ability of the people who staff them. A clear budget calendar is an obvious system element, since the timing of information greatly affects its utility. It is also important to have strong sector planning processes.
Poverty reduction strategy papers and sector approachesA general finding is that poverty reduction strategy papers (PRSPs) have strengthened the common focus of government and donors, although PRSPs vary in their quality as planning documents. There are continuing efforts to link PRSPs to more disciplined MTEFs. Sector approaches have helped to strengthen sector planning and the focus on capturing aid at sector level. Indeed, aid is integrated successfully only when it is integrated well at sector level. Coordination belongs at the central level, but integration occurs at the sector level. If sector-level processes are weak, the capture of aid on plan and subsequently on budget will be ineffective.
Hence, sector approaches and sector working groups can be very important. Their role was highlighted, for example, in the country studies for Mozambique, Rwanda, Tanzania and Uganda. However, not all working groups were effective in including donor projects in sector planning and budgeting. In some cases their focus was exclusively on policy, or they lacked effective participation from the government or from key donors. Donors that did participate did not always provide useful information about their programmes. On the positive side, sector wide approaches (SWAps) were often found to have developed and supported technical analysis capacity as well as contributed to dialogue and the flow of information for planning purposes. It is a challenge to ensure that sector processes (including public expenditure reviews – PERs – in Uganda, Tanzania and Kenya) are equally owned by government and donors; and that sector processes are fully articulated with overall plan and budget. (Uganda is a strong example of this kind of articulation).
Bringing different modalities on planAs with other dimensions there is a hierarchy among aid modalities in their likelihood of being captured on plan:
In principle, general budget support (GBS) is fully integrated with domestic resources. In practice, •even for GBS, the quality of capture on plan may be undermined by unpredictability. At the same time, budget support helps drive the capture of other modalities: the dialogue around GBS creates a demand for coherent medium-term planning documents, and also reinforces efforts to strengthen PFM generally (thereby influencing all modalities and dimensions).
As already noted, sector approaches often mean that planning for the sector is more systematic. Like •GBS, they provide a focus for government–donor dialogue that can improve the quality of planning. Sector budget support and pooled funds are usually very visible for planning purposes (though some pooled funds have followed parallel planning processes and, as discussed later, pooled funds may be off budget in many dimensions).
Projects are more likely to be on plan if financed by loans than if financed by grants. Projects themselves •take many forms. The planning (and financial) implications of large-scale infrastructure projects are quite different from smaller and more adaptable interventions. Separate “investment plans” and “development budgets” are traditional devices for bringing projects on plan and on budget, but may cause problems in disconnecting aid from domestic resources. (See further discussion below, especially Box 27.)
The forms of aid least likely to be captured on plan are technical assistance and other in-kind aid (some •in-kind aid is emergency in nature and therefore difficult to plan for in detail), and aid that is disbursed
37
via non-government bodies. Aid provided directly to decentralised levels of government is also harder to capture.
The paradox of aid agency projectionsThere are systematic biases in the information provided by aid donors. They are very cautious about providing aid forecasts that are not based on firm commitments. This leads to a paradox. Donor projections of disbursements within the budget year tend to be too optimistic (because donors rarely meet their commitments in full21). At the same time, donor projections of aid over the medium term tend to be too pessimistic, because donors are unwilling to make projections above what they have formally committed to, even when they know that new projects and programmes are likely to appear. Box 22 uses Uganda to illustrate this problem.
Box 22: Paradoxical aid projections (Uganda illustration)
Donor aid projections typically overestimate short-term disbursements and underestimate medium-term aid flows. This paradox is well illustrated in Uganda. The Uganda case study notes:
Quality of aid data: Aid data provided by donors is largely based on financial commitments, and donors are either unwilling or unable to provide medium-term estimates of aid other than what has been committed. This provides two problems. Firstly that aid data provided by sectors and by donor agencies for the budget year are usually fairly comprehensive, but over-optimistic – i.e. in aggregate not all donor commitments are realised. Secondly aid projections for the outer years usually underestimate future levels of aid, as existing commitments tail off over the medium term. The diagram below shows data compiled under the Division of Labour Exercise and the (undiscounted) Macroeconomic Department data compiled directly from donors. Both show a significant tailing off of aid over the medium term
(2006/07 was the budget year).
comparison of aid projections at the time of uganda’s 2006/07 budget
At the aggregate level, for the purpose of macroeconomic management the Ministry of Finance tries to make realistic projections of aid, by discounting aid for the budget year in question (different discounts are used for budget support and projects), and making realistic projections of budget support (i.e. projecting it above committed levels) over the medium term. The latter is an inexact science, especially because of the lack of realism in aid data from donors and sectors.
Meanwhile the allocations which appear in the Annual Budget – the Estimates of Revenue and Expenditure – are based on annual donor commitments, rather than the discounted numbers which appear in the Macroeconomic Framework and the MTEF. This is because it is difficult to predict which individual project or programme is likely to under-perform.
Source: Putting Aid On Budget: Uganda Case Study (diagram reproduced from ODI: 2007).
21 Even if committed funds are fully available, expenditure may fall short, especially if funds arrive late in the budget year.
0
200
400
600
800
1000
1200
1400
1600
2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10
Financial Year
US$
Mill
ion Financial Data Tool (Division of Labour
Exercise)
GOU Budget/Development ManagementSystem
Macroeconomic Dep't (incl. BS fromFDT)
Financial data tool (division of labour exercise)
GOU Budget/Development Management System
Macroeconomic Department (incl. BS from FDT)
38
Good practice considerations
Progress in strategic budgetingThe government’s own progress in strategic budgeting is a key factor in bringing aid on plan. Although long-term planning has to address trade-offs between sectors, useful progress can be made at the sector level. Government leadership of the process is crucial. The principles of strategic budgeting are general, but where aid is prominent, it is important:
not to take donor information at face value, but to use it as the basis for scenario planning – see Box 23•
to engage donors with a government-led planning process – see Box 24.•
Box 23: Putting aid on plan: Realistic projections and scenario planning
A number of governments have recognised the biases and uncertainties in donor projections of their own aid flows – see Box 22 – and are responding by discounting estimates of short-term disbursements, and developing their own alternative scenarios of aid flows in the medium and long term.
discounted budget estimates: The Rwanda study provides a good example and rationale for discounting:
Because some projects have a poor record of project execution and in the past the Development Budget tended to be often “under-executed”, expected project disbursements are discounted (at variable rates) in the budget preparation process by the Budget Department. Amounts included in the Development Budget are therefore often lower than the disbursement estimates provided by project coordinators (who, for opposite reasons, have shown in the past a tendency to overestimate expected
disbursements). (Putting Aid On Budget: Rwanda Case Study)
Uganda has used discounted estimates of budget support disbursements, based on experience, to avoid authorising unfunded expenditures. Ghana also applies discounts.
Scenario planning: Several countries have published different funding scenarios for their PRSPs/medium-term expenditure plans. Ethiopia describes different expenditure scenarios for different levels of donor funding. Mozambique’s most recent PRSP is based on three expenditure scenarios. Because of the unpredictability of its budget support, Kenya does not include budget support in its main budget but only in an alternative scenario expenditure plan. In Tanzania, government and donors are working towards the joint definition of alternative scenarios.
Source: Putting aid on budget case studies for Rwanda, Ghana, Ethiopia, Mozambique, Kenya and Tanzania.
Box 24: Engaging donors in medium-term planning
Medium-term planning processes in Uganda, Tanzania and, recently, Kenya, seek to engage donors transparently in analysis and review linked to planning and budgeting. In Tanzania, for example:
The planning and budgeting process structures require the participation of donors and other stakeholders in sector and thematic working groups. The wide use of sector SWAps supports the inclusion of aid in a more meaningful way on plan. The quality varies across sectors depending on how well structures work
and the analytical capacity in the sector. (Putting Aid On Budget, Tanzania Case Study.)
A positive element in Uganda is the transparency of the process and the availability of sector and national budget framework papers to stakeholders in sector working groups and across government. Kenya also publishes all sector and national framework papers.
In all three countries, efforts are made to fit systematic public expenditure reviews into the planning and budgeting calendar.
The multi-donor budget support (MDBS) process in Ghana is another example of effective engagement between the government and donors (and a government-led process).
Engagement in such processes makes it more likely that information (in both directions) will be provided, queried and used.
Source: Putting aid on budget case studies for Uganda, Kenya, Tanzania and Ghana.
39
Drawbacks of fragmented planning and budgeting systems
Separate development budgets can be seen as a device for keeping aid on plan and on budget while deferring to aid agency wishes for their funds to be separately identified. Over time, however, “investment budgets” often degenerate into “aid budgets”; so that aid appears in the development budget even if it is financing recurrent costs. The disadvantages of dual budgeting are exacerbated when responsibilities for development and recurrent budgets are divided between separate planning and finance ministries. In many countries, some of the disadvantages of dual budgeting persist, but there is scope to mitigate them by adopting more integrated budget classifications, documents and processes. For example, in Rwanda, the recurrent/capital split is being mitigated by a new chart of accounts, and all ministry expenditure is being shown by ministry instead of in a separate development budget. (See also Box 27.)
6.3 Putting aid on budgetThe relevance of putting aid on budget
Annual budgets are a universal management tool. The formal budget document is the basis of legislative authority for expenditure, and subsequently the basis for controlling budget disbursement and monitoring budget implementation. However, there may be deviations between a formally approved budget and what happens subsequently. The budget documents are only a starting point in discovering whether, and how well, aid is incorporated in the formulation, authorisation and implementation of government budgets.
Budgets should show the sources of finance for public spending as well as the details of expenditure. Both aspects are important in giving the legislature a full picture of the expenditure choices it is being asked to endorse. As noted in the discussion of bringing aid on plan, it is important to factor available aid resources into strategic planning at an early stage – before their detailed uses have been determined. In an annual budget document, the extent to which uses are specified will depend largely on the aid modality. Earmarked project funding may be tracked to specific expenditure items in the budget, while general budget support is treated in the same way as most domestic revenue – it is not tagged to specific expenditure items, but underpins the budget as a whole.
Factors influencing aid on budget
Incentives and disincentives for including aid on budgetOn the donor side, the most common deterrents to putting aid on budget are the overall credibility of the budget, and the specific visibility of donor funds within it. Credibility is undermined by deviations between budgets and actual out-turns. (See Box 25.) The desire for visibility has a fiduciary aspect (as a means to show that funds have been used as agreed), but may also relate to an agency’s broader need to be able to demonstrate results – that its aid has made a difference. Thus, government classifications which make it hard to identify different uses and different donors tend to discourage aid from coming on budget.
On the government side, there are differences among case study countries in their attitude towards incorporating aid on budget. Some proactive governments, such as Rwanda, take the view that aid should appear in the national budget only if meaningful national discretion is applied. Others, such as Ethiopia, have legislation which requires them to show all aid in the budget, even if it does not pass through government disbursement procedures. (And, as already noted in Chapter 5, MDAs may have an interest in protecting their discretion by keeping off the budget aid that their sectors receive directly.)
On the government side too, there are concerns about credibility (will promised donor funds arrive?) and practical concerns about compatibility of information systems (is donor information provided on time? and in a format compatible with the government budget – for example, in terms of detailed classification and time period?). Looking further ahead, will information for the ex ante budget be matched by timely information on actual expenditures?
In many cases it is clear that, while efforts to include aid could be quite onerous for both parties, the failure to include aid on budget did not have significant consequences for those directly involved. For example, there were cases where donor officials never checked whether their aid appeared in government budgets or not, implying that its inclusion in the budget or not made no practical difference to them.
40
Box 25: Deviations between budgets and actual spending
Credibility of budgets is recognised as a key criterion of effective public finance management (reflected in the first four PEFA indicators – see Box 6). In several of the case study countries, large deviations between approved budgets and actual out-turns have undermined donor confidence and provided a rationale for not putting aid on budget.
Budget discrepancies may occur in different ways:
• Cashmay fall short,with the result thatbudgeted fundsarenot released.This is commonin countries that have cash rationing systems, sometimes necessitated by macroeconomic instability, but often because fiscal forecasting is not reliable.
• The initiallyapprovedbudgetmayberadically revisedduring implementation.Expendituredemand in some areas may be more than budgeted, resulting in funds being deviated from a budgeted area of spending to areas that experience shortfalls. The demand may be more on account of lack of discipline in budget execution, when activities are undertaken which were not in the plan, although they are not authorised by the budget and are not unforeseen or unavoidable. It may also be on account of changes in the external environment, such as natural disasters or other emergencies. Often, however, it is because budgeting itself did not estimate properly what the cost of programmes will be, or because changes to key cost drivers (such as salaries) take place out of step with the budget process.
• Somediscrepanciesmayreflectthewaysaidismanaged.Budgetedaidfundsmaynotarrive(with consequences for domestic counterpart funds as well), or government may fail to meet trigger conditions for their disbursement. Aid expenditures via Channels 2 and 3 may be under-recorded if aid agencies or project implementation units do not supply expenditure data on time and in compatible formats.
Source: Putting aid on budget case studies.
Approaches to different aid modalitiesDifferent aid modalities tend to be treated differently in terms of efforts to incorporate them on budget. Loans and credits are likely to be incorporated, because of the need to budget for debt service, as well as legal stipulations. Budget support inherently has to be incorporated (though there are striking differences in the quality of information provided about budget support as a source of funds; in one case – Mali – it was not explicitly noted in the ex ante budget). In some countries (Ghana, Mali) being off budget is taken as an inherent characteristic of projects. In Mali’s case this is partly a reflection of a francophone tradition that, in effect, excludes from the main budget any expenditure that is not also on treasury. Such restrictive assumptions are unhelpful as they preclude active consideration of how project aid might be brought on budget, and may simultaneously result in very restrictive interpretations of sector budget support. (See Box 26.)
Good practice considerations
Priorities for PFM reformPFM reform programmes should seek to increase the willingness of partners to bring aid on budget by increasing the budget’s credibility, as well as facilitating the practical process of incorporating aid in the budget. Key (interlocking) areas for PFM strengthening to bring aid on budget thus include:
charts of accounts and budget classifications that enable sources and uses of funds to be clearly identified •in the budget and monitored during budget execution
measures to ensure discipline in budget management (so that there is close correspondence between •budget and out-turns – see Box 25)
a planning and budgeting calendar that shows when donor inputs are required, complemented by clear •lines of communication within government and between government and aid agencies.
41
Box 26: “Sector Budget Support” in Mali
As discussed in Chapter 2, according to the standard definitions, one of the characteristics of budget support is that it is not linked to specific project activities. In Mali, however, this does not apply to the modality known as sector budget support. As described in the Mali case study:
The recent increase in sectoral budget support has come about with the switch from pooled-funding arrangements to sector budget support, largely from donors who previously had contributed to the pooled-funding arrangements.The detailed tables by ministry include a line item separately identifying expenditures financed by sector budget support (separately for each separate donor providing sector budget support). Thus, through the relevant classification codes, it is possible to track the use (execution) of sector budget support by economic item.In theory, sector budget support should be planned in the same way as domestic resources, with the only difference being that the relevant sector ministries are supposed to achieve particular results in accordance with the accompanying sector policy matrix. In Mali, these resources are in practice treated differently to domestic resources; in effect, they are treated as if they were project resources since they are planned separately, their use is separately identified as line items and tracked in the budget, and in budget execution reports. Furthermore, the resources are directed towards particular purposes (e.g. training workshops), which may or may not be how these resources would have been used had they
been planned together with the rest of the ministry’s resources (as opposed to planned separately).
Although this approach does not conform to the generally accepted definition of budget support, it does suggest a mechanism by which project aid in Mali might be brought on budget.
Source: Putting Aid On Budget: Mali Case Study.
Joined-up budgeting – recurrent and development budgetsSeparate development budgets may make aid more visible, but often at the expense of fragmenting the planning and budget system. Rwanda’s reforms seek to integrate recurrent and development budgets in a way that still enables sources and uses of aid funds to be clearly identified. (See Box 27.)
Some other countries with dual budgets have not gone as far in integration as Rwanda plans to. However, even if development and recurrent budgets are presented separately, the links between them can be strengthened by applying a common classification system. Another important mitigation is to ensure that the same administrative unit prepares both budgets (at central and sector levels). This can reduce the problems of integrating project and programme aid into budgets.
Box 27: Integrating recurrent and development budgets: Rwanda’s approach
Until the 2007 budget, all development projects were classified in the “development budget”, using a specific economic classification which did not allow the separation of recurrent and capital expenditure. (75 percent of the development budget was externally financed.)
Major reforms in the budget presentation should start being implemented with the 2008 budget. Domestically and externally financed expenditures will be jointly presented under each ministry, instead of having a separate development budget covering donor projects. The same chart of accounts will be used to classify both domestically financed expenditures and donor projects.
The 2008 budget will start using the new chart of accounts which no longer provides for a separate “development budget”, but clearly separates recurrent and capital expenditures. As a result, project expenditures will need to be classified according to the same chart of accounts as the rest of the budget, i.e. separating recurrent and capital expenditure.
The integration of projects into the national budget will also require training of project coordinators and line ministries to ensure that the new budget classification is applied appropriately to projects, and sub-dividing big, multi-sectoral projects into project components in order to ensure they are accounted for where they belong (sector, ministry, programme) instead of being lumped in one place.
Source: Putting Aid On Budget: Rwanda Case Study.
42
Showing how aid has been usedKeeping aid off budget is one strategy for an aid agency that wishes to ensure that its aid is separately identified and channelled to specific activities. However, such earmarking can be accommodated on budget if the budget codes are able to identify both the sources and the uses of funds in detail (and if the regular accounting system provides timely and credible reports in the same format).
For many donors, however, earmarking is not an end in itself. They have a broader concern that their aid should add to the total resources addressing particular objectives (for example, to achieve the Millennium Development Goals [MDG] for health or education). A number of countries have adopted broader tracking mechanisms which, in the context of dialogue with aid partners, help to demonstrate the focus and the additionality of aid, without requiring programmatic aid to be strictly earmarked to particular budget items. For example:
In Uganda a notional Poverty Action Fund has been used to channel HIPC resources and budget support to •priority sectors, and to identify budget lines that are protected from within-year budget cuts.22
In Ethiopia, the Protection of Basic Services (PBS) project supplements federal grants to regions and districts, •enabling them to maintain and expand primary health, education and other services. Aid agencies and the federal government jointly monitor the additionality of PBS funding, the fairness of sub-national allocations, and regional/district performance, but without compromising local authorities’ discretion to manage their own budgets.
In Tanzania, the Strategic Budget Allocation System (SBAS) links budget spending not to priority sectors in •the poverty reduction strategy (PRS), but to priority outcomes.
These mechanisms bring more aid fully on budget while allowing for risk management regarding effectiveness and fiduciary concerns.
Making budget information more accessibleInformation contained in the budget is often obscured by poor presentation. Country studies found that information on aid as a source of funding was often too highly aggregated to be checked. The division of information between recurrent and development budgets may make it difficult to see an overall picture (for example, at the sector or agency level). Budget documents may present masses of line item detail without analytical summaries or narrative explanation. Poor presentation frequently undermines the translation of strategic planning for aid into annual budgeting, and impairs the quality of capture of aid on budget.
6.4 Putting aid on parliamentThe relevance of putting aid on parliament
The strict interpretation of putting aid on parliament is that external financing should be included in the revenue and appropriations approved by parliament. Parliament is legally responsible for authorising expenditure by the executive branch, and inclusion of aid within the appropriated budget signifies that the government should be accountable to parliament for aid-funded expenditure.
In practice, however, aid may be brought to parliament’s attention in ways that fall short of formal appropriation. The country case studies therefore reviewed the overall role of parliaments in the budget process, checking the extent to which parliaments are informed about aid, and the extent to which aid is subject to parliamentary debate and review, even if not formal appropriation. Three levels of parliamentary involvement may be considered: at the level of the strategic framework for budgets, at the level of approving annual budgets and particular aid agreements, and at the level of reviewing audit reports. (See the discussion under “Putting aid on audit” on page 47.)
Factors influencing aid on parliamentScope of parliamentary approvalsThe case study countries varied in the extent to which external finance was incorporated, and made visible, in the budgets submitted for parliamentary approval. This partly reflected differences in legal frameworks, but also different degrees of diligence in following the legal requirements. Even in countries where parliament explicitly
22 The Ghana multi-donor budget support (MDBS) has also been used very effectively to channel donor funds to addressing poverty reduction. See the 2007 evaluation of Ghana MDBS (Lawson et al, 2007).
43
does approve project aid (and is not simply informed about it), parliamentary discussion tends to be limited. In Rwanda’s case, for example:
Parliament approves the Budget Law as a whole, comprising budget support and the donor-financed projects presented in the development budget. The discussion in parliament nevertheless focuses on recurrent budget and on internally financed development budget. The only aid modality on which Parliament has a real say is therefore budget support. In practice, although the externally financed part of the budget goes to parliament, it only includes projects that have been signed already, therefore projects that parliament cannot modify. Regarding loans, as stated in Article 54 of the Law on State Finance and Property, each of them should be approved by parliament. Therefore, while the Minister of Finance is requested to approve all externally financed support, parliament approves only loans. (Putting Aid On Budget: Rwanda Case Study)
Quality of parliamentary reviewThe quality of parliamentary review was generally weak. This applies to all resources, and reflects weaknesses in the capacity and staffing of parliaments themselves, characteristically rushed timetables for review, poorly presented budget documents, and, often, a political environment in which proposals presented to parliament are not strongly contested (for example, because of strong party cohesion across the executive and parliament23). The Kenya researcher noted a general finding:
Lack of parliamentary scrutiny of proposals regarding the use of external funding has less to do with transparency on the proposed use than with weak and even inappropriate engagement of parliament in the budget process, and weak capacity. (Putting Aid On Budget: Kenya Case Study)
Mozambique presents a similar scenario, where capacity constraints have limited the extent and quality of parliamentary debates on aid and the budget. The Mozambique case study argues that these problems have been heightened by the little attention that parliament has received from donor agencies in the past decade:
... the quality of both the approval and review of the budget and its accompanying documents is generally considered weak (see, for example, de Renzio and Sulemane, 2006). This is partly because parliament’s lack of technical, human and financial resources undermines its ability to undertaking a thorough analysis and control of budget proposals and execution, including in its external component. While some donors are, directly or indirectly, supporting capacity building initiatives for the Mozambican parliament, the small size of these projects and parliament’s very weak technical and human capabilities base suggest there is significant scope to work at this level and, this way, improve parliamentary oversight of the national budget, including in its external component. (Putting Aid On Budget: Mozambique Case Study)
Another important factor identified in the Mozambique country study is the political context in which parliamentary activities relating to the budget (such as budget approval, parliamentary expenditure control, and so on) take place:
In addition to parliament’s weak capacity, the “politicisation” and the underlying tensions between the (FRELIMO) government and the main opposition party, RENAMO, impede an in-depth and more objective examination by this chamber of the government’s budget proposals and subsequent budgetary performance. Ultimately, these tensions end up by affecting the quality of parliamentary budget debates. (Putting Aid On Budget: Mozambique Case Study)
Good practice considerations
Reasons for weak parliamentary engagement in budgeting are deep-seated and not likely to be quickly transformed. Issues with whether and how aid is “on parliament” have little to do with how aid is managed, but rather with how the budget as a whole is managed. Nevertheless there is merit in:
ensuring that legal requirements concerning parliament’s role are observed by governments and donor •partners alike
23 This is a factor in many donor country parliaments too.
44
ensuring that parliament is informed about aid flows even in cases where it does not formally approve •them
strengthening the engagement of parliaments in the budget process generally, for example, by:•
engaging parliament in the discussion of medium-term strategieso
improving the accessibilityo 24 and transparency of the budget itself and of related documents
strengthening parliamentary procedures and staffing (for example, strengthening the staffing of o
specialist committees)
supporting civil society efforts to hold governments accountable for public expenditure (which may •include strengthening the role that political parties play in parliamentary budget processes)
seeking accountability to donors in ways that complement, and do not undermine, governments’ •domestic accountability, including accountability to parliament.
6.5 Putting aid on treasuryThe relevance of putting aid on treasury
Aid is on treasury when “external financing is disbursed into the main revenue funds of government and managed through government’s systems” (definition in Box 3). The rationale for putting aid on treasury is to support government’s financial management systems and capacity, to facilitate efficient cash management and reinforce financial discipline, and to facilitate aid capture on other dimensions. Not all funds channelled to government bodies are on treasury. (See the distinction between Channel 1 and Channel 2 disbursements, also explained in Chapter 2.) Equally, aid may be on budget without being on treasury. However, on treasury is a pivotal dimension, both because of its direct importance and because of its influence on other dimensions: thus, when aid is disbursed through the treasury system, the treasury has to be provided with necessary information to trigger disbursements, and this supports the on budget and on accounting dimensions too.
Factors influencing aid on treasuryIncentives and disincentives for including aid on treasuryDonors’ reliance on treasury systems for disbursement of aid depends on their assessment of the financial risks involved, as well as the technical capacity of the treasury system to meet other donor requirements (which, as discussed under “on budget”, may include the separate identification and earmarking of a donor’s funds). Thus, many of the factors discussed under “on budget” are also relevant to the on treasury dimension. In addition, there are frequently concerns about the efficiency, not just the probity, of treasury disbursement procedures. MDAs, as well as donors, may prefer parallel disbursement systems if these are considered more likely to ensure (timely) delivery of funds to particular sectors or sub-national levels. In practice, concerns about public procurement standards act as a deterrent to putting aid on treasury, although aid can be, and frequently is, on treasury while following special donor procurement procedures.25
Approaches to different aid modalitiesAll forms of budget support are, by definition, on treasury. On the other hand, the case study countries included several where the use of treasury systems for project aid disbursements is minimal. As noted, this is treated almost as a matter of definition in Ghana, and as a corollary of the legal framework in the francophone countries. In Uganda, no system for Channel 1 disbursement by aid projects has been introduced. (This is despite the fact that, since 2003, parliament has legally approved project aid as part of the budget, and seems to reflect government’s preoccupation with budget support as the preferred modality for incorporating aid fully into government systems.) Tanzania, in contrast, has deliberately created a channel to disburse project aid through the treasury system.
Several countries have made notable gains by bringing pooled sector funds on treasury that were formerly managed in parallel. This has been a focus of joint government–donor efforts to bring more aid on budget in Mozambique,
24 “Accessibility” implies user-friendliness and not just availability of documents.25 South Africa’s RDP fund provides a special case: funds disbursed through the RDP fund are thereby on treasury, but as they are not being disbursed through the
central revenue fund, their procurement procedures can be specified in the aid agreement. More generally, earmarked aid that is on treasury may have special procurement stipulations attached.
45
and also in Tanzania. Pooled funds have also been moved on treasury in Burkina Faso and Mali. (See Box 26.) In Ethiopia, Channel 2 disbursements have largely given way to Channel 1 (with and without earmarking).
Good practice considerationsPriorities for PFM reformThe same features of PFM systems that are important for bringing aid on budget are also relevant to bringing aid on treasury. (See discussion under “Good practice considerations” on page 43.) In order for donors (and beneficiaries) to have confidence in treasury disbursement, it is particularly important to address both the discipline and the efficiency of treasury disbursement. Measures that have been effective in the case study countries include the strengthening of expenditure processing linked to reformed budget classifications, and capacity building. In some cases (for example, the SISTAFE system in Mozambique) the introduction of integrated financial management information system (IFMIS) modules has contributed to the strengthening of treasury systems.26
Across a range of countries, the introduction of single treasury accounts (STAs) and related efforts to strengthen bank account management have had a strong effect. The Mozambique country report provides an example. (See Box 28.) Ghana and Rwanda are among other countries pursuing similar reforms. In Uganda, as a separate measure, the government has required bank accounts for project aid to be held at the central bank, rather than commercial banks; this has improved information and supervision by the accountant general.
Box 28: Bringing aid on treasury through the single treasury account in Mozambique
The Mozambique government, together with its aid partners, has been working to merge all financial accounts into one single treasury account – Conta Única do Tesouro, CUT – and to bring all available sources of funding, both internal and external, under the umbrella of the Mozambican integrated budget and PFM system. Donor agencies that belong to the Programme Aid Partners group have agreed to disburse at least 60 percent of its development assistance to the government through the single treasury account system by 2009.
Since 2007 some of the larger sectoral common funds are now being managed through the CUT and are, therefore, on treasury. These include common funds in agriculture, health and education (as well as a fund that supports PFM reform!).
The establishment of the single treasury account has been instrumental in the process of putting aid on budget in Mozambique and bringing its associated PFM benefits. Its creation directly addresses one of the main causes behind the phenomenon of off budgets: the lack of transparency within the treasury system, which deters donors from using local public financial management systems. At the same time it is perhaps one of the most visible manifestations of the PFM benefits brought by aid capture on treasury, given the immediate benefits the government/treasury draws from having the number of accounts it uses for its treasury operations reduced to one.
Source: Putting Aid On Budget: Mozambique Case Study.
Donor practicesDonors can contribute to more effective aid capture on treasury in a variety of ways:
by giving early attention to the design of disbursement procedures when preparing aid interventions, •and taking account of the nuts and bolts of the partner country’s PFM system when doing so
by bringing existing pooled funds on treasury, and designing new pooled funds to be on treasury from •the outset
by improving the predictability and front-loading of budget support disbursements (to assist government’s •liquidity and cash-flow management).
26 However, the introduction of an IFMIS is not a panacea. Research reported in 2003 (Dorotinsky, 2003, cited in Parry, 2005) showed that for 34 IFMIS projects in 27 countries funded by the World Bank (at an average of US$12.3m per country): only 21 percent were successful; only 6 percent were regarded as likely to be sustainable; and, they took an average of seven years to complete.
46
6.6 Putting aid on accountingThe relevance of putting aid on accounting
Aid is on accounting when “external financing is recorded and accounted for in government’s accounting system, in line with government’s classification system” (definition in Box 3). Aid is not fully on accounting if special additional financial reports are stipulated. (See discussion under “Putting aid on report” on page 49.) The rationale for using the government accounting system as much as possible is to strengthen government financial systems, to minimise transaction costs and demands on government capacity, and to ensure that government accounts provide a comprehensive view of the government’s finances.
Aid may simultaneously be on accounting and not on some other dimensions. For example, there are ways in which aid that is not on treasury, or not appropriated through the budget, may still be recorded in government accounts. However, there is valuable synergy if aid is on treasury as well as on accounting, since the need to justify treasury disbursements provides a strong incentive to keep government accounting records up to date. Credible and timely accounting for disbursements may be a crucial factor in aid agency willingness to use Channel 1.
Factors influencing aid on accountingIncentives and disincentives for including aid on accounting Whether aid is brought on accounting is largely a corollary of the incentives for being on treasury, but the incentive works in both directions, as reliable accounting is a prime concern in either case. Donor preferences are strongly influenced by the issues in budget classification and charts of accounts that have already been discussed, but a crucial factor is the ability of the government accounting system to deliver reliable and timely expenditure reports. In several of the case study countries, large backlogs in closing government accounts, as well as delays in reporting, have been a major problem, but impressive strides have been made in reducing backlogs and strengthening accounting systems as part of PFM reforms and capacity development.
Approaches to different aid modalitiesBudget support is inherently on accounting, and its use gives donors a greater interest in the quality of government accounting as a whole. PFM reform programmes and conditions linked to HIPC and budget support have been instrumental in strengthening government accounting systems.
Government approaches to putting project aid on accounting are influenced by different legal traditions. Mali illustrates a francophone approach, where the responsibility of the financial controller is expressly limited to the accounting of funds that are on treasury; apart from budget support, aid funds are disbursed by project implementation units and accounted for separately, while producing consolidated accounts seems not to be a priority for government or donors. Tanzania has, more recently, made strenuous efforts to bring all modalities on accounting. It has arrangements whereby funds not on treasury are on account (because of pooled funds with parallel disbursement systems, but government accounting); the same mechanism can incorporate aid-in-kind into the accounts. (See Box 29.)
Good practice considerations
Since disbursement through treasury is not practical for all donors (or appropriate for all types of aid), mechanisms to record aid-in-kind in government accounts are a good practice. (See Box 29.) (Kenya also tries to capture aid-in-kind, in line with its financial legislation). More fundamentally, there has been a major contribution from the accounting reforms which have raised the speed and quality of government accounting in many countries.
For donors who are not disbursing through Channel 1, good practice requires the timely provision of expenditure information in formats that match the government’s accounting classifications.
The key good practices for governments seeking to bring external funds on accounting are:
to ensure that the accounting system provides a proper framework for reporting transactions, including •those involving external funds
capacity development to ensure the quality and timeliness of government accounts at all levels. •
There are now internationally recognised accounting standards for aid which provide attainable standards for countries using cash-based accounting systems. (See Box 30.)
47
Box 29: Tanzania: Aid-in-kind and dummy entries
The following extracts from Tanzania guidelines illustrate an approach to recording aid-in-kind in the mainstream government accounting system:
direct disbursements to the Project and ‘in kind’ AssistanceThere are instances where donors disburse directly to the project or provide in-kind assistance (donor procurement, recruitment of consultants, project staff engaged at the donor office). […] The government is encouraging all donors to route their resources through the Exchequer. However, where direct disbursements occur, the procedure used to account for such transactions is as follows:i. The donor disburses funds or provides goods or services directly to the project.ii. The project then reports on the amounts, goods, services or equipment value received to the
relevant MDA with relevant supporting evidence of the disbursements and purchases.iii. The MDA raises a TFN 358 application form in the normal manner. However, since expenditure has
in this case been already made, copies of payment vouchers and other supporting documents will be attached to the TFN 358 […].
iv. The Budget Division, upon satisfying itself, will grant a release warrant to the Accountant General.v. The Accountant General will now issue a Dummy Exchequer Issue Notification (for record purposes
only since there is no actual cash outflow) and such issuance will be recorded in the Integrated Financial Management System. At this stage, the expenditure will now have been incorporated in the government accounts.
Source: Government of Tanzania, 2005.
Box 30: Good practice: Follow international accounting standards for aid (IPSAS)
In 2003 the internationally recognised accounting standard, the International Public Sector Accounting Standard (IPSAS) for cash-based financial reporting was issued by the International Public Sector Accounting Standards Board (IPSASB) of the International Federation of Accountants. This is relevant to developing countries as the vast majority of their governments do not have accrual-based systems.
When this was adopted there were calls for an internationally accepted accounting standard for reporting external assistance. Since 2005, the IPSASB has undertaken a consultation process on exposure drafts to amend the cash-basis IPSAS to include additional required and encouraged disclosures for recipients of external assistance, issuing Exposure Draft (ED) 24 in 2005 and ED 32 in 2006. ED 32 proposes a split into mandatory and recommended disclosures under the cash basis of accounting.
The mandatory requirements include: “show separately total external assistance received in cash, show [aid-in-kind] i.e. payments made on government’s behalf by a third party, break down by different providers, show grants and loans separately” and others.
Following field tests in 2007, the IPSASB approved IPSAS, Financial Reporting Under the Cash Basis of Accounting – Disclosure Requirements for Recipients of External Assistance. The IPSAS (to be housed within the cash-basis IPSAS), is planned to be issued early in 2008. Full details are available from the IPSASB website: http://www.ifac.org/PublicSector/ProjectHistory.php?ProjID=0067
Source: http://www.ifac.org/PublicSector/Board.
6.7 Putting aid on auditThe relevance of putting aid on audit
Aid is on audit when “external financing is audited by government’s auditing system” (definition in Box 3). As with other dimensions, the country case studies looked beyond the formal question of whether aid is included within the scope of government audits and also considered the quality of auditing that takes place. Of course donors (and especially budget support donors) have an interest in the auditing of domestically financed expenditures as well. The rationale for using national audit systems is to conform to national lines of accountability and to support the strengthening of national systems.
Factors influencing aid on audit
Whether aid falls within the scope of the national audit process depends on the national legal framework, on
48
the availability, in practice, of the records on which audits focus, and on the capacity of the audit bodies and the auditing process.
The national legal requirements for the audit of externally financed expenditures often exceed what actually occurs. For example, public finance legislation in Tanzania explicitly requires that all aid, whether paid in cash or in the form of goods and services, should be subject to audit by the controller and auditor-general. Kenyan legislation similarly requires that all donor-funded expenditure should be audited by the auditor-general. In practice, the scope of government audit is commonly limited to funds that are on budget and on accounting, so may implicitly exclude most aid (though budget support is inherently covered). In a number of cases auditors themselves have highlighted this as an anomaly. Thus:
The Burkina Faso study noted that non-GBS ODA expenditures are not included in government •accounts, and are therefore not regularly checked by the court of auditors. The court is aware of this. Recommendations made while checking the application of the financial management law of 2003 and repeated since are that “external funding processed by [the directorate for cooperation] should be sent to the treasury for accounting”. This is scheduled for 2008.
In Uganda, in the 2004/05 report the auditor-general states: •A scrutiny of the approved budget revealed that various projects received funding from un-appropriated sources/donors. From a sample tested $23,972,298.87 was disbursed to projects outside the approved budget. It appears therefore that information provided to parliament for approval is inadequate with regard to donor funded projects. (Putting Aid On Budget: Uganda Case Study)
Even where national audit offices do work that is of a high standard, they tend to be thinly resourced and limited in the volume of work they can manage; hence the proportion of funds/agencies subject to audit in a given year may be worryingly low. Auditors are further constrained by delays in the submission of accounts. As a consequence, their reports may be submitted late, and this contributes to a generally poor record of parliamentary follow-up of audit findings. (See also the discussion of parliamentary capacity in relation to putting aid on parliament on page 42.)
Against this background, audit arrangements are often the subject of special conditions attached to specific aid instruments. These may require an external audit, or an undertaking by the auditor-general to conduct (or to commission) a special audit.
Good practice considerationsSupport to audit capacity Support to the capacity of audit bodies is likely to be a long-term requirement, and could complement long-term action to strengthen parliamentary public accounts committees, as well as civil society demand for accountability. Direct support can be complemented by efforts to strengthen the demand for audit services. For example, the Uganda case study notes that, in the context of the move towards budget support, the auditor-general has benefited from significant institutional support in recent years, and also the timely preparation of audits has been a key focus of conditionality. This has helped underpin improvements in quality and timeliness of audit reports. Similar developments are taking place in Mozambique where, with the growth of budget support to the government, several donors have started supporting the administrative tribunal (the body responsible for external auditing) with various capacity building initiatives which have improved the scope and quality of its audit exercises on the budget, including its externally funded components.
Where special audits are to be commissioned, donors can avoid sidelining national audit authorities by involving them in the commissioning of audits. (In several countries, such involvement is required by law.) There is also scope for harmonisation of donor audit requirements (as happens through various forms of pooled funding).
Complementary approaches to fiduciary and other assurances sought by donorsFormal ex post audit is only one element of fiduciary assurance. Donors often seek much more immediate comfort about the use of their funds, and are concerned with wider questions about how well resources are used. Relevant complements include internal audit as an aspect of the control environment when funds are put on treasury, ongoing financial reports and reviews to verify the application (and in some cases the additionality) of aid funds, and diagnostic studies such as public expenditure reviews and public expenditure tracking surveys. There is thus a need to think broadly about the range of “audit” that may be required both for accountability and for monitoring and improving aid effectiveness, and to implement in ways that engage governments and harmonise donor efforts.
49
6.8 Putting aid on reportRelevance of putting aid on report
Under “on report” this study considered whether external financing is included in ex post reports by government (definition in Box 3). This dimension acknowledges that the monitoring of public expenditures involves a much wider range of (financial and other) reports than those provided by the formal accounting system. It is possible that aid which is not incorporated in the formal public accounts may nevertheless appear in such reports.
The definition of relevant reports is rather open-ended, but the study focused on reports which are owned by governments, and which are consolidated. In addition to the formal accounting records that are produced, there is potential for other reports (for example, budget execution reports, consolidated sector reports) to provide valuable management information and to integrate externally funded activities in the process. For example, such reports may provide additional detail on funds that, in budgeting terms, are disbursed as block grants to agencies or tiers of government.
Factors influencing aid on report
A general lack of strong government monitoring and evaluation systems means that monitoring and evaluation of aid is also incomplete. In-year reporting is often of poor quality, and, in some cases, government reluctance to share information limits its wider value. There are systematic efforts, in some cases (for example, Rwanda and Kenya) to produce systematic reports on aid at the national level or by ministry, but problems in government capacity to obtain, assemble and use such information are shown by persistent discrepancies between information assembled from different sources (such as country-assembled data versus aid flows reported to the OECD DAC). At the same time, the DAC harmonisation agenda has stimulated more efforts to supply and use relevant information on aid. Sector approaches and sector working groups have stimulated the preparation of sector-specific and general reports.
The most ambitious efforts are so-called aid information management systems (AIMS) – see the discussion in the Putting Aid On Budget Literature Review: paras 41–43 – of which the most conspicuous in the study’s sample countries are the Aid Management Platform (AMP) in Ethiopia, and the ODAMOZ database in Mozambique. The latter is functioning as a useful adjunct to efforts to bring aid directly on budget – see Box 31 – but the AMP in Ethiopia, though more ambitious, is not yet operating as an effective shared resource between government and donors. In many cases (of which Rwanda and Tanzania were conspicuous examples) a multiplicity of expenditure, accounting and reporting databases have accumulated. However, weak linkages between these databases detract from the comprehensiveness and quality of reporting. This makes it difficult to link financial reports, to non-financial (activity, outputs, outcomes and impacts) reporting. Often the multiplicity of reports and databases reflects a fragmentation of responsibilities within government, and there is a need for rationalisation and streamlining of such reporting. There is a risk that reports, when available, do not reach domestic stakeholders, but are oriented primarily to the demands of external aid agencies.
Box 31: The ODAMOZ database in Mozambique
International development agencies have made efforts in Mozambique to systematically compile, report and publish information on the ODA projects they operate or support. This is now being reported on a regular basis in the ODAMOZ web-based database (see www.odamoz.org.mz), which includes detailed information for all projects funded by agents participating in this aid dissemination project and is an important source of ODA information for the government. Since 2006 this database is being managed by the Ministry of Planning and Development, whilst efforts are being made to encourage its use by both donors and government agencies (GOM/PAP, 2007c: Annex 4:14).
Whilst ODAMOZ has been an important step in improving information flows between donors and the government, worthy of being emulated by other countries, efforts to put aid on budget should not be seen as limited only to a process of greater information generation and sharing.
Source: Putting Aid On Budget: Mozambique Case Study.
Good practice considerations
Reporting has costs, and the quality of reports depends on reporters having a continuing incentive to provide timely and accurate information. Consolidated reports can add value by capturing more aid than is directly reflected
50
in the government’s financial management systems, especially if financial data can be linked to information on activities, outputs and outcomes. Incompatibility of reporting categories, timetables and formats is a continuing challenge. Donors need to adapt to government, classifications, currency, time periods and reporting timetables in order to increase the value of such reports. At the same time, demands for reporting should take account of capacity, and be linked to efforts to strengthen analytical capacity in particular.
6.9 decentralisation
The case study countries show a variety of approaches to decentralisation. Ethiopia has a fully fledged federal system, with lower tiers mainly funded by non-earmarked transfers from the federal government. Uganda also has a highly developed intergovernmental transfer system, but based on much more tightly controlled specific–purpose grants. Several other countries (for example, Rwanda and Mozambique) have decentralisation reforms under way which aim to increase responsibilities and public expenditures at local levels, and are simultaneously urging aid agencies to channel local-level financial support through the government financial system.
In all cases, decentralisation poses a special challenge to bringing aid on budget. On the one hand, aid that its channelled to decentralised bodies or focused on specific geographical areas is often hard to capture for planning and budgeting purposes. On the other, one of the deterrents to bringing aid on budget is a fear that disbursement via central government (Channel 1) will not reach local levels fairly or efficiently. Mozambique is a prime example of this dilemma. Government and aid partners have worked to bring more aid on budget, through pooled sector funds; these efforts have been linked to the strengthening of financial management sub-systems and the introduction of a computerised financial management system. There has been substantial progress, but the delivery of resources to local level through the government system is still problematic.
The Uganda case study included a special review of decentralisation in the context of putting aid on budget. Its conclusions are summarised in Box 32. Both the Ethiopian27 and Ugandan experiences show that, instead of creating parallel mechanisms through projects and basket funds, sector budget support can be channelled though a government’s intergovernmental transfer system in support of local service delivery. This helps build local government budgetary and service delivery systems, and donor fiduciary concerns can be addressed through tracking resources via the transfer system, and building local systems. However, both cases benefited from a strong political support for decentralisation, and a capable central finance ministry able to ensure the efficient disbursement of grants to sub-national levels.
Box 32: Decentralised services on budget in Uganda
The Uganda case study demonstrates first, that capacity can effectively be built while using the intergovernmental transfer system and local government systems, even from a very weak starting point. It also demonstrates that strong government initiative has helped bring aid both on budget and on financial management systems in local governments – this has meant that off-budget pooled funds became far less prevalent in Uganda than other similarly aid dependent countries (for example, Tanzania and Mozambique).
The following lessons can be drawn: • Firstistheimportanceofthegovernmentdevelopingaclearandtransparentfiscaltransfer
system, where grants to local governments are clearly identifiable in the budget. • Secondistheneedtoaddressdonorfiduciaryconcernsdirectlybydevelopinginitiativeswhich
ensure transfers are protected and can be tracked, while strengthening local government budgeting, accounting and reporting systems.
• Third,demandingcapacityfromlocalgovernmentshelpsbuildcapacity,andthiscanbedoneby enforcing existing systems, and providing explicit incentives through initiatives such as the Local Government Development Programme (LGDP) assessment.
Yet the study also shows that a fragmented government system can have a detrimental effect just like fragmented projects, and increase transaction costs. These costs need to be weighed up against the benefits of using earmarked sectoral transfers as a means of tracking resources to basic services.
Source: Putting Aid On Budget: Uganda Case Study.
27 See detail on page 42 on the Protection of Basic Services (PBS) project in Ethiopia.
51
7
cOncluSiOnS And RecOmmendAtiOnS
This chapter provides recommendations (for individual governments and donors, and for joint action). The recommendations focus on practical ways of strengthening aid effectiveness by bringing aid on budget.
7.1 introduction
Overall, this study has found undoubted progress, achieved and under way, in bringing aid on budget, but there is still a lot of scope to improve the alignment and integration of external and domestic resources. This requires attention to the quality of capture on all dimensions as well as the volume of aid that is recorded in budget documents. Chapter 6 has identified many good practices (and some bad ones), related to each of the “on budget” dimensions considered by this study. This final chapter draws on them to present a coherent set of recommendations for governments and for donors, and for governments and donors acting jointly. It first considers the overall approach, and the scope for progress with different aid modalities.
7.2 Overall approach to improving aid captureDrivers of progress in putting aid on budget
The country studies have highlighted the ways in which government and donors working together (especially if led by government) can make progress in strengthening effectiveness, including putting aid on budget. Often the dialogue around budget support and sector wide approaches (SWAps) is a valuable focus. (Hence, as noted, the most striking gains in bringing aid on budget have been through the use of budget support and by bringing pooled funds on budget.)
Improvements in PFM are a fundamental factor in bringing aid on budget. At a general level they can increase donor confidence in government systems, and support a virtuous circle in which the integration of aid in country systems strengthens those systems and encourages the further integration of aid. At a specific level, the nuts and bolts for integrating aid are the nuts and bolts of the PFM system.
There has been growing consensus on the characteristics of effective PFM, supported by many good practice guidelines and encapsulated in the Public Expenditure and Financial Accountability (PEFA) criteria. And – as all the country studies observe – the DAC harmonisation and alignment agenda, now embodied in the Paris Declaration, has had a galvanising effect on the efforts of donors and partner governments at country level.
At the same time, factors other than the quality of PFM are also important – as implied by the rather weak correlation between overall measures of PFM quality and the use of country budget systems. (See Box 15 and the associated discussion on page 23.) A recent study of donors’ approaches to risk in the use of country systems explores both the financial and non-financial factors involved, and is also highly relevant to the issues discussed in this report. (Mokoro Ltd and CIPFA, 2008)
Where are the potential gains?
Budget support is automatically on budget in most if not all dimensions. For this and for other reasons it makes
52
sense for partner countries to create favourable conditions for (general and sector) budget support, and to encourage donors to provide more of their aid in that form. There are risks attached to budget support, for recipients as well as donors, and it is in both sides’ interest to design forms of budget support that mitigate these risks.
However, budget support has limits, and project aid has been growing. The problems associated with poorly integrated project aid still loom large. The bigger challenge, therefore, is to bring project aid on budget. (See Box 14.) The Rwanda study noted the challenges as follows:
The shift to budget support clearly is and remains the most efficient and comprehensive way of “putting aid on budget”. Nevertheless, a lot can be done to improve the way projects are integrated into the national planning, budgeting, accountability, accounting and auditing processes. Somehow the international pressure that has led some donors to shift to budget support has so far led to less concrete results in terms of shifting projects on budget. This may be partly linked to the complexity of the issue. In that respect, making sure aid is reflected on the budget is only a first step in the process. It is an essential step, since transparency will allow better planning, allocation, and accountability.
Nevertheless it is essential for local donors and government to push the reflection further than mere data collection, in particular towards how to ensure project planning and pipeline is aligned to government priorities, how projects can be integrated into the single treasury account, and how project execution can pass through SMARTGOV [the financial management system] and abide by the Law on State Finance and Property. (Putting Aid On Budget: Rwanda Case Study)
Several countries have paid special attention to designing procedures for managing externally financed projects within government systems. (Tanzania is one example; Mali’s so-called sector budget support – described in Box 26 – could be considered another.) In other cases, the integration of projects has been surprisingly neglected. The Uganda study notes:
As well as appearing on the revenue budget, project aid may use government procurement systems, but no projects use the government’s accounting systems. However, to date, a mechanism has not been developed through which project expenditures funded by donor aid can use either the automated accounting systems at central government, or a manual one.
A key problem has been the fact that project-based budgetary systems have not changed over time to evolve with the MTEF and sectoral processes. This is largely because it was widely perceived that budget support would become the dominant modality, but project aid has remained and is now growing again. Another implication of this is the lack of thinking that has gone on with respect to how projects can use domestic budget execution systems. (Putting Aid On Budget: Uganda Case Study)
Priorities and sequencing
Different countries have different institutional settings, their governments have different preferences in how they relate to donors, and their PFM systems are at different stages of development and reform. Strategies for bringing aid on budget need to be tailored to these different circumstances. Relevant considerations include:
Aid on budget issues are generally more important in more aid-dependent countries.• 28
Sectors where substantial aid flows support public services merit more attention.•
Capacity everywhere (including aid agency capacity) is limited, and in any case technical solutions will •not work unless incentives are right.
Bringing aid on budget needs to be considered as part of an overall strategy for PFM, which will have a •strong influence on the sequence of reforms.
28 On the on the hand, less aid-dependent countries, such as South Africa, tend to have stronger PFM systems. Therefore, while putting aid on budget in these countries may be less critical, it should also be easier, and donors should take advantage of this.
53
The analytical framework of this study, with its distinction between different dimensions of the budget system, may be a useful template in considering the choice and design of aid instruments. Chapter 6 has commented on which modalities are more/less likely to be captured on different dimensions. Often, the good practices it identifies have implications for the design of specific aid instruments as well as the broad choice of modalities. It is worth considering which dimensions are more important in a particular case, and whether disadvantages of non-capture in some dimensions can be mitigated on other ones. (For example, if it is not practical to put aid on budget or on treasury in a particular case, is it at least on plan and on report? Or, as in the case of Tanzania, bringing aid on account may mitigate failure to capture it on treasury.) Also, given that different donors have different preferences and constraints concerning modality and instrument design, can they focus on sectors and interventions that play to their strengths?
The Mozambique report exemplifies this way of thinking:
It is necessary to reflect on the urgency and necessity of bringing all donor funds on budget at all levels, especially given the recentness of some of the key reforms undertaken in Mozambique in the sphere of public financial management, and the ongoing challenges that still need to be addressed. In some instances – for example, in the case of large infrastructure projects with a high development impact, such as the construction of roads, bridges or large water and energy infrastructures – the benefits of maintaining ODA off treasury (although not necessarily off plan or off budget) may outweigh the potential costs of processing these projects’ funds through national PFM systems. This may be the case, for instance, if existing PFM systems are weak and unreliable and, thus, incapable of managing the funding and accounting requirements associated with this type of project with the reliability and quality necessary to ensure their timely undertaking. (Putting Aid On Budget: Mozambique Case Study)
7.3 Recommendations for governments The PFM agenda
Bringing aid on budget should be seen as part of the overall PFM reform agenda. It should be addressed within the comprehensive, country-led programmes of PFM reform that are envisaged in the PEFA “strengthened approach” (Public Expenditure Working Group, 2005). It is important for governments to publicise their systems, highlight improvements, and make their procedural requirements known to donors.
The following PFM priorities are especially relevant to the integration of aid within country systems:
Strengthen transparent medium-term expenditure planning (to bring aid on plan).•
Clear planning and budgeting calendar to integrate aid management with the planning and budgeting •process.
Better budget classification and presentation. (The newly approved IPSAS standards – see Box 30 – are •relevant here.) Especially relevant improvements include:
accurate identification of recurrent and capital expenditures (but moving away from separate o
“recurrent” and “development” budgets)
“source” codes to allow tracking of donor funds if requiredo
Discipline in budget execution (which can be reinforced by many of the points which follow).•
Single treasury accounts (and strict control of government bank accounts).•
More timely reports and accounts.•
Transparent and efficient mechanisms for funding sub-national governments.•
More transparent budget documents:•
to highlight plan/budget strategic linkso
to show aid sources of budget finance, and aggregate uses of aido
to facilitate more meaningful parliamentary review.o
54
Strategies for aid management
Few countries have well articulated strategies for the management of aid (as opposed to its mobilisation). Often the responsibilities for aid management within government are poorly defined, and the lines of communication with donors are confused. Preferences for budget support may be stated, but there is little guidance on how donors should deliver other forms of aid in ways that use country systems. The interconnections between aid and PFM are not always recognised. Some countries have recognised the need for more explicit strategies: for example, Rwanda is in the process of developing an aid management manual to augment its 2006 aid policy; South Africa has such a manual; Ghana is preparing an aid strategy – see Box 33 – and so is Mozambique.
Explicit aid effectiveness strategies should:
Link aid management strategy to strengthened public finance management reform (effectiveness of all •resources). The link begins with the making of policy through the budget process, and continues through all the phases of the budget cycle.
As part of this: address unclear and overlapping responsibilities for the coordination of planning and •budgeting. (See the earlier comments on the integration of recurrent and development budgets and of the responsibilities for their preparation, and the need to avoid disconnecting aid management and budgeting responsibilities.)
Use donor information as an input into scenario planning for the poverty reduction strategy papers •(PRSPs)/medium-term expenditure plans, and use scenario planning as a basis for seeking longer-term aid commitment.
Go beyond indication of preference for budget support modality, to specify good donor practices and •standard government procedures for other modalities.
Allow for dialogue about the use and the additionality of aid funds so as to reduce donor demand •for earmarking and micro-management of donor funds. (See page 42 for the discussion on Ethiopia’s Protection of Basic Services project and the Uganda Poverty Action Fund, and also the discussion on wider on audit approaches on page 48.)
Box 33: Ghana: Preparing a comprehensive aid strategy
The Ghana government is in the process of developing a comprehensive aid policy, known as the external resources mobilisation and management policy. It is currently at the Concept Note stage.
The strategy is in response to government and donor concerns expressed following the results of the 2006 Survey on Monitoring the Paris Declaration. This revealed:
• alackofsharedunderstandingbetweenthegovernmentanddevelopmentpartnersonthedefinitions and concepts related to aid effectiveness
• inconsistencies in reporting on activities financed by donor grants and loans, leading tosubstantial differences in the data available to the government and the development partners
• fragmentationofdonorefforts(bothloansandgrants)duetothelackoftherelevantpolicyprovisions guiding allocation of foreign aid
• thelackofcomprehensivenessofaidincludedinthebudget • uncoordinatedprovisionoftechnicalassistance.
There is also the recognition of a need for stronger government leadership and guidance for improving aid effectiveness.
In response to these needs, the objectives of the aid policy are twofold: (a) to increase the effectiveness of external resources to Ghana; and (b) to provide a basis on which the additional aid required by Ghana to meet its short- to mid-term investment needs can be mobilised.
The aid policy will include criteria, guidelines, and procedures for: the mobilisation of aid resources, procedures for handling different types of aid, aid reporting, and guidelines for improving sector policy frameworks. It will also specify the roles and responsibilities for aid management at all levels of government.
Source: Putting Aid On Budget: Ghana Case Study.
55
7.4 Recommendations for donors
The Paris Declaration commitment to increase the use of programme-based approaches should provide continued impetus for the greater integration of aid with government systems. Aid agencies should review their aid instruments designs, centrally and at country level, from the perspective of aligning and integrating with country systems along each of the dimensions reviewed in this study.
Agencies should in particular review the extent to which they use regular government systems. This requires investing in understanding the systems and procedures of specific recipients, and training aid agency staff accordingly. It is important to make the use of government systems a design consideration in the early stages of designing any intervention, not an afterthought.
Agencies should improve the medium-term predictability of their aid (an area in which there has been relatively little progress) and provide aid flow information that is linked to the government fiscal year, the government planning and budget calendar, and government budget classifications.
7.5 Recommendations for joint action
Working together, governments and their aid partners should:
recognise the continuing importance of joint government–donor action (at general and sector level) to •address the greater use of government systems
incorporate “aid on budget” objectives within country-led public finance management strategies•
strengthen monitoring of the Paris Declaration commitments:•
set standardised and less flexible definitions at international levelo
monitor the proportion of project aid that is on budgeto
continue the country-level focus on use of government systems, and set country-level targets o
for progress
rationalise aid-related databases and collaborate in sharing information on all types of aid.•
Jointly setting and monitoring standards and targets for the use of country budget systems is especially important, because there are few effective sanctions on donor performance, apart from donor peer pressure and the voluntary obligations of mutual accountability between donors and partner governments.
7.6 next steps
Many steps can be taken in the short term, but a long-term perspective is required. As the Mozambique country study noted:
Given its structural and overarching nature, the process of putting aid on budget is likely to require time and a long-term perspective from all parties involved, and a gradualist approach that takes into account the many interdependencies that exist between the various policy spheres if these objectives of bringing aid on budget at all levels of the budget cycle are to be achieved. (Putting Aid On Budget: Mozambique Case Study)
Many of the “good aid on budget practices” identified in this report are general good practices. It has been emphasised that putting aid on budget is a subset of a wider agenda for aid and public finance management. Pursuing aid on budget should not become a separate, fragmented chore; it should be incorporated in wider programmes for PFM strengthening, and included among the tasks of existing general and sector forums.
56
BiBliOgRAPhy
This bibliography includes documents referenced in the text, plus a selection of other key documents. The Putting Aid On Budget Literature Review provides a fuller bibliography and the putting aid on budget country reports list country-specific documents used in the case study research.
Outputs for the putting aid on budget studyCarter, R. (2008). Putting Aid On Budget Literature Review. Pretoria: CABRI and SPA.
Mokoro Ltd (2007). Putting Aid On Budget Inception Report. Pretoria: CABRI and SPA.
Mokoro Ltd (2008). Good Practice Note: Using Country Budget Systems. Pretoria: CABRI and SPA.
country case studies
Group A (background working papers):Lanser, P. (2008). Putting Aid On Budget. Burkina Faso Case Study – Working Paper. Oxford: Mokoro Ltd.
Bladon, R. (2008). Putting Aid On Budget. Ethiopia Case Study – Working Paper. Oxford: Mokoro Ltd.
Fölscher, A. (2008). Putting Aid On Budget. Kenya Case Study – Working Paper. Oxford: Mokoro Ltd.
Fölscher, A. (2008). Putting Aid On Budget. South Africa Case Study – Working Paper. Oxford: Mokoro Ltd.
Fölscher, A. (2008). Putting Aid On Budget. Tanzania Case Study – Working Paper. Oxford: Mokoro Ltd.
Group B (free-standing reports):Betley, M. (2008). Putting Aid On Budget. Ghana Case Study. Oxford: Mokoro Ltd (2008).
Betley, M. (2008). Putting Aid On Budget. Mali Case Study. Oxford: Mokoro Ltd (2008).
Warren-Rodríguez, A. (2007). Putting Aid On Budget. Mozambique Case Study. Oxford: Mokoro Ltd.
Chiche, M. (2008). Putting Aid On Budget. Rwanda Case Study. Oxford: Mokoro Ltd.
Williamson, T. (2008). Putting Aid On Budget. Uganda Case Study. Oxford: Mokoro Ltd.
Other referencesBooth et al (2008). Good governance, aid modalities and poverty reduction. From better theory to better practice. Final
Synthesis Report. Synthesis by David Booth and Verena Fritz. Research project of the Advisory Board for Irish Aid. CDD-Ghana, Economic and Social Research Council, ODI and Chr. Michelsen Institute.
Dorotinsky, W. (2003). Implementing Financial Management Information System Projects: The World Bank Experience. Washington D.C.: World Bank.
Government of Tanzania (2005). A Practical Guide to Using the Government of Tanzania Exchequer System.
High Level Forum (2005). Paris Declaration on Aid Effectiveness. Revised edition with targets.
IDD & Associates (2006). Evaluation of General Budget Support: Synthesis Report. London: DFID.
International Monetary Fund (1998). Manual on Fiscal Transparency. Prepared by the Fiscal Affairs Department. Washington D.C.: IMF.
57
International Monetary Fund (2001). Manual on Fiscal Transparency. Prepared by the Fiscal Affairs Department. (March 23 2001). Washington D.C.: IMF.
International Monetary Fund (2007a). Code of Good Practices on Fiscal Transparency. Washington D.C.: IMF.
International Monetary Fund (2007b). Manual on Fiscal Transparency. Washington D.C.: IMF.
International Public Sector Accounting Board (2006). Financial Reporting Under the Cash Basis of Accounting – Disclosure Requirements for Recipients of External Assistance. Exposure Draft 32. Proposed Amendment to International Public Sector Accounting Standard – Financial Reporting Under the Cash Basis of Accounting. (November 2006). New York: International Federation of Accountants.
Lawson et al (2007). Joint Evaluation of Multi-Donor Budget Support to Ghana. Final Report. Volume One: Evaluation Results and Recommendations on Future Design & Management of Ghana MDBS. Report to the Government of Ghana and to the MDBS Partners by Andrew Lawson, Gyimah Boadi, Ato Ghartey, Adom Ghartey, Tony Killick, Zainab Kizilbash Agha and Tim Williamson. ODI and CDD-Ghana.
Millennium Challenge Corporation (2006). Fiscal Accountability Guidelines.
Mokoro and CIPFA (2008 forthcoming). Stocktake on Donor Approaches to Managing Risk When Using Country Systems.
ODI (2007). Interim Report on the Uganda Division of Labour Exercise. London: ODI.
OECD Development Assistance Committee (2005). Paris Declaration on Aid Effectiveness. High Level Forum, Paris, March 2. Paris: OECD.
OECD Development Assistance Committee (2006a). Monitoring the Paris Declaration on Aid Effectiveness. Documents 1–5. Explanatory Note, Donor Questionnaire, Government Questionnaire, Country Worksheet, Definitions & Guidance. Paris: OECD.
OECD Development Assistance Committee (2006b). DAC Guidelines and Reference Series: Harmonising Donor Practices for Effective Aid Delivery. Volume 2. Paris: OECD.
OECD (2007a). 2006 Survey on Monitoring the Paris Declaration. Overview of the Results. Pre-Publication Copy. Paris: OECD.
OECD (2007b). Definitions and Guidance for PFM Indicators. Joint Venture on Public Finance Management. Paris: OECD.
Parry, M. (2005). Why Government IFMS procurements so often get it wrong. ACCA International Public Sector Bulletin, Issue 4.
PEFA Secretariat (2006a). Public Financial Management Performance Measurement Framework. Washington D.C.: World Bank.
PEFA Secretariat (2006b). PFM Performance Measurement Framework. Report on Early Experience from Application of the Framework. PEFA.
Public Expenditure Working Group (2005). Supporting Better Country Public Financial Management Systems: Towards a Strengthened Approach to Supporting PFM Reforms. Current Consultative Draft.
De Renzio, P. and Sulemane, J.A. (2006). Integrating Reporting of PRS and Budget Implementation: The Mozambican Case. London: ODI.
Schick, A. (1998). A Contemporary Approach to Public Expenditure Management. Prepared for the Economic Development Institute of the World Bank Core Course “Budgetary Processes and the Analysis and Management of Public Expenditure”. International Bank for Reconstruction and Development/World Bank.
Sida (2007). Public Finance Management in Development Co-operation. A Handbook for Sida Staff. Stockholm: Sida.
Williamson, T. and Kizilbash Agha, Z. (2008). Building Blocks or Stumbling Blocks? The Effectiveness of New Approaches to Aid Delivery at the Sector Level. Working Paper 6. Tim Williamson and Zainab Kizilbash Agha, with Liv Bjornstad, Gerald Twijukye, Yamungu Mahwago and George Kabelwa. Research project of the Advisory Board for Irish Aid. CDD-Ghana, Economic and Social Research Council, ODI and Chr. Michelsen Institute.
World Bank (1998). Public Expenditure Management Handbook. PREM (Poverty Reduction and Economic Management) Network. Washington D.C.: The International Bank for Reconstruction and Development/World Bank.
58
Annexure
SummARy OF cOuntRy Study FindingS
60
61
cOntentS
Acronyms and abbreviations 62
Introduction 63
Group A country studies 65
Burkina Faso 65
Ethiopia 67
Kenya 72
South Africa 76
Tanzania 81
Group B country studies 85
Ghana 85
Mali 89
Mozambique 94
Rwanda 100
Uganda 105
Boxes
Box 1: Different dimensions of on budget/capturing aid 63
Box 2: Structure of country study reports 64
62
AcROnymS And ABBReViAtiOnS
AfDB African Development Bank
AFRODAD African Forum and Network on Debt and Development
DAC Development Assistance Committee (of the OECD)
DFID Department for International Development [UK]
Danida Danish International Development Agency
EC European Commission
GAVI Global Alliance for Vaccines and Immunisation
GBS general budget support
GFATM Global Fund for Aids, Tuberculosis and Malaria
HIPC Heavily Indebted Poor Countries Initiative
IDA International Development Association
IFMIS integrated financial management and investment system
IMF International Monetary Fund
IP international partner
IT information technology
MDA ministries, departments and agencies
MTEF medium-term expenditure framework
NGO non-governmental organisation
ODA official development assistance
OECD Organisation of Economic Co-operation and Development
OECD DAC OECD Development Assistance Committee
PEFA Public Expenditure and Financial Accountability
PFM public finance management
PIP public investment programme
PIU project implementation unit
PRSP poverty reduction strategy paper
STA single treasury account
SWAp sector wide approach
SWG sector working group
TA technical assistance
UK United Kingdom
UN United Nations
UNDP United Nations Development Programme
UNICEF United Nations Children’s Fund
US United States of America
USAID United States Agency for International Development
US$ United States Dollar
63
intROductiOn
The terms of reference for the Putting Aid On Budget study call for two types of country study: a set with a “lighter touch” and a more in-depth set. All the countries underwent “lighter touch” research: an initial factual scan of the country’s situation regarding putting aid on budget and reflecting it in the financial accounts. The research in the other countries went further to understand what drives the identified country practices and the effects of efforts to improve aid transparency. The countries that only underwent the “lighter touch” research are classified as Group A. Group B countries underwent more in-depth research.
group A group B
Burkina Faso Ghana
Ethiopia Mali
Kenya Mozambique
South Africa Rwanda
Tanzania Uganda
Both overall and among the Group B countries, there is a good range of geographical locations, administrative/linguistic inheritance, and degrees of aid management and PFM sophistication. The Group B countries include several where concerted efforts to bring aid on budget offer scope for learning. The Group A countries include a number of large aid recipients and some with federal systems.
“Putting aid on budget” means capturing aid at different phases of the budget cycle. Box 1 defines the framework of related terms as they are used in the Putting Aid On Budget study.
Box 1: Different dimensions of on budget/capturing aid
Term Definition
On plan Programme and project aid spending integrated into spending agencies’ strategic planning and supporting documentation for policy intentions behind the budget submissions.
On budget* External financing, including programme and project financing, and its intended use reported in the budget documentation.
On parliament (or through budget)
External financing included in the revenue and appropriations approved by parliament.
On treasury External financing disbursed into the main revenue funds of a government, and managed through the government’s systems.
On accounting External financing recorded and accounted for in a government’s accounting system, in line with the government’s classification system.
On audit External financing audited by a government’s auditing system.
On report External financing included in ex post reports by a government.
* It should be clear from the context if ‘on budget’ is meant in a more general sense than this precise definition.
64
The country studies report on each of these dimensions.
Box 2: Structure of country study reports
Section A: country context The aim in this section is to identify the key characteristics relevant to aid on budget.
A1 A note on key information sources.
A2 Country institutions: structure of the government and the main institutions for planning, budgeting and budget implementation.
A3 Aid context: overview of aid flows, their importance, the aid modalities/instruments in use, and the main donors active.
A4 Institutions for aid management: government, donor and joint donor/government structures and procedures for aid management.
A5 Legal and practical framework for budgeting: legal requirements and actual practice, including the budget calendar, budget structure and classification, accounting and reports, any special provisions for the incorporation of aid, key weaknesses and any recent or pending reforms.
Section B: evidence and assessment of capturing aidThis is the core of the study and the report. The matrix shows the structure of the study/reporting against each dimension of capturing aid. For Group B countries, there is additional work to assess the quality of capture, and especially to explore reasons for success or failure in capture.
DimensionEvidence of capture (What is captured?)
Quality of capture (How well? How useful?)
Explanations (Why/Why not?)
B1 On plan
B2 On budget
B3 On parliament
B4 On treasury
B5 On account
B6 On audit
B7 On report
Section c: Reflections • IdentifiablepatternsintheSectionBfindings,andpossibleexplanations. • Particularexamplesofgoodpractice. • Particularexamplesofbadpracticeorreformsthatdonotwork. • Possiblerecommendations(atcountrylevelorwider). • Issuestobefollowedupinfurtherwork(especiallywhereafullerGroupBreportistobe
prepared).
This annexure summarises Section B for each country from the country studies.
65
Gr
ou
p A
co
un
tr
y s
tu
die
s
Bu
rK
inA
FA
so
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(Ho
w w
ell?
Ho
w u
sefu
l?)
exp
lana
tio
ns(W
hy/W
hy n
ot?
)
On
pla
nPr
og
ram
me
and
pro
ject
aid
is in
clud
ed in
the
var
ious
pla
nnin
g
do
cum
ents
, the
mo
st s
igni
fican
t o
f whi
ch is
the
PIP
. Thi
s lo
ng-
term
pro
gra
mm
e is
up
dat
ed a
nnua
lly. T
he P
RSP
’s Pr
iorit
y A
ctio
ns
Pro
gra
mm
e in
clud
es t
he P
IP p
rog
ram
mes
and
pro
ject
s w
hich
hav
e p
riorit
y o
r st
rate
gic
imp
ort
ance
for
the
go
vern
men
t.
Mo
st c
om
mitm
ents
are
tak
en in
to a
cco
unt.
The
200
6 Su
rvey
on
Mo
nito
ring
the
Par
is D
ecla
ratio
n re
po
rts
that
dis
bur
sem
ents
ar
e re
cord
ed a
t a
rate
of 9
2% o
f OD
A o
ver
a th
ree-
year
per
iod
. Th
e D
irect
ora
te G
ener
al fo
r C
oo
per
atio
n (D
irect
ion
Gén
éral
e d
e la
Coo
pér
atio
n) re
po
rts
a sl
ight
ly lo
wer
rat
e o
f 85%
. Thi
s d
irect
ora
te p
asse
s in
form
atio
n to
the
Dire
cto
rate
Gen
eral
for
the
Co
ord
inat
ion
and
Ass
essm
ent
of I
nves
tmen
ts (D
irect
ion
Gén
éral
e d
e la
Coo
rdin
atio
n et
de
l’Eva
luat
ion
des
Inve
stis
sem
ents
), w
hich
is
resp
ons
ible
for
the
PIP.
All
the
dat
a re
late
to
OD
A b
etw
een
go
vern
men
ts. T
he O
DA
p
rovi
ded
dire
ctly
to
the
priv
ate
sect
or,
NG
Os
or
civi
l so
ciet
y is
no
t ta
ken
into
acc
oun
t.
GB
S is
cap
ture
d v
ia t
he F
inan
ce M
inis
try’
s Pe
rman
ent
Secr
etar
iat
for
the
Mo
nito
ring
of F
inan
cial
Po
licie
s an
d P
rog
ram
mes
, whi
ch is
als
o
resp
ons
ible
for
dra
win
g u
p t
he o
vera
ll M
TEF,
form
ing
par
t o
f the
p
lann
ing
.
The
Dire
cto
rate
Gen
eral
for
Co
op
erat
ion
is re
spo
nsib
le fo
r th
e no
n-G
BS
OD
A. I
t p
resi
des
ove
r th
e ne
go
tiatio
ns fo
r lo
an a
nd g
rant
ag
reem
ents
, whi
ch in
clud
e th
e g
ove
rnm
ent’s
ow
n co
ntrib
utio
ns. I
t ta
kes
par
t in
bila
tera
l del
iber
atio
n p
roce
sses
tha
t en
able
gra
nts
that
d
o n
ot
have
a g
ove
rnm
ent
cont
ribut
ion
to b
e m
oni
tore
d. O
nly
aid
b
y sm
all a
nd in
freq
uent
do
nors
can
the
refo
re s
lip t
hro
ugh
the
net,
or
aid
kep
t q
uiet
by
conn
ivan
ce b
etw
een
the
do
nor
and
the
ben
efici
ary
stru
ctur
e.
On
bud
get
All
kno
wn
aid
is re
gis
tere
d a
s re
ceip
ts a
nd e
xpen
ses
in t
he
finan
ce la
w fo
r th
e g
ove
rnm
ent
bud
get
. In
rece
ipts
, the
bud
get
d
oes
no
t d
istin
gui
sh b
etw
een
gra
nts
and
loan
s. In
exp
ense
s,
pro
gra
mm
e an
d p
roje
ct a
id is
list
ed in
div
idua
lly in
the
sec
tion
rela
ting
to
inve
stm
ents
. The
typ
e o
f exp
ense
s is
no
t sp
ecifi
ed, b
ut is
su
mm
aris
ed in
the
sin
gle
line
: “Pr
oje
ct e
xpen
ses”
.
GB
S is
mix
ed in
with
the
go
vern
men
t’s o
wn
reso
urce
s an
d c
anno
t b
e tr
acke
d in
thi
s w
ay, w
hich
is w
hy t
he s
umm
ary
bud
get
incl
udes
a
spec
ial t
able
list
ing
GB
S b
y IP
s.
Mo
st o
f the
GB
S is
list
ed in
the
bud
get
if it
is b
roug
ht t
o li
ght
in
time.
Und
erta
king
s ca
rrie
d o
ut la
ter
in t
he y
ear
can
be
incl
uded
b
y th
e Fi
nanc
e M
inis
ter,
who
is a
utho
rised
to
neg
otia
te w
ith a
ny
inte
rnal
or
exte
rnal
so
urce
of f
und
ing
, in
ord
er t
o b
alan
ce t
he
go
vern
men
t b
udg
et.
In a
dd
itio
n to
the
am
oun
ts k
now
n to
be
par
t o
f the
pla
n, b
etw
een
80%
and
90%
are
est
imat
ed t
o b
e lis
ted
in t
he b
udg
et.
Fore
ign
aid
is d
etai
led
by
sour
ce o
f fun
din
g, b
y p
roje
ct a
nd b
y b
enefi
ciar
y m
inis
try.
Sim
ply
des
igna
ted
as
“Pro
ject
exp
ense
s”, t
here
is
no
diff
eren
tiatio
n b
etw
een
per
sonn
el, o
per
atio
nal a
nd in
vest
men
t co
sts.
The
sam
e is
tru
e o
f the
go
vern
men
t’s o
wn
cont
ribut
ion,
whi
ch
is e
nter
ed in
to t
he s
ame
sect
ion
of t
he b
udg
et.
GB
S is
par
t o
f the
bud
get
by
defi
nitio
n.
Loan
s ar
e re
sour
ces
take
n o
ut b
y th
e g
ove
rnm
ent
and
mus
t b
e en
tere
d in
to t
he b
udg
et.
Gra
nts
for
inve
stm
ent
pro
ject
s re
qui
ring
a p
art
cont
ribut
ion
fro
m
the
go
vern
men
t ar
e en
tere
d in
the
bud
get
to
get
her
with
the
g
ove
rnm
ent
cont
ribut
ion
– a
pre
req
uisi
te fo
r th
e g
ove
rnm
ent
to
mak
e th
e m
one
y av
aila
ble
.
TA a
nd g
rant
s th
at re
qui
re n
o c
ont
ribut
ion
fro
m t
he g
ove
rnm
ent
may
be
excl
uded
fro
m t
he b
udg
et, e
spec
ially
if t
hey
are
man
aged
d
irect
ly b
y th
e d
ono
r w
itho
ut b
eing
mo
ved
via
a n
atio
nal a
utho
risin
g
offi
cer.
On
p
arlia
men
tA
s th
e an
nual
bud
get
is v
ote
d o
n b
y th
e N
atio
nal A
ssem
bly
, OD
A is
im
plic
itly
incl
uded
.
The
Nat
iona
l Ass
emb
ly is
no
t g
ener
ally
invo
lved
in d
raw
ing
up
ag
reem
ents
: tha
t is
the
go
vern
men
t’s re
spo
nsib
ility
. But
the
Nat
iona
l A
ssem
bly
is k
ept
abre
ast
of s
uch
agre
emen
ts in
sofa
r as
it r
atifi
es
them
.
Som
e ag
reem
ents
hav
e b
een
criti
cise
d fo
r th
eir
hig
h o
per
atin
g
cost
s. T
he N
atio
nal A
ssem
bly
em
pha
sise
s th
e p
rinci
ple
tha
t O
DA
sh
oul
d b
e us
ed fo
r in
vest
men
ts, e
spec
ially
reim
bur
sab
le lo
ans.
The
B
urki
na F
aso
co
untr
y re
po
rt fo
r th
e jo
int
eval
uatio
n o
f GB
S1 no
tes
case
s w
here
the
Nat
iona
l Ass
emb
ly re
que
sts
amen
dm
ent
in fa
vour
o
f dep
artm
ents
rece
ivin
g li
ttle
or
no O
DA
bef
ore
the
y w
ill r
atify
an
agre
emen
t.
Fro
m t
he p
ersp
ectiv
e o
f the
do
nors
, OD
A is
mai
nly
used
to
fund
ac
tiviti
es li
sted
in t
he P
RSP
Prio
rity
Act
ions
Pro
gra
mm
e. T
he P
RSP
is
kno
wn
to p
arlia
men
t, b
ut it
s m
ain
fram
e o
f ref
eren
ce is
the
gen
eral
in
tere
st. G
iven
tha
t so
cial
initi
ativ
es, i
nclu
din
g b
asic
ed
ucat
ion
and
hea
lth, a
lread
y re
ceiv
e si
gni
fican
t am
oun
ts o
f fo
reig
n ai
d, t
he
Nat
iona
l Ass
emb
ly t
end
s to
lean
to
war
ds
oth
er a
reas
, see
king
inte
r-se
cto
ral b
alan
ce.
66
Bu
rK
inA
FA
so
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(Ho
w w
ell?
Ho
w u
sefu
l?)
exp
lana
tio
ns(W
hy/W
hy n
ot?
)
On
trea
sury
All
GB
S p
asse
s th
roug
h tr
easu
ry b
y d
efini
tion.
Pro
gra
mm
e ai
d g
iven
in
the
form
of p
oo
led
fund
s is
eith
er p
aid
to
a t
reas
ury
acco
unt
or
to
the
Cen
tral
Ban
k o
f Wes
t A
fric
an S
tate
s.
Oth
er p
rog
ram
me
and
pro
ject
aid
do
es n
ot
gen
eral
ly p
ass
thro
ugh
trea
sury
, but
is p
aid
to
acc
oun
ts in
the
Cen
tral
Ban
k o
f Wes
t A
fric
an
Stat
es o
r to
loca
l co
mm
erci
al b
anks
for
the
par
t im
ple
men
ted
in
Bur
kina
Fas
o.
Cap
ture
d b
y d
efini
tion.
No
t ap
plic
able
.
On
acco
unt
GB
S is
ent
ered
in fu
ll al
ong
sid
e th
e g
ove
rnm
ent’s
ow
n re
sour
ces
and
is m
anag
ed u
sing
the
ap
plic
able
nat
iona
l pro
ced
ures
. It
is t
hus
incl
uded
in g
ove
rnm
ent
acco
untin
g a
nd c
ons
eque
ntly
in t
he B
udg
et
Rev
iew
Act
.
No
n-G
BS
OD
A is
no
t ta
ken
into
acc
oun
t in
go
vern
men
t ac
coun
ting
. W
here
ver
po
ssib
le, e
xpen
ses
are
kep
t in
sha
do
w a
cco
unts
and
in
clud
ed a
s su
ch in
the
Bud
get
Rev
iew
Act
.
No
n-G
BS
OD
A d
isb
urse
men
ts a
re c
ont
rolle
d a
nd fl
agg
ed if
the
y p
ass
via
a na
tiona
l aut
horis
ing
offi
cer.
For
oth
er d
isb
urse
men
ts, t
he
Dire
cto
rate
Gen
eral
for
Co
op
erat
ion
relie
s o
n in
form
atio
n p
rovi
ded
b
y th
e d
ono
r o
r p
erfo
rman
ce s
truc
ture
. Inf
orm
atio
n is
oft
en p
atch
y,
and
pro
vid
ed la
te. J
oin
t ef
fort
s to
imp
rove
thi
s ha
ve b
een
agre
ed
on.
Som
e O
DA
is m
anag
ed o
utsi
de
natio
nal s
yste
ms.
Info
rmat
ion
dep
end
s o
n w
heth
er t
he d
ono
r is
fort
hco
min
g. F
urth
erm
ore
, ac
coun
ting
ter
ms
are
not
alw
ays
com
pat
ible
. The
re a
re s
ever
al
dat
abas
es t
hat
use
diff
eren
t cl
assi
ficat
ion
syst
ems.
The
Dire
cto
rate
G
ener
al fo
r C
oo
per
atio
n is
in t
he p
roce
ss o
f cha
ngin
g it
s cl
assi
ficat
ion
syst
em, m
ovi
ng fr
om
the
UN
DP
syst
em t
o t
he O
EC
D
DA
C s
yste
m fo
r g
reat
er h
arm
oni
satio
n.
The
dat
a is
no
t ye
t fu
lly d
igiti
sed
. The
Dire
cto
rate
Gen
eral
for
Co
op
erat
ion
is in
the
pro
cess
of d
raw
ing
up
the
Inte
gra
ted
Sys
tem
fo
r E
xter
nal F
inan
ces
(Circ
uit I
ntég
ré d
es F
inan
ces
Exté
rieur
s), w
hich
w
ill h
ave
an in
terf
ace
enab
ling
co
mm
unic
atio
n w
ith t
he In
teg
rate
d
Exp
end
iture
Sys
tem
(Circ
uit I
nfor
mat
isé
de
la D
épen
se) a
nd t
he
inte
gra
tion
of d
ata
in t
he n
atio
nal a
cco
unts
.
On
aud
itTh
e na
tiona
l acc
oun
ts a
nd t
he B
udg
et R
evie
w A
ct a
re c
heck
ed
ever
y ye
ar b
y th
e C
our
t o
f Aud
itors
bef
ore
sub
mis
sio
n to
the
N
atio
nal A
ssem
bly
. GB
S is
par
t o
f the
nat
iona
l acc
oun
ts b
y d
efini
tion.
The
non-
GB
S O
DA
do
es n
ot
com
e w
ith a
n o
pin
ion
fro
m t
he C
our
t o
f Aud
itors
and
is u
sual
ly c
heck
ed b
y ex
tern
al a
uditi
ng fi
rms
in
acco
rdan
ce w
ith t
he c
ond
itio
ns s
et o
ut in
the
ag
reem
ents
.
No
n-G
BS
OD
A e
xpen
ses
are
not
incl
uded
in t
he n
atio
nal a
cco
unts
, and
are
the
refo
re n
ot
reg
ular
ly c
heck
ed b
y th
e C
our
t o
f Aud
itors
. The
co
urt
is a
war
e o
f thi
s. It
mad
e re
com
men
dat
ions
whi
le c
heck
ing
the
ap
plic
atio
n o
f the
fina
ncia
l man
agem
ent
law
of 2
003,
and
has
rep
eate
d
sinc
e th
at e
xter
nal f
und
ing
pro
cess
ed b
y th
e D
irect
ora
te G
ener
al fo
r C
oo
per
atio
n sh
oul
d b
e se
nt t
o t
he t
reas
ury
for
acco
untin
g. T
his
is
sche
dul
ed fo
r 20
08.
On
rep
ort
All
min
istr
ies
and
the
ir d
epar
tmen
ts a
re re
qui
red
to
pro
duc
e an
nual
ac
tivity
rep
ort
s, in
clud
ing
the
ir fu
ndin
g.
Ther
e ar
e an
nual
PR
SP a
nd s
ecto
r p
olic
y re
view
s at
co
nsul
tatio
n le
vel b
etw
een
the
go
vern
men
t an
d IP
s.
The
mo
st im
po
rtan
t re
po
rt fo
r al
l OD
A is
Dev
elo
pm
ent
Co
op
erat
ion,
join
tly p
ublis
hed
by
the
Dire
cto
rate
Gen
eral
for
Co
op
erat
ion
and
the
UN
DP.
At
inte
rnat
iona
l lev
el, t
here
are
OE
CD
DA
C s
tatis
tics
and
rece
ntly
th
e 20
06 S
urve
y o
n M
oni
torin
g t
he P
aris
Dec
lara
tion.
The
Dire
cto
rate
Gen
eral
for
Co
op
erat
ion/
UN
DP
rep
ort
has
oft
en
bee
n p
ublis
hed
late
, with
mis
sing
dat
a. E
ffort
s fo
r im
pro
vem
ents
ha
ve b
een
agre
ed o
n. T
he 2
005
rep
ort
ap
pea
red
in J
uly
2007
, a
sig
nific
ant
imp
rove
men
t.
The
vario
us re
po
rts
diff
er o
n va
rious
asp
ects
, and
are
diffi
cult
to
com
par
e. T
he D
irect
ora
te G
ener
al fo
r C
oo
per
atio
n/U
ND
P re
po
rt
is b
ased
on
dis
bur
sem
ents
. The
OE
CD
DA
C re
po
rt is
bas
ed o
n co
mm
itmen
ts. T
he c
lass
ifica
tion
syst
ems
also
diff
er. T
he D
irect
ora
te
Gen
eral
for
Co
op
erat
ion
/UN
DP
rep
ort
is m
ore
co
mp
rehe
nsiv
e o
n in
tern
atio
nal fi
nanc
ial i
nstit
utio
ns d
ata,
whi
le t
he O
EC
D D
AC
rep
ort
ha
s m
ore
info
rmat
ion
abo
ut b
ilate
ral s
upp
ort
.
The
harm
oni
satio
n an
d a
lignm
ent
pro
cess
es s
tart
ed b
y th
e R
om
e D
ecla
ratio
n an
d fo
llow
ed b
y th
e Pa
ris D
ecla
ratio
n ha
ve c
ont
ribut
ed
to b
ols
terin
g p
artn
er c
oun
try
ow
ners
hip
and
dev
elo
pm
ent
par
tner
will
ing
ness
to
use
nat
iona
l pro
ced
ures
and
pro
vid
e tim
ely
dis
bur
sem
ent
dat
a.
67
etH
iop
iA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(Ho
w w
ell?
Ho
w u
sefu
l?)
exp
lana
tio
ns(W
hy/W
hy n
ot?
)
On
pla
nE
thio
pia
’s se
cond
-gen
erat
ion
PRSP
, the
Pla
n fo
r A
ccel
erat
ed a
nd
Sust
aine
d D
evel
op
men
t fo
r th
e E
limin
atio
n o
f Po
vert
y, h
as n
ow
b
een
cost
ed fo
r al
l sec
tors
. It
has
bee
n al
igne
d w
ith E
thio
pia
’s M
acro
eco
nom
ic a
nd F
isca
l Fra
mew
ork
– a
thr
ee-y
ear,
rolli
ng fe
der
al
fram
ewo
rk –
for
its li
fesp
an (2
005/
06–2
009/
10).
It is
fina
lised
in d
raft
b
efo
re J
anua
ry o
f eac
h ye
ar, a
nd is
use
d a
s th
e b
asis
of t
he a
nnua
l b
udg
et. T
he b
udg
et is
tab
led
by
the
fed
eral
go
vern
men
t to
the
H
ous
e o
f Peo
ple
s’ R
epre
sent
ativ
es in
May
. The
co
stin
g fo
r th
e Pl
an
for
Acc
eler
ated
and
Sus
tain
ed D
evel
op
men
t fo
r th
e E
limin
atio
n o
f Po
vert
y in
clud
es re
sour
ces
that
hav
e to
be
rais
ed e
xter
nally
.
As
a se
cond
sta
ge
of t
he p
lann
ing
cyc
le, a
PIP
is a
lso
pre
par
ed,
cove
ring
cap
ital e
xpen
ditu
re o
ver
a th
ree-
year
per
iod
.
In g
ener
al, a
t fe
der
al le
vel,
pla
nnin
g p
roce
sses
tak
e g
oo
d a
cco
unt
of e
xter
nal s
upp
ort
, and
the
re a
re jo
int
sect
ora
l rev
iew
s ac
ross
so
me
sect
ors
(e.g
. ed
ucat
ion)
. But
the
re is
no
MTE
F lin
king
pla
nnin
g
pro
cess
es t
o e
xpen
ditu
re fo
reca
sts
and
bud
get
ing
sys
tem
s,
and
the
Mac
roec
ono
mic
and
Fis
cal F
ram
ewo
rk o
nly
go
es s
om
e w
ay t
o p
rovi
din
g t
he n
eces
sary
link
ages
. Giv
en t
he d
isco
nnec
ts,
item
s in
clud
ed a
t th
e p
lann
ing
pha
se m
ight
mo
re e
asily
no
t b
e ca
ptu
red
in b
udg
ets,
or
vice
ver
sa. I
n g
ener
al, t
he li
nks
bet
wee
n st
rate
gic
pla
nnin
g a
nd b
udg
et a
lloca
tion
are
po
or,
a fa
ct t
hat
was
co
mm
ente
d o
n in
Eth
iop
ia’s
PEFA
ass
essm
ent.
The
lack
of t
imel
ines
s o
f inf
orm
atio
n flo
ws
is p
rob
lem
atic
, lim
iting
the
ext
ent
to w
hich
do
nors
are
ab
le t
o m
ake
pre
dic
tab
le
com
mitm
ents
to
aid
flo
ws
so t
hat
thes
e in
tur
n ca
n b
e us
ed
mea
ning
fully
in g
ove
rnm
ent
pla
nnin
g. T
ang
ible
co
mm
itmen
ts a
re
usua
lly o
nly
mad
e o
n a
year
-on-
year
bas
is fo
r m
ost
of t
he b
ilate
ral
do
nors
(the
UK
bei
ng a
no
tab
le e
xcep
tion)
. Thi
s is
one
of t
he
driv
ers
beh
ind
the
go
vern
men
t o
f Eth
iop
ia’s
mo
ve t
o c
ost
the
Pla
n fo
r A
ccel
erat
ed a
nd S
usta
ined
Dev
elo
pm
ent
for
the
Elim
inat
ion
of P
ove
rty
usin
g s
cena
rios:
diff
eren
t le
vels
of p
lan
inte
rven
tion
are
pre
dic
ated
on
vary
ing
deg
rees
of d
ono
r co
mm
itmen
t an
d
sup
po
rt. T
he M
acro
eco
nom
ic a
nd F
isca
l Fra
mew
ork
is o
ften
bas
ed
on
estim
ates
of d
ono
r su
pp
ort
rat
her
than
firm
fig
ures
. Und
er t
he
curr
ent
arra
ngem
ents
, do
nors
do
no
t us
ually
sup
ply
ind
icat
ive
figur
es u
ntil
Janu
ary
(i.e.
six
mo
nths
bef
ore
the
sta
rt o
f the
fina
ncia
l ye
ar),
and
man
y d
ono
rs s
upp
ly t
his
info
rmat
ion
even
late
r. Th
e g
ove
rnm
ent
has
atte
mp
ted
to
cha
nge
this
, am
ong
oth
er t
hing
s,
by
seek
ing
ag
reem
ent
at t
he D
evel
op
men
t A
ssis
tanc
e G
roup
tha
t in
form
atio
n sh
oul
d b
e su
pp
lied
in S
epte
mb
er fo
r th
e fin
anci
al
year
sta
rtin
g in
the
follo
win
g J
uly,
and
by
rais
ing
the
issu
e at
the
H
igh
Leve
l Fo
rum
(the
foru
m w
here
hig
h-le
vel d
ialo
gue
tak
es
pla
ce b
etw
een
Dev
elo
pm
ent
Ass
ista
nce
Gro
up m
emb
ers
and
th
e g
ove
rnm
ent)
. Ho
wev
er, t
he D
evel
op
men
t A
ssis
tanc
e G
roup
b
elie
ves
that
Sep
tem
ber
is t
oo
ear
ly in
the
bud
get
cal
end
ars
of
man
y d
ono
rs fo
r th
em t
o g
ive
firm
co
mm
itmen
ts t
hen.
Giv
en t
he
com
ple
xity
and
unc
erta
intie
s o
f Eth
iop
ia’s
pla
nnin
g a
nd b
udg
etin
g
syst
ems,
whe
re s
ub-n
atio
nal b
udg
ets
are
dep
end
ent
on
a fe
der
al
sub
sid
y, t
here
is t
he p
ote
ntia
l fo
r se
rious
kno
ck-o
n ef
fect
s. E
vid
ence
su
gg
ests
, ho
wev
er, t
hat
in p
ract
ice
the
fed
eral
co
mm
itmen
ts u
sed
fo
r su
b-n
atio
nal b
udg
et c
eilin
gs
are
pre
dic
tab
le, e
ven
whe
n b
udg
et
sup
po
rt h
as b
een
susp
end
ed.
The
reco
ncili
atio
n b
etw
een
cost
ed p
lans
for
the
Plan
for
Acc
eler
ated
and
Sus
tain
ed D
evel
op
men
t fo
r th
e E
limin
atio
n o
f Po
vert
y an
d d
iscl
ose
d re
ceip
ts o
f aid
is a
lso
pro
ble
mat
ic. T
here
ha
ve b
een
very
larg
e d
iscr
epan
cies
bet
wee
n w
hat
the
go
vern
men
t ha
s d
iscl
ose
d a
s th
e fu
nds
need
ed t
o o
per
atio
nalis
e th
e p
lan,
and
w
hat
do
nors
say
has
bee
n d
isb
urse
d. D
ono
r re
cord
s ap
par
ently
sh
ow
mo
re fu
nds
dis
bur
sed
tha
n ar
e d
iscl
ose
d in
go
vern
men
t re
cord
s.
Som
e o
f the
relu
ctan
ce a
roun
d lo
ng-t
erm
co
mm
itmen
ts fr
om
do
nor
par
tner
s is
pro
bab
ly d
ue t
o t
he p
erce
ived
vo
latil
ity o
f the
do
mes
tic
po
litic
al s
ituat
ion
in E
thio
pia
. Dire
ct b
udg
et s
upp
ort
arr
ang
emen
ts
wer
e se
vere
ly d
isru
pte
d a
nd t
hen
rep
acka
ged
aft
er t
he 2
005
elec
tions
, and
do
nors
do
no
t se
em w
illin
g t
o c
om
mit
bey
ond
re
lativ
ely
near
tim
e ho
rizo
ns. I
t ha
s al
so re
sulte
d in
the
des
ign
of a
id
mo
dal
ities
whi
ch, w
hile
the
y us
e C
hann
el 1
, are
rela
tivel
y in
flexi
ble
, ha
ving
pro
-po
or
serv
ice
del
iver
y ex
pen
ditu
re a
t th
e d
ecen
tral
ised
le
vel a
s th
eir
targ
et fo
cus
At
sub
-nat
iona
l lev
el, t
he p
lann
ing
pro
cess
has
litt
le m
eani
ng
bec
ause
of r
eso
urce
env
elo
pes
. In
real
ity, d
iscr
etio
n is
ver
y lim
ited
, an
d s
ub-n
atio
nal e
ntiti
es h
ave
to t
ake
on
unfu
nded
man
dat
es,
par
ticul
arly
at
wer
eda
(dis
tric
t ad
min
istr
atio
n) le
vel.
68
etH
iop
iA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(Ho
w w
ell?
Ho
w u
sefu
l?)
exp
lana
tio
ns(W
hy/W
hy n
ot?
)
On
bud
get
The
exte
nt t
o w
hich
aid
can
be
said
to
be
on
bud
get
var
ies
acro
ss
aid
mo
dal
ities
. The
200
6 Su
rvey
on
Mo
nito
ring
the
Par
is D
ecla
ratio
n p
uts
aid
on
bud
get
at
74%
. Fun
din
g t
hro
ugh
the
larg
e st
rate
gic
in
stru
men
ts is
fully
on
bud
get
– t
he P
ublic
Sec
tor
Cap
acity
Bui
ldin
g
Pro
gra
mm
e, t
he P
rod
uctiv
e Sa
fety
Net
Pro
gra
mm
e, a
nd P
rote
ctin
g
Bas
ic S
ervi
ces.
Lo
ans
are
also
sho
wn.
Acc
ord
ing
to
Min
istr
y o
f Fi
nanc
e an
d E
cono
mic
Dev
elo
pm
ent
dat
a, m
ost
bila
tera
l sup
po
rt
is t
hro
ugh
Cha
nnel
1 o
r C
hann
el 2
, and
is c
aptu
red
. One
item
tha
t is
po
ssib
ly n
ot
cap
ture
d is
free
-sta
ndin
g T
A t
hat
is p
rovi
ded
no
t as
p
art
of d
ono
r p
roje
cts
(whi
ch d
o g
et c
aptu
red
) but
as
sep
arat
e, s
elf-
cont
aine
d in
terv
entio
ns.
An
ano
mal
y o
f the
fed
eral
sys
tem
, and
an
issu
e fr
om
tim
e to
tim
e,
is s
upp
ort
at
sub
-reg
iona
l lev
el. S
uch
sup
po
rt c
an b
e o
n b
udg
et
and
on
trea
sury
for
a re
gio
n, y
et re
mai
n o
ff b
udg
et o
r no
t re
po
rted
at
fed
eral
leve
l. So
me
do
nors
mig
ht u
se C
hann
el 1
at
sub
-reg
iona
l le
vel,
but
no
t re
po
rt s
uch
sup
po
rt a
t fe
der
al le
vel,
havi
ng s
et u
p
thei
r o
wn
rep
ort
ing
req
uire
men
ts. I
n th
eory
thi
s is
no
t p
erm
itted
(a
s ag
reem
ents
hav
e to
be
ratifi
ed b
y th
e fe
der
al g
ove
rnm
ent)
, but
it
hap
pen
s in
pra
ctic
e, a
nd is
exa
cerb
ated
by
the
mo
dus
op
eran
di
of p
artic
ular
do
nors
(e.g
. UN
ICE
F). N
GO
sup
po
rt fo
r su
b-n
atio
nal
entit
ies
has
bee
n su
bje
ct t
o t
he s
ame
pro
ble
ms.
In t
heo
ry a
gai
n, t
he
go
vern
men
t sh
oul
d k
now
all
abo
ut a
id fr
om
any
NG
O, s
ince
eac
h N
GO
has
to
get
reg
iona
l go
vern
men
t ap
pro
val f
or
its o
per
atio
ns.
But
in p
ract
ice
ther
e ha
ve b
een
inst
ance
s w
here
op
erat
ions
rem
ain
off
bud
get
.
Ther
e ar
e al
so s
om
e an
om
alie
s o
f fun
ds
bei
ng o
n b
udg
et a
t re
gio
nal l
evel
but
no
t at
fed
eral
leve
l (e.
g. t
he P
rod
uctiv
e Sa
fety
Net
Pr
og
ram
me)
, whi
ch h
as im
plic
atio
ns fo
r th
e ac
hiev
emen
t o
f PFM
o
bje
ctiv
es fo
r su
b-n
atio
nal e
xpen
ditu
re.
Bud
get
s at
sub
-nat
iona
l lev
el t
ake
into
acc
oun
t ai
d fl
ow
s (e
.g. I
rish
bud
get
sup
po
rt in
Tig
ray,
Sw
edis
h su
pp
ort
in A
mha
ra).
Unr
epo
rted
aid
at
sub
-nat
iona
l lev
el is
dec
linin
g. T
he M
inis
try
of F
inan
ce a
nd E
cono
mic
Dev
elo
pm
ent
rep
ort
s im
pro
vem
ents
in
do
nor
beh
avio
ur. I
ncre
asin
gly
do
nors
are
mo
ving
to
war
ds
Cha
nnel
1, m
eani
ng t
hat
mo
re a
nd m
ore
aid
flo
ws
are
cap
ture
d, a
nd
flow
s th
at a
re n
ot
cap
ture
d a
re t
he e
xcep
tion
rath
er t
han
the
rule
.
Wha
t is
cle
ar is
tha
t th
e g
ove
rnm
ent
is it
self
do
ing
muc
h to
cap
ture
as
muc
h ai
d o
n b
udg
et a
s p
oss
ible
, evi
den
ced
by
atte
mp
ts t
o c
lose
o
ut s
ub-n
atio
nal a
gre
emen
ts t
hat
byp
ass
the
Min
istr
y o
f Fin
ance
an
d E
cono
mic
Dev
elo
pm
ent,
and
to
get
do
nors
to
co
mm
it to
tim
ely
info
rmat
ion
on
aid
co
mm
itmen
ts a
nd d
isb
urse
men
ts. (
See
und
er
“On
pla
n” a
bo
ve.)
Info
rmat
ion
is c
red
ible
in t
hat
it is
gen
erat
ed in
par
tner
ship
–
info
rmat
ion
is p
asse
d o
n b
y d
ono
rs. I
ncre
asin
gly
, the
ho
pe
is t
hat
the
Aid
Man
agem
ent
Plat
form
can
be
used
as
a ve
hicl
e fo
r th
is
– w
ith d
ono
rs p
lug
gin
g in
tend
ed s
upp
ort
fig
ures
dire
ctly
into
the
sy
stem
.
Ther
e ar
e so
me
issu
es a
bo
ut t
he a
ccur
acy
of t
he d
ata
for
aid
-in-
kind
. It
is n
ot
alw
ays
po
ssib
le t
o p
red
ict
such
aid
pro
per
ly, e
spec
ially
if
it is
link
ed t
o u
nfo
rese
en e
vent
s su
ch a
s em
erg
enci
es (e
.g.
floo
din
g in
Dire
Daw
a in
200
6). T
he C
entr
al A
cco
unts
Dep
artm
ent
is
ultim
atel
y re
spo
nsib
le fo
r d
iscl
osi
ng a
ll ai
d-in
-kin
d t
hat
has
actu
ally
b
een
dis
bur
sed
, but
thi
s is
diffi
cult
to c
aptu
re a
t b
udg
et s
tag
e.
Mo
re t
imel
y in
form
atio
n fr
om
do
nors
wo
uld
imp
rove
the
co
mp
lete
ness
of b
udg
et in
form
atio
n.
A m
ove
to
war
ds
sing
le t
reas
ury
syst
ems
in s
ub-n
atio
nal j
uris
dic
tions
an
d a
n im
pro
vem
ent
in w
hat
the
go
vern
men
t ha
s te
rmed
“U
N
beh
avio
ur”
are
help
ing
to
red
uce
the
chan
ces
of u
ndis
clo
sed
aid
at
sub
-nat
iona
l lev
el. H
ow
ever
, the
re a
re in
cent
ives
for
sub
-nat
iona
l ju
risd
ictio
ns t
o le
ave
such
sup
po
rt o
ff b
udg
et, a
cco
unt
and
rep
ort
if
in d
oin
g s
o t
hey
avo
id a
pp
aren
tly b
eing
pen
alis
ed b
y o
ffset
s in
the
b
lock
gra
nt. H
ow
ever
, the
fed
eral
go
vern
men
t is
ad
aman
t th
at o
ffset
m
echa
nism
s ar
e ne
eded
, bo
th t
o p
rese
rve
equi
ty b
etw
een
reg
ions
an
d t
o c
reat
e d
isin
cent
ives
for
do
nors
to
up
set
the
fine
bal
ance
of
the
dec
entr
alis
ed fi
scal
fram
ewo
rk.
On
par
liam
ent
Info
rmat
ion
abo
ut a
id fl
ow
s is
incl
uded
in t
he b
udg
et in
form
atio
n p
ack
whi
ch is
pre
sent
ed t
o p
arlia
men
t as
par
t o
f the
ove
rall
bud
get
ap
pro
val p
roce
ss.
An
Ap
pro
pria
tion
Bill
is a
pp
rove
d, a
nd t
his
bec
om
es t
he c
urre
nt
year
’s A
pp
rop
riatio
n Pr
ocl
amat
ion.
By
imp
licat
ion,
ext
erna
l aid
flo
ws
are
note
d (o
ther
wis
e, g
iven
Eth
iop
ia’s
dep
end
ence
on
aid
, the
b
udg
etar
y ex
pen
ditu
re a
pp
rove
d c
oul
d n
ot
be
finan
ced
).
One
vie
w is
tha
t ai
d is
no
t ap
pro
ved
by
par
liam
ent
as s
uch:
p
arlia
men
t is
info
rmed
ab
out
aid
rath
er th
an a
ctiv
ely
cons
ider
ing
aid
ag
reem
ents
. The
Min
istr
y o
f Fin
ance
and
Eco
nom
ic D
evel
op
men
t co
ntes
ted
thi
s vi
ew a
t th
e tim
e o
f the
Put
ting
Aid
on
Bud
get
st
udy,
sta
ting
tha
t ai
d a
gre
emen
ts, i
nclu
sio
n in
the
bud
get
and
the
d
isb
urse
men
t o
f fun
ds
are
sub
mitt
ed fo
r p
arlia
men
tary
ap
pro
val.
Ther
e ar
e d
oub
ts a
bo
ut t
he e
xten
t to
whi
ch p
arlia
men
t en
gag
es
in t
he p
roce
ss o
f ap
pro
ving
aid
. Any
thin
g m
ore
tha
n a
rub
ber
st
amp
ing
func
tion
is u
ncer
tain
.
Parli
amen
taria
ns d
o n
ot
und
erst
and
the
ir ro
le o
r th
e in
stitu
tiona
l fr
amew
ork
in w
hich
the
y o
per
ate.
The
re h
ave
bee
n si
mila
r g
aps
of
und
erst
and
ing
ab
out
the
ext
erna
l aud
it.
The
reg
ulat
ory
fram
ewo
rk a
llow
s fo
r p
arlia
men
tary
bo
die
s at
fe
der
al a
nd s
ub-n
atio
nal l
evel
s to
scr
utin
ise
bo
th t
he b
udg
et a
nd
bud
get
ary
out
-tur
n, b
ut in
pra
ctic
e lit
tle t
ime
is g
iven
to
thi
s –
esp
ecia
lly a
t th
e su
b-n
atio
nal l
evel
. The
re a
re a
lso
issu
es w
ith t
he
cap
acity
of p
arlia
men
taria
ns, a
nd in
par
ticul
ar w
heth
er t
here
is a
n un
der
stan
din
g o
f wha
t th
e ro
le o
f a p
arlia
men
taria
n is
. Thi
s al
so
imp
acts
on
the
effe
ctiv
enes
s o
f the
ext
erna
l aud
it fu
nctio
n.
69
etH
iop
iA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(Ho
w w
ell?
Ho
w u
sefu
l?)
exp
lana
tio
ns(W
hy/W
hy n
ot?
)
On
trea
sury
Man
y ai
d m
od
aliti
es a
re o
n tr
easu
ry. P
artic
ular
ly n
ota
ble
is a
m
ove
aw
ay fr
om
Cha
nnel
2 a
nd C
hann
el 3
sup
po
rt fr
om
mo
st
do
nors
, to
war
ds
Cha
nnel
1 –
incr
easi
ngly
sup
po
rt is
go
ing
thr
oug
h g
ove
rnm
ent
trea
sury
sys
tem
s. F
or
exam
ple
: (i)
Pro
tect
ing
Bas
ic
Serv
ices
use
s C
hann
el 1
in C
om
po
nent
1, b
ut C
om
po
nent
2 u
ses
IDA
pro
cure
men
t p
roce
dur
es a
nd C
om
po
nent
4 is
Cha
nnel
3. (
ii)
The
Pro
duc
tive
Safe
ty N
et P
rog
ram
me
is c
om
ing
on
to C
hann
el 1
. Th
e is
sues
wer
e ar
oun
d re
po
rtin
g r
athe
r th
an d
isb
urse
men
t, b
ut
all f
und
s g
o t
hro
ugh
trea
sury
sys
tem
s at
fed
eral
and
sub
-nat
iona
l le
vels
. At
sub
-nat
iona
l lev
el, t
he P
rod
uctiv
e Sa
fety
Net
Pro
gra
mm
e is
no
t re
po
rted
. (iii
) Pub
lic S
ecto
r C
apac
ity B
uild
ing
Pro
gra
mm
e su
pp
ort
is a
lso
on
trea
sury
(alth
oug
h ID
A p
roce
dur
es a
re u
sed
in
pro
cure
men
t, o
nce
mo
re).
(iv) S
ecto
r su
pp
ort
use
s a
varie
ty o
f m
od
aliti
es, s
om
e o
f whi
ch u
se t
reas
ury
mec
hani
sms
(but
so
me
of
whi
ch a
re C
hann
els
2 an
d 3
).
Onl
y C
hann
el 3
pro
ject
s se
em n
ow
to
be
tota
lly o
ff tr
easu
ry, b
ut
thes
e in
terv
entio
ns a
re d
eclin
ing
in n
umb
er a
nd im
po
rtan
ce. D
ata
pro
vid
ed fo
r th
e PE
FA re
view
sho
ws
a m
ove
to
dire
ct b
udg
et
sup
po
rt b
y d
ono
rs in
the
fina
ncia
l yea
rs 2
003
and
200
4, a
tre
nd t
hat
was
onl
y in
terr
upte
d b
y d
ifficu
lties
ass
oci
ated
with
the
ele
ctio
n.
Ho
wev
er, s
om
e m
ajo
r d
ono
rs (e
.g. t
he U
S) s
till c
hann
el a
lot
of
sup
po
rt t
houg
h C
hann
el 3
.
If am
oun
ts a
re b
udg
eted
, the
use
of b
ank
acco
unts
nee
ds
trea
sury
ap
pro
val.
Ban
k ac
coun
t us
e is
ver
y st
rictly
co
ntro
lled
. But
so
me
ban
k ac
coun
ts –
e.g
. fo
r su
pp
ort
to
go
vern
men
t fu
nds
(road
fund
, etc
.) –
are
not
stric
tly s
pea
king
und
er t
he d
irect
co
ntro
l of t
he g
ove
rnm
ent,
ev
en t
houg
h th
e fu
nd it
self
can
be
seen
as
an a
gen
cy o
f the
g
ove
rnm
ent,
with
ap
pro
pria
tions
aut
horis
ed b
y p
arlia
men
t, e
tc.
In s
om
e in
stan
ces,
the
re h
ave
bee
n in
cent
ives
for
sub
-nat
iona
l ju
risd
ictio
ns n
ot
to d
iscl
ose
par
ticul
ar e
lem
ents
of s
upp
ort
(p
artic
ular
ly s
upp
ort
go
ing
to
sec
tor
bur
eaux
) to
avo
id p
oss
ible
o
ffset
in t
he b
lock
gra
nt. H
ow
ever
, thi
s is
bec
om
ing
incr
easi
ngly
d
ifficu
lt b
ecau
se re
form
pro
cess
es a
re g
rad
ually
imp
lem
entin
g
sing
le t
reas
ury
syst
ems
coun
try
wid
e. T
his,
with
rig
oro
us c
ont
rol
ove
r b
ank
acco
unts
, mea
ns t
hat
any
sup
po
rt g
oin
g t
o s
ub-n
atio
nal
sect
ors
is u
sual
ly b
eing
cap
ture
d in
tre
asur
y sy
stem
s.
Ther
e ar
e st
ill in
cent
ives
to
mai
ntai
n m
ore
tra
diti
ona
l “st
ove
-p
ipe”
gra
nts
to s
ecto
rs, p
artic
ular
ly a
t th
e su
b-n
atio
nal l
evel
. But
a
tech
nica
l ref
orm
is m
akin
g t
his
loo
pho
le m
ore
diffi
cult
to u
se. I
t se
ems
that
ind
ivid
ual d
ono
r b
ehav
iour
as
muc
h as
any
thin
g e
lse
is
a fa
cto
r he
re.
Ther
e ar
e a
num
ber
of i
ncen
tives
for
wan
ting
to
avo
id u
sing
the
g
ove
rnm
ent
trea
sury
sys
tem
s:
Polit
ical
– T
his
bec
ame
par
ticul
arly
pro
min
ent
afte
r th
e 20
05
elec
tions
, and
rela
tes
to d
ono
r re
luct
ance
to
bei
ng a
sso
ciat
ed w
ith
wha
t th
ey s
ee a
s a
po
or
hum
an r
ight
s re
cord
on
the
par
t o
f the
g
ove
rnm
ent
of E
thio
pia
.
Offs
et –
Offs
et o
f the
blo
ck g
rant
is e
xplic
itly
mad
e ag
ains
t d
ono
r su
pp
ort
giv
en t
o re
gio
ns (e
.g. r
egio
nal b
udg
et s
upp
ort
). Th
is
dem
ons
trat
es t
he g
ove
rnm
ent’s
co
mm
itmen
t to
the
prin
cip
le o
f eq
uity
bet
wee
n re
gio
ns a
nd it
s b
elie
f tha
t th
is is
co
mp
rom
ised
b
y th
e us
e o
f dis
trib
utio
n ch
anne
ls o
utsi
de
thei
r o
wn.
But
thi
s ha
s af
fect
ed o
n b
udg
et a
nd o
n tr
easu
ry is
sues
, bec
ause
in t
he p
ast
do
nors
hav
e so
ught
to
giv
e su
pp
ort
to
ent
ities
or
pro
gra
mm
es a
t su
b-n
atio
nal l
evel
, avo
idin
g c
ore
go
vern
men
t d
isb
urse
men
t an
d
rep
ort
ing
sys
tem
s.
Ad
diti
onal
ity –
The
nee
d t
o s
how
tha
t d
ono
r su
pp
ort
has
resu
lted
in
gre
ater
reso
urce
s at
sub
-nat
iona
l lev
el h
as h
eavi
ly in
fluen
ced
the
d
esig
n o
f the
Pro
tect
ing
Bas
ic S
ervi
ces
mo
dal
ity. B
ut in
the
pas
t d
ono
rs a
pp
ear
to h
ave
bee
n re
luct
ant
to t
rust
go
vern
men
t sy
stem
s fo
r th
is.
Syst
emic
– D
ono
rs h
ave
sig
nific
ant
do
ubts
ab
out
the
effi
cien
cy o
f g
ove
rnm
ent
syst
ems
to d
eliv
er s
upp
ort
whe
re it
is n
eed
ed, o
r –
criti
cally
– a
bo
ut fi
duc
iary
issu
es, i
.e. t
hat
fund
s w
ill n
ot
be
used
for
thei
r in
tend
ed p
urp
ose
s, o
r m
ay e
ven
leak
fro
m t
he s
yste
m. R
ecen
t p
rog
ress
in P
FM re
form
has
mad
e th
is d
ono
r p
osi
tion
less
ten
able
.
70
etH
iop
iA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(Ho
w w
ell?
Ho
w u
sefu
l?)
exp
lana
tio
ns(W
hy/W
hy n
ot?
)
On
acco
unt
All
Cha
nnel
1 fu
nds
use
the
go
vern
men
t ac
coun
ting
sys
tem
.
Cha
nnel
3 s
upp
ort
do
es n
ot
use
the
go
vern
men
t ac
coun
ting
sy
stem
, but
the
re a
re re
gul
atio
ns t
o e
nsur
e th
at re
po
rtin
g s
uch
activ
ities
follo
ws
go
vern
men
t fo
rmat
s, w
hich
allo
ws
them
to
be
put
o
n re
po
rt.
Sup
po
rt t
o o
ff b
udg
et fu
nds,
suc
h as
the
fuel
fund
, the
Eth
iop
ian
Ro
ads
Aut
horit
y, e
tc.,
is a
lso
str
ictly
sp
eaki
ng o
ff ac
coun
t, a
s th
ese
entit
ies
have
the
ir o
wn
acco
untin
g s
yste
ms.
Ho
wev
er, t
he in
itial
d
isb
urse
men
ts w
ill b
e o
n tr
easu
ry a
nd a
lso
on
par
liam
ent,
and
fund
in
form
atio
n is
rep
rod
uced
as
a fo
rm o
f no
te t
o t
he a
cco
unts
.
Ther
e ha
s b
een
a te
nden
cy t
o c
reat
e hy
brid
fund
ing
cha
nnel
s,
whe
re m
one
y is
dire
cted
sim
ilarly
to
Cha
nnel
1 fu
ndin
g, b
ut
with
sep
arat
e re
po
rtin
g re
qui
rem
ents
. But
the
cur
rent
tre
nd is
in
crea
sing
ly t
o u
se g
ove
rnm
ent
acco
untin
g s
yste
ms.
The
go
vern
men
t ac
coun
ting
sys
tem
allo
ws
for
exp
end
iture
to
be
clas
sifie
d b
y so
urce
. The
cha
rt o
f acc
oun
ts fo
llow
s th
e G
ove
rnm
ent
Fina
nce
Stat
istic
s m
od
el. C
od
ing
allo
ws
exp
end
iture
to b
e id
entifi
ed
bo
th in
ter
ms
of s
our
ce o
f fina
ncin
g a
nd d
etai
led
exp
end
iture
dat
a (in
clud
ing
all
exp
end
iture
thr
oug
h su
b-n
atio
nal j
uris
dic
tions
). Th
e IB
EX
(go
vern
men
t ac
coun
ting
) sys
tem
at
bo
th fe
der
al a
nd s
ub-
natio
nal l
evel
s ap
pea
rs t
o p
rovi
de
very
co
ncis
e in
form
atio
n ab
out
b
oth
the
so
urce
s o
f fun
ds
and
the
ir us
e.
In t
he p
ast,
fina
ncia
l inf
orm
atio
n ha
s no
t al
way
s b
een
timel
y, fo
r b
oth
in-y
ear
rep
ort
ing
and
the
clo
sing
of a
cco
unts
. At
one
sta
ge
ther
e w
ere
four
-yea
r b
ackl
og
s. L
ack
of t
imel
ines
s w
as p
artic
ular
ly a
n is
sue
for
the
cons
olid
ated
acc
oun
ts s
how
ing
fed
eral
and
reg
iona
l in
form
atio
n, d
ue t
o c
apac
ity c
ons
trai
nts
at t
he w
eake
st w
ered
as
(dis
tric
t ad
min
istr
atio
ns).
Thes
e is
sues
hav
e b
een
reso
lved
, mai
nly
due
to
suc
cess
ful r
efo
rm p
roce
sses
.
Stea
dy
pro
gre
ss in
PFM
refo
rm is
per
hap
s o
ne c
ritic
al re
aso
n w
hy t
here
has
bee
n a
mo
ve t
ow
ard
s C
hann
el 1
sup
po
rt. T
he
Dec
entr
alis
atio
n Su
pp
ort
Act
ivity
Pro
ject
, par
t o
f the
go
vern
men
t’s
Exp
end
iture
Man
agem
ent
and
Co
ntro
l Pro
gra
mm
e, h
as a
dd
ress
ed
a nu
mb
er o
f iss
ues,
incl
udin
g: (
i) a
new
cha
rt o
f acc
oun
ts a
nd c
od
ing
st
ruct
ures
; (ii)
a m
ove
to
do
uble
-ent
ry b
oo
kkee
pin
g; (
iii) a
mo
ve t
o
sing
le t
reas
ury
acco
untin
g; a
nd (i
v) t
he in
teg
ratio
n o
f bud
get
and
ac
coun
ting
sys
tem
s.
As
a re
sult,
acc
oun
ting
sys
tem
s ar
e m
ore
cre
dib
le a
nd p
rod
uce
mo
re t
imel
y in
form
atio
n. T
his
has,
for
man
y d
ono
rs, r
emo
ved
the
im
ped
imen
t to
co
nsid
erin
g C
hann
el 1
fund
ing
mo
dal
ities
.
On
aud
itTh
e au
dits
of t
he fe
der
al A
udito
r G
ener
al w
ill, a
s a
mat
ter
of c
our
se,
pic
k up
all
Cha
nnel
1 s
upp
ort
.
Aud
iting
act
iviti
es u
nder
the
Pro
tect
ing
Bas
ic S
ervi
ces
mo
dal
ity
are
do
ubly
sup
po
rted
by
the
cont
inuo
us a
udit
pro
cess
, whe
re
acco
untin
g fi
rms
app
oin
ted
by
the
Aud
itor
Gen
eral
aud
it th
e us
e o
f Pro
tect
ing
Bas
ic S
ervi
ces
mo
nies
to
ens
ure
add
itio
nalit
y an
d
that
the
y w
ere
used
for
sub
-nat
iona
l ser
vice
del
iver
y ac
tiviti
es
(rat
her
than
for
oth
er a
ctiv
ities
no
t re
late
d t
o d
irect
ser
vice
d
eliv
ery,
e.g
. def
ence
exp
end
iture
). Th
is is
a p
artic
ular
cha
nge
in
focu
s fr
om
tha
t o
f tra
diti
ona
l aud
iting
, whi
ch u
sual
ly lo
oks
onl
y at
th
e ap
pro
pria
tene
ss o
f exp
end
iture
(and
sup
po
rtin
g s
yste
ms
of
reco
rdin
g, e
tc.),
plu
s, a
t b
est,
giv
es s
om
e co
nsid
erat
ion
to e
ffici
ency
an
d e
ffect
iven
ess.
All
IDA
fund
s (lo
ans
and
gra
nts)
are
sub
ject
to
aud
it sc
rutin
y th
roug
h a
pro
cess
ove
rsee
n b
y th
e A
udito
r G
ener
al (a
ltho
ugh
the
Aud
itor
Gen
eral
may
ap
po
int
an e
xecu
ting
ag
ent)
.
UN
pro
cess
es, s
uch
as N
atio
nal E
xecu
tion
Pro
ject
s, a
re a
lso
sub
ject
to
ext
erna
l aud
it b
y th
e A
udito
r G
ener
al.
Ther
e ha
ve b
een
inst
ance
s, b
oth
in a
udit
field
wo
rk a
nd in
co
mm
ents
in
aud
it re
po
rts,
of t
he A
udito
r G
ener
al p
icki
ng u
p a
nom
alie
s in
the
d
iscl
osu
re o
f aid
(fo
r in
stan
ce, a
pp
aren
t d
iscr
epan
cies
bet
wee
n fo
od
ai
d b
udg
ets
and
eve
ntua
l out
-tur
n).
Ther
e ar
e a
num
ber
of d
oub
ts a
bo
ut t
he q
ualit
y o
f ext
erna
l aud
it p
roce
sses
in E
thio
pia
.
Firs
t, t
here
are
issu
es a
bo
ut a
udit
cove
rag
e an
d t
he re
spec
tive
man
dat
es o
f the
fed
eral
Aud
itor
Gen
eral
and
reg
iona
l Aud
itors
G
ener
al. I
n p
ract
ice,
sub
-nat
iona
l co
vera
ge
of f
ront
line
exp
end
iture
–
muc
h o
f whi
ch w
ill h
ave
bee
n fu
nded
by
loan
or
gra
nt –
is p
atch
y (h
ence
the
nee
d fo
r co
ntin
uous
aud
it o
f Pro
tect
ing
Bas
ic S
ervi
ces
– in
a s
ense
, an
atte
mp
t to
byp
ass
stan
dar
d e
xter
nal a
uditi
ng
syst
ems)
. Whi
le t
he d
isb
urse
men
t o
f the
blo
ck g
rant
to
reg
ions
(w
hich
will
co
ntai
n el
emen
ts o
f fo
reig
n as
sist
ance
) is
aud
ited
by
the
fed
eral
Aud
itor
Gen
eral
, the
use
of s
uch
mo
nies
falls
und
er
the
cont
rol o
f reg
iona
l Aud
itor
Gen
eral
s. B
ut t
he re
gio
nal A
udito
r G
ener
als’
ext
erna
l aud
it ca
n b
e ve
ry li
mite
d in
dee
d (s
om
etim
es,
acco
rdin
g t
o t
he re
gio
nal P
EFA
, les
s th
an 3
0%),
due
to
insu
ffici
ent
reso
urce
s o
f the
ir o
ffice
s. A
nd c
apac
ity d
evel
op
men
t un
der
take
n b
y th
e fe
der
al A
udito
r G
ener
al is
no
t m
irro
red
in m
any
of t
he re
gio
nal
offi
ces
(alth
oug
h T
igra
y an
d O
rom
iya
cam
e o
ut o
f the
reg
iona
l PE
FA a
sses
smen
t w
ell,
and
Am
hara
, a re
gio
n no
t co
vere
d b
y th
e as
sess
men
t, h
as a
lso
a re
put
atio
n fo
r a
stro
ng o
ffice
). N
ot
onl
y is
met
hod
olo
gy
lack
ing
, ins
titut
iona
l fra
mew
ork
s ar
e al
so w
eak.
Th
ere
is a
wid
esp
read
lack
of u
nder
stan
din
g a
mo
ng t
he m
ajo
r st
akeh
old
ers
of t
he ro
le o
f an
Aud
itor
Gen
eral
, res
ultin
g, a
mo
ng
oth
ers,
in p
oo
r p
arlia
men
tary
revi
ew a
nd fo
llow
-up
.
Seco
nd, a
udit
rep
ort
s ha
ve n
ot
bee
n tim
ely.
Bac
klo
gs
have
bee
n se
vere
, and
the
imp
act
of a
udit
rep
ort
s, h
ow
ever
go
od
tho
se
rep
ort
s m
ight
be,
has
bee
n co
mp
rom
ised
bec
ause
the
info
rmat
ion
is o
ut o
f dat
e. A
cco
untin
g re
form
s ha
ve n
ow
larg
ely
rem
edie
d t
his,
an
d b
ackl
og
s ar
e ab
out
to
be
elim
inat
ed. N
ever
thel
ess,
thi
s is
sue
has
had
an
imp
act.
Aud
its h
ave
bee
n d
elay
ed b
ecau
se o
f acc
oun
ting
bac
klo
gs
rath
er
than
del
ays
in t
he a
uditi
ng fu
nctio
n its
elf.
This
is g
rad
ually
bei
ng
rem
edie
d t
hro
ugh
the
ong
oin
g re
form
s.
One
are
a w
here
Eth
iop
ia’s
oth
erw
ise
stea
dily
imp
ress
ive
refo
rm
pro
cess
has
had
no
imp
act
(and
ver
y lit
tle p
rog
ress
) is
stre
ngth
enin
g
the
inst
itutio
nal f
ram
ewo
rks
in w
hich
the
Aud
itor
Gen
eral
can
o
per
ate
(at
bo
th fe
der
al a
nd s
ub-n
atio
nal l
evel
s). P
arlia
men
taria
ns,
at fe
der
al a
nd re
gio
nal l
evel
s in
par
ticul
ar, d
o n
ot
seem
to
fully
un
der
stan
d a
udit
pro
cess
es, a
nd a
re p
artia
lly re
spo
nsib
le fo
r re
gio
nal a
udit
offi
ces
bei
ng, i
n th
e m
ain,
sub
stan
tially
und
er-
reso
urce
d.
71
etH
iop
iA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(Ho
w w
ell?
Ho
w u
sefu
l?)
exp
lana
tio
ns(W
hy/W
hy n
ot?
)
On
rep
ort
Info
rmat
ion
abo
ut a
id fl
ow
s is
pro
duc
ed b
y th
e M
ultil
ater
al a
nd
Bila
tera
l Aid
Dep
artm
ents
in t
he M
inis
try
of F
inan
ce a
nd E
cono
mic
D
evel
op
men
t. T
his
info
rmat
ion
is n
ot
usua
lly a
vaila
ble
in t
he p
ublic
d
om
ain.
Info
rmat
ion
abo
ut t
he b
udg
et –
whi
ch in
clud
es m
ost
aid
flo
ws
– is
av
aila
ble
in t
he F
eder
al N
egar
it G
azet
a, t
he o
ffici
al g
ove
rnm
ent
pap
er.
Ho
wev
er, t
here
is n
o p
ublic
acc
ess
to t
he in
-yea
r b
udg
et e
xecu
tion
rep
ort
s. B
ut t
he M
inis
try
of F
inan
ce a
nd E
cono
mic
Dev
elo
pm
ent
web
site
pro
vid
es q
uart
erly
info
rmat
ion
on
in-y
ear
bud
get
exe
cutio
n.
The
year
-end
fina
ncia
l sta
tem
ents
pre
par
ed b
y th
e M
inis
try
of
Fina
nce
and
Eco
nom
ic D
evel
op
men
t ar
e no
t av
aila
ble
in t
he p
ublic
d
om
ain.
Fin
al a
cco
unts
with
aud
it re
po
rts
fro
m t
he fe
der
al A
udito
r G
ener
al, w
hile
the
ore
tical
ly a
vaila
ble
to
the
pub
lic, a
re, i
n p
ract
ice,
in
lim
ited
circ
ulat
ion.
Som
e ad
diti
ona
l inf
orm
atio
n is
ava
ilab
le t
o d
ono
rs. B
oth
the
ed
ucat
ion
and
hea
lth s
ecto
r d
evel
op
men
t p
rog
ram
mes
pro
duc
e p
erio
dic
rep
ort
s (w
hich
refle
ct a
ll d
isb
urse
men
t fr
om
do
nors
, re
gar
dle
ss o
f cha
nnel
s).
Spec
ific
rep
ort
ing
is a
lso
ava
ilab
le fo
r th
e Pu
blic
Sec
tor
Cap
acity
B
uild
ing
Pro
gra
mm
e an
d fo
r th
e Pr
od
uctiv
e Sa
fety
Net
Pro
gra
mm
e.
The
Pro
tect
ing
Bas
ic S
ervi
ces
pro
gra
mm
e al
so p
rod
uces
in
form
atio
n o
n b
asic
ser
vice
sp
end
ing
, alth
oug
h th
is (p
artic
ular
ly
thro
ugh
the
cont
inuo
us a
udit
pro
cess
) see
ms
to b
e fo
r th
e b
enefi
t o
f do
nors
.
Ther
e ha
ve lo
ng b
een
dis
crep
anci
es b
etw
een
go
vern
men
t fig
ures
(pro
duc
ed in
the
go
vern
men
t g
azet
te, e
tc.)
and
tho
se o
f the
inte
rnat
iona
l co
mm
unity
.
The
qua
lity
of c
aptu
re c
oul
d b
e im
pro
ved
if t
he A
id M
anag
emen
t Pl
atfo
rm w
as a
ble
to
func
tion
pro
per
ly, a
llow
ing
eac
h d
ono
r to
inp
ut
com
mitm
ent
and
dis
bur
sem
ent
dat
a d
irect
ly. T
his
info
rmat
ion
coul
d b
e us
ed b
y th
e M
inis
try
of F
inan
ce a
nd E
cono
mic
Dev
elo
pm
ent
at
com
mitm
ent
for
pla
nnin
g p
urp
ose
s, a
nd c
oul
d b
e tr
iang
ulat
ed w
hen
fund
s ar
e d
isb
urse
d. B
ut t
he re
al a
dva
ntag
e is
tha
t th
e p
latf
orm
is a
tr
ansp
aren
t p
ort
al w
hich
effe
ctiv
ely
calls
do
nors
to
acc
oun
t an
d c
an p
rovi
de
info
rmat
ion
in t
he p
ublic
do
mai
n.
72
Ke
nyA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/w
hy n
ot?
)
On
pla
nTh
ere
is s
igni
fican
t ev
iden
ce t
hat
effo
rts
are
bei
ng m
ade
to in
clud
e ai
d in
sp
end
ing
pla
ns:
The
Inve
stm
ent
Plan
of t
he E
cono
mic
Rec
ove
ry S
trat
egy
pro
vid
es a
co
mp
rehe
nsiv
e, o
wn
and
dev
elo
pm
ent
par
tner
reso
urce
fram
ewo
rk.
At
min
iste
rial l
evel
, aid
is s
upp
ose
d t
o b
e in
clud
ed in
Min
iste
rial
Pub
lic E
xpen
ditu
re R
evie
ws.
In m
inis
trie
s th
at a
re s
igni
fican
t re
cip
ient
s o
f aid
and
wo
rkin
g t
ow
ard
s a
SWA
p, t
here
are
inte
gra
ted
st
rate
gic
pla
ns s
upp
ort
ed b
y an
nual
op
erat
iona
l pla
ns p
rovi
din
g
a fr
amew
ork
for
cap
turin
g a
id o
n p
lan.
Aid
is a
lso
incl
uded
in t
he
SWG
rep
ort
s.
Aid
is in
clud
ed in
Dis
tric
t D
evel
op
men
t Pl
ans,
by
pro
ject
and
by
do
nor.
Mid
way
bet
wee
n th
e p
lann
ing
pro
cess
and
the
bud
get
are
the
B
udg
et O
utlo
ok
Pap
er a
nd t
he B
udg
et S
trat
egy
Pap
er. A
id is
ca
ptu
red
in t
hese
do
cum
ents
in t
he fo
llow
ing
way
s: (i
) It
form
s p
art
of t
he fi
scal
fram
ewo
rk a
nd s
ecto
r ce
iling
s p
rese
nted
in t
he
Bud
get
Out
loo
k Pa
per
. At
fisca
l fra
mew
ork
leve
l, ex
tern
al g
rant
s an
d lo
ans
have
a s
epar
ate
line.
(ii)
It is
eve
n m
ore
exp
licitl
y ca
ptu
red
in
the
Bud
get
Str
ateg
y Pa
per
, at
fram
ewo
rk le
vel a
nd a
t th
e le
vel o
f in
div
idua
l sp
end
ing
cei
ling
s.
The
qua
lity
and
co
mp
lete
ness
of t
he in
form
atio
n in
inst
rum
ents
su
ch a
s th
e M
inis
teria
l Pub
lic E
xpen
ditu
re R
evie
ws
diff
er a
cro
ss
min
istr
ies.
Eve
n if
info
rmat
ion
is c
aptu
red
, it
is n
ot
alw
ays
cert
ain
that
the
re a
re s
ubst
antiv
e p
lann
ing
pro
cess
es b
ehin
d t
he
do
cum
enta
tion
to e
nsur
e in
crea
sed
aid
effe
ctiv
enes
s. R
ecip
ient
s o
f hig
h am
oun
ts o
f aid
who
hav
e p
ut S
WA
ps
and
bas
ket
fund
ar
rang
emen
ts in
, o
r ar
e in
the
pro
cess
of d
oin
g t
his,
cap
ture
aid
p
artic
ular
ly w
ell o
n p
lan.
The
narr
ativ
e o
n ai
d fl
ow
s in
the
Bud
get
Str
ateg
y Pa
per
co
uld
be
mo
re c
om
ple
te, p
artic
ular
ly a
t m
inis
try
leve
l.
Ken
ya h
as re
spo
nded
wel
l to
find
ing
s in
the
last
few
yea
rs t
hat
the
go
vern
men
t d
oes
no
t p
rovi
de
suffi
cien
t le
ader
ship
in in
teg
ratin
g a
id
in p
lann
ing
pro
cess
es a
nd in
stru
men
ts, a
nd t
hat
ther
e is
a h
arm
ful
sep
arat
ion
bet
wee
n d
evel
op
men
t an
d re
curr
ent
spen
din
g p
lann
ing
. Th
e in
teg
ratio
n b
etw
een
dev
elo
pm
ent
and
recu
rren
t sp
end
ing
, and
d
ono
r an
d g
ove
rnm
ent
fund
ing
, is
the
resu
lt o
f the
go
vern
men
t’s
des
ire t
o in
teg
rate
pla
nnin
g in
stru
men
ts a
nd p
ut a
n ef
fect
ive
bud
get
pre
par
atio
n p
roce
ss in
pla
ce.
In m
inis
trie
s w
hich
rece
ive
hig
h le
vels
of a
id, p
artic
ular
ly c
ond
uciv
e en
viro
nmen
ts h
ave
bee
n cr
eate
d (s
upp
ort
ed b
y si
gni
fican
t TA
) fo
r in
teg
rate
d p
lann
ing
for
go
vern
men
t an
d d
evel
op
men
t p
artn
er
fund
ing
and
the
dev
elo
pm
ent
and
recu
rren
t b
udg
ets.
In m
inis
trie
s w
here
aid
info
rmat
ion
is w
eake
r an
d t
he in
teg
ratio
n b
etw
een
sour
ces
of f
und
ing
and
the
dev
elo
pm
ent
and
recu
rren
t b
udg
ets
is n
ot
effe
ctiv
e, b
udg
etin
g is
stil
l do
ne in
iso
latio
n fr
om
pla
nnin
g a
nd in
line
with
mo
re h
isto
rical
ap
pro
ache
s. N
ew
req
uire
men
ts b
y th
e M
inis
try
of F
inan
ce, s
uch
as t
he M
inis
teria
l Pu
blic
Exp
end
iture
Rev
iew
s an
d p
artic
ipat
ion
in t
he S
WG
s, a
re m
ore
a
mat
ter
of c
om
plia
nce
than
driv
ers
of r
efo
rm a
t m
inis
try
leve
l. O
ne
cont
ribut
ing
fact
or
is t
he s
epar
atio
n o
f the
Min
istr
y o
f Fin
ance
and
M
inis
try
of P
lann
ing
and
Dev
elo
pm
ent
at c
entr
al le
vel,
whi
ch re
sults
in
ove
rlap
pin
g re
qui
rem
ents
, ove
rbur
den
ing
min
istr
ies.
73
Ke
nyA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/w
hy n
ot?
)
On
bud
get
Aid
is c
aptu
red
wel
l on
bud
get
. Acc
ord
ing
to
the
200
6 Su
rvey
o
n M
oni
torin
g t
he P
aris
Dec
lara
tion,
91%
of a
id d
isb
urse
d t
o t
he
go
vern
men
t w
as c
aptu
red
on
bud
get
. Thi
s is
refle
cted
by
the
AFR
OD
AD
rep
ort
, whi
ch s
tate
s th
at, r
ecen
tly, c
ove
rag
e o
f aid
in t
he
bud
get
is m
uch
mo
re c
om
ple
te. T
he g
ove
rnm
ent’s
ow
n b
oo
klet
o
n ai
d o
n th
e d
evel
op
men
t b
udg
et fo
r 20
05/0
6 st
ates
tha
t 87
% o
f ex
pec
ted
do
nor
dis
bur
sem
ents
is re
flect
ed o
n th
e d
evel
op
men
t b
udg
et. T
he E
xter
nal R
eso
urce
s D
epar
tmen
t an
d t
he B
udg
et
Sup
plie
s D
epar
tmen
t ar
e co
nfid
ent
that
all
aid
is c
aptu
red
on
bud
get
, par
ticul
arly
bec
ause
thi
s is
leg
ally
req
uire
d.
Aid
is c
aptu
red
on
bud
get
in t
he fo
llow
ing
way
s:
In t
he re
curr
ent b
udg
et, e
xpec
ted
bud
get
sup
po
rt fl
ow
s ar
e sh
ow
n in
an
alte
rnat
ive
fund
ing
sce
nario
. Ken
ya a
s a
rule
do
es n
ot
incl
ude
exp
ecte
d b
udg
et s
upp
ort
in it
s m
ain
fisca
l fra
mew
ork
and
bud
get
. Th
e re
curr
ent
bud
get
est
imat
es in
clud
e an
alte
rnat
ive
scen
ario
ta
ble
, whi
ch s
how
s ho
w b
udg
et s
upp
ort
wo
uld
be
used
by
sect
or,
sho
uld
it m
ater
ialis
e.
In t
he d
evel
opm
ent b
udg
et e
stim
ates
, sig
nific
ant
det
ail o
n ai
d
flow
s is
pre
sent
ed. I
n th
e su
mm
ary
tab
les,
aid
is s
how
n b
y m
inis
try,
w
heth
er it
is lo
ans
or
gra
nts,
and
whe
ther
it is
dis
bur
sed
thr
oug
h th
e g
ove
rnm
ent
or
app
rop
riate
d a
s A
pp
rop
riatio
ns in
Aid
. It
is
also
sho
wn
by
dev
elo
pm
ent
par
tner
, by
sect
or
and
by
pro
ject
or
pro
gra
mm
e, b
y lo
an o
r g
rant
and
by
app
rop
riatio
ns in
aid
or
flow
ing
th
roug
h th
e g
ove
rnm
ent
syst
em. F
urth
er d
etai
l is
then
pro
vid
ed
for
each
min
istr
y, w
here
dev
elo
pm
ent
bud
get
exp
end
iture
s ar
e se
t o
ut b
y b
udg
et h
ead
, by
pro
ject
or
pro
gra
mm
e, a
nd b
y ec
ono
mic
cl
assi
ficat
ion
agai
nst
the
sour
ce o
f fun
din
g.
The
cap
ture
is q
uite
co
mp
rehe
nsiv
e an
d d
etai
led
.
The
lack
of n
arra
tive
with
bud
get
do
cum
enta
tion
affe
cts
dev
elo
pm
ent
par
tner
fund
s as
muc
h as
ow
n re
venu
e.
The
lack
of i
nteg
ratio
n o
f the
recu
rren
t an
d d
evel
op
men
t b
udg
ets
det
ract
s fr
om
par
liam
ent’s
ab
ility
to
vie
w a
id s
upp
ort
in t
he c
ont
ext
of o
vera
ll sp
end
ing
.
The
sep
arat
e ca
ptu
re o
f bud
get
sup
po
rt in
the
recu
rren
t b
udg
et is
o
n ac
coun
t o
f the
unp
red
icta
bili
ty o
f the
se re
sour
ces.
Whe
n b
udg
et
sup
po
rt w
as a
ffect
ed in
200
4, t
he g
ove
rnm
ent
mad
e a
bud
get
rul
e no
t to
incl
ude
it in
the
firs
t fis
cal f
ram
ewo
rk s
cena
rio, b
ut t
o p
lan
for
it th
roug
h a
seco
nd s
cena
rio, r
eflec
ted
in t
he b
udg
et e
stim
ates
b
ut n
ot
exp
licitl
y vo
ted
. If t
he b
udg
et s
upp
ort
do
es m
ater
ialis
e, it
is
vote
d s
ubse
que
ntly
thr
oug
h an
ad
just
men
t b
udg
et.
The
fram
ewo
rk fo
r ca
ptu
ring
aid
thr
oug
h th
e d
evel
op
men
t b
udg
et
has
long
bee
n in
pla
ce. I
n re
cent
yea
rs t
he q
ualit
y o
f the
info
rmat
ion
has
imp
rove
d. T
he 2
006
Surv
ey o
n M
oni
torin
g t
he P
aris
Dec
lara
tion
offe
rs s
ever
al e
xpla
natio
ns fo
r w
hy a
id d
isb
urse
men
t is
hig
her
than
ai
d p
lann
ed o
n b
udg
et: i
n su
mm
ary,
it h
as m
ore
to
do
with
do
nors
ad
just
ing
pla
ns d
urin
g t
he y
ear
(par
tly o
n ac
coun
t o
f inc
om
pat
ible
fis
cal y
ears
) tha
n w
ith d
elib
erat
ely
leav
ing
aid
off
bud
get
. “A
ran
ge
of f
acto
rs e
xpla
in t
he s
hort
fall
of a
id re
cord
ed o
n b
udg
et in
rela
tion
to a
id d
isb
urse
d t
o t
he g
ove
rnm
ent
sect
or.
On
the
go
vern
men
t si
de,
for
man
y se
cto
rs c
lear
and
fully
co
sted
pla
ns w
ith w
hich
do
nors
ca
n al
ign
thei
r as
sist
ance
hav
e no
t b
een
dev
elo
ped
. On
the
do
nor
sid
e, s
om
e d
ono
rs a
re in
the
hab
it o
f sp
end
ing
fund
s d
irect
ly
with
out
info
rmin
g t
he g
ove
rnm
ent,
and
of f
ailin
g t
o c
ons
ult
the
bud
get
est
imat
es a
nd a
s a
resu
lt so
met
imes
pro
vid
ing
exc
essi
ve
aid
to
par
ticul
ar p
roje
cts.
Bo
th p
ract
ices
are
bo
und
to
red
uce
the
pro
po
rtio
n o
f aid
to
go
vern
men
t w
hich
is re
po
rted
in t
he b
udg
et. I
n ad
diti
on,
rep
ort
ing
pro
ced
ures
are
no
t st
rictly
follo
wed
, and
the
re
are
inco
nsis
tenc
ies
bet
wee
n g
ove
rnm
ent
and
do
nor
app
roac
hes,
an
d m
ism
atch
ed fi
scal
yea
rs. A
s a
resu
lt, b
oth
the
go
vern
men
t an
d
do
nors
fail
to a
cco
unt
adeq
uate
ly fo
r ai
d re
sour
ces.
”2 The
se v
iew
s ar
e re
flect
ed in
the
Ken
ya E
xter
nal R
eso
urce
s Po
licy.
The
sep
arat
ion
of t
he re
curr
ent
and
dev
elo
pm
ent
bud
get
s is
long
-st
and
ing
pra
ctic
e. In
rece
nt y
ears
, the
dev
elo
pm
ent
bud
get
has
b
een
coo
rdin
ated
fro
m t
he s
ame
dep
artm
ent
in t
he M
inis
try
of
Fina
nce
as
the
recu
rren
t b
udg
et: t
he in
tro
duc
tion
of M
inis
teria
l Pu
blic
Exp
end
iture
Rev
iew
s an
d S
WG
s, w
here
sp
end
ing
on
the
two
b
udg
ets
is t
reat
ed m
ore
co
here
ntly
, has
imp
rove
d p
ract
ice.
74
Ke
nyA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/w
hy n
ot?
)
On
par
liam
ent
Aid
to
the
go
vern
men
t is
ap
pro
pria
ted
by
par
liam
ent
by
law
(the
C
ons
titut
ion
and
the
Ext
erna
l Lo
ans
and
Cre
dit
Act
), w
heth
er it
flo
ws
thro
ugh
go
vern
men
t sy
stem
s o
r no
t. W
hen
it flo
ws
thro
ugh
go
vern
men
t sy
stem
s it
is a
pp
rop
riate
d a
s re
venu
e. W
hen
it flo
ws
thro
ugh
sep
arat
e sy
stem
s it
is a
pp
rop
riate
d a
s A
pp
rop
riatio
ns in
A
id.
Eve
n th
oug
h ag
gre
gat
ed a
nd d
isag
gre
gat
ed in
form
atio
n o
n th
e al
loca
tion
of e
xter
nal r
eso
urce
s is
pro
vid
ed, t
here
is li
ttle
evi
den
ce
of a
str
ong
eng
agem
ent
by
par
liam
ent
on
the
alig
nmen
t o
f tho
se
reso
urce
s w
ith n
atio
nal p
riorit
ies
and
on
whe
ther
the
y ar
e lik
ely
to
be
used
effi
cien
tly, t
rans
par
ently
and
eco
nom
ical
ly fo
r th
e p
urp
ose
s fo
r w
hich
the
y ar
e al
loca
ted
.
Lack
of p
arlia
men
tary
scr
utin
y o
f pro
po
sals
on
the
use
of e
xter
nal
fund
ing
has
less
to
do
with
tra
nsp
aren
cy o
n th
e p
rop
ose
d u
se t
han
with
par
liam
ent’s
wea
k an
d e
ven
inap
pro
pria
te e
ngag
emen
t in
th
e b
udg
et p
roce
ss, a
nd w
eak
cap
acity
. Bef
ore
200
5/06
, whe
n th
e cl
assi
ficat
ion
refo
rms
wer
e in
tro
duc
ed, s
trat
egic
eng
agem
ent
on
aid
al
loca
tion
agai
nst
natio
nal p
riorit
ies
wo
uld
hav
e b
een
diffi
cult
giv
en
the
op
aque
ness
of a
lloca
tions
. Whi
le t
here
is ro
om
for
imp
rove
men
t –
esp
ecia
lly t
he p
rog
ram
mat
ic p
rese
ntat
ion
of e
xpen
ditu
re –
p
arlia
men
t no
w h
as m
uch
bet
ter
info
rmat
ion
to w
ork
with
, inc
lud
ing
in
form
atio
n o
n ai
d d
isb
urse
men
t an
d u
se in
the
pas
t. T
he H
uman
R
ight
s C
om
mis
sio
n in
Ken
ya h
as n
ote
d t
hat
a ke
y w
eakn
ess
in t
he
bud
get
sys
tem
is t
hat
a si
gni
fican
t p
art
of s
pen
din
g e
scap
es p
rop
er
scru
tiny.
On
trea
sury
Aid
in K
enya
is d
isb
urse
d t
hro
ugh
two
cha
nnel
s: (i
) go
vern
men
t sy
stem
s (C
hann
el 1
); o
r (ii
) par
alle
l str
uctu
res,
suc
h as
PIU
s an
d
sect
or
finan
cial
man
agem
ent
agen
ts o
r th
roug
h d
ono
rs t
hem
selv
es
(Cha
nnel
3).
Ther
e is
no
evi
den
ce o
f aid
bei
ng d
isb
urse
d t
hro
ugh
Cha
nnel
2, i
.e. t
hro
ugh
acco
unts
tha
t ar
e w
ithin
the
co
ntro
l of
the
go
vern
men
t b
ut n
ot
und
er t
reas
ury
cont
rol o
r d
irect
scr
utin
y.
Bud
get
sup
po
rt is
pro
vid
ed t
hro
ugh
Cha
nnel
1. P
rog
ram
me
sup
po
rt, i
n th
e fo
rm o
f lo
ans,
gra
nts
and
TA
, is
pro
vid
ed t
hro
ugh
Cha
nnel
s 1
and
3. P
roje
ct s
upp
ort
, in
the
form
of l
oan
s, g
rant
s an
d
TA, i
s p
rovi
ded
thr
oug
h C
hann
els
1 an
d 3
.
The
Ken
ya E
xter
nal R
eso
urce
s Po
licy
do
cum
ent
note
s th
at K
enya
’s re
po
rted
util
isat
ion
of a
id h
as b
een
low
(30%
to
40%
on
aver
age)
. Th
is is
rela
ted
to
the
po
or
pre
dic
tab
ility
of d
isb
urse
men
ts fr
om
d
ono
rs. T
he P
EFA
sco
re (2
006
asse
ssm
ent)
in t
his
reg
ard
is a
D.
Poo
r p
red
icta
bili
ty is
rela
ted
to
pre
-dis
bur
sem
ent
and
dis
bur
sem
ent
cons
trai
nts.
The
go
vern
men
t’s in
abili
ty t
o m
eet
agre
ed c
ond
itio
ns,
inad
equa
te c
oun
terp
art
fund
s, a
nd c
hang
es in
dev
elo
pm
ent
par
tner
co
nditi
ona
lity
mid
-str
eam
, hav
e al
l ad
ded
to
low
pre
dic
tab
ility
.
Low
leve
ls o
f dev
elo
pm
ent
par
tner
tru
st in
go
vern
men
t fin
anci
al
man
agem
ent
and
pro
cure
men
t sy
stem
s ha
ve c
ont
ribut
ed t
o t
he
esta
blis
hmen
t o
f par
alle
l str
uctu
res
to m
anag
e ai
d, i
nclu
din
g
finan
cial
man
agem
ent
syst
ems
and
pro
cure
men
t sy
stem
s.
On
acco
unt
Aid
tha
t is
ap
pro
pria
ted
as
reve
nue
and
flo
ws
thro
ugh
the
trea
sury
sy
stem
is a
cco
unte
d fo
r b
y g
ove
rnm
ent
syst
ems
at m
inis
try
or
dis
tric
t le
vels
and
rep
ort
ed t
o t
he A
cco
unta
nt G
ener
al, w
here
it is
co
nso
lidat
ed in
to t
he g
ove
rnm
ent
acco
unts
(Cha
nnel
1).
Aid
tha
t is
ap
pro
pria
ted
as
Ap
pro
pria
tions
in A
id is
no
t ac
coun
ted
fo
r th
roug
h g
ove
rnm
ent
syst
ems
at t
he le
vel o
f min
istr
ies
(Cha
nnel
3)
. Ho
wev
er, s
ince
the
go
vern
men
t ne
eds
to re
po
rt t
o p
arlia
men
t o
n al
l ap
pro
pria
tions
, Ap
pro
pria
tions
in A
id is
cap
ture
d a
t ce
ntra
l le
vel b
y th
e A
cco
unta
nt G
ener
al (t
hro
ugh
the
Ext
erna
l Res
our
ces
Dep
artm
ent
and
the
Bud
get
Sup
plie
s D
epar
tmen
t in
the
Min
istr
y o
f Fin
ance
), an
d c
ons
olid
ated
in t
he g
ove
rnm
ent
acco
unts
ag
ains
t ap
pro
ved
exp
end
iture
.
Ther
e ar
e si
gni
fican
t p
rob
lem
s w
ith b
oth
Cha
nnel
s 1
and
2.
For
aid
tha
t is
dis
bur
sed
und
er C
hann
el 1
(tre
asur
y sy
stem
), w
eak
and
inco
mp
lete
use
of t
he n
ew IF
MIS
and
wea
knes
ses
in t
he v
ario
us s
yste
ms
esta
blis
hed
by
dep
artm
ents
to
reco
rd
com
mitm
ents
, rec
eip
ts a
nd p
aym
ents
affe
ct a
id a
s m
uch
as it
do
es
the
go
vern
men
t’s o
wn
spen
din
g. A
lso
, rep
ort
s ar
e ve
ry o
ften
late
, le
adin
g t
o re
gul
atio
ns n
ot
to d
isb
urse
min
istr
ies’
allo
catio
ns u
nles
s re
po
rts
are
rece
ived
on
time.
For
aid
tha
t is
dis
bur
sed
as
Cha
nnel
3 (p
aral
lel s
yste
ms)
, wea
knes
ses
in li
ne m
inis
try
info
rmat
ion
syst
ems
and
dis
tric
t-le
vel i
nfo
rmat
ion
syst
ems
cont
ribut
e to
po
or
rep
ort
ing
by
par
alle
l str
uctu
res
on
aid
d
isb
urse
men
t an
d u
se. T
his
caus
es s
igni
fican
t p
rob
lem
s fo
r th
e M
inis
try
of F
inan
ce in
co
mp
iling
go
vern
men
t ac
coun
ts a
gai
nst
app
rove
d e
xpen
ditu
re.
Wea
k p
rofe
ssio
nal c
apac
ity fo
r fin
anci
al m
anag
emen
t in
the
p
ublic
sec
tor
cont
ribut
es t
o t
he w
eak
imp
lem
enta
tion
of fi
nanc
ial
man
agem
ent
refo
rms,
suc
h as
the
IFM
IS. H
ow
ever
, eve
n w
here
ca
pac
ity is
str
ong
er, s
uch
as in
the
Min
istr
y o
f Fin
ance
, the
old
cu
sto
m-d
evel
op
ed s
yste
m fo
r re
cord
ing
pay
men
ts a
nd a
cco
untin
g
for
exp
end
iture
s in
ord
er t
o p
rod
uce
rep
ort
s to
the
Acc
oun
tant
G
ener
al is
stil
l pre
ferr
ed o
ver
the
IFM
IS. T
his
po
ints
to
wea
knes
ses
in t
he s
yste
m’s
dev
elo
pm
ent
and
/or
imp
lem
enta
tion
pro
gra
mm
es.
Wea
k fin
anci
al m
anag
emen
t d
isci
plin
e in
tur
n co
ntrib
utes
to
the
tim
elin
ess
of r
epo
rts.
Wea
k re
po
rtin
g o
n A
pp
rop
riatio
ns in
Aid
exp
end
iture
is re
late
d
to p
aral
lel i
mp
lem
enta
tion
units
and
/or
do
nors
no
t ad
herin
g t
o
go
vern
men
t d
ead
lines
for
rep
ort
ing
on
exp
end
iture
, and
wea
k ca
pac
ity (b
oth
hum
an re
sour
ce a
nd in
form
atio
n sy
stem
s) in
m
inis
trie
s to
ens
ure
rep
ort
s ar
e p
rovi
ded
on
time
for
incl
usio
n in
m
inis
trie
s’ re
po
rts
to t
he A
cco
unta
nt G
ener
al.
75
Ke
nyA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/w
hy n
ot?
)
On
aud
itTh
e O
EC
D D
AC
rep
ort
sho
ws
that
fund
s th
at a
re d
isb
urse
d t
hro
ugh
go
vern
men
t sy
stem
s ar
e g
ener
ally
aud
ited
by
the
Aud
itor
Gen
eral
. A
n ex
cep
tion
is G
AV
I: th
oug
h G
AV
I’s s
upp
ort
is n
ot
dis
bur
sed
th
roug
h g
ove
rnm
ent
syst
ems,
it is
rep
ort
ed t
o b
e au
dite
d b
y th
e A
udito
r G
ener
al. T
he m
and
ate
of t
he A
udito
r G
ener
al re
qui
res
that
all
do
nor
fund
exp
end
iture
be
aud
ited
by
the
Aud
itor
Gen
eral
. Fun
ds
that
are
dis
bur
sed
thr
oug
h C
hann
el 3
are
aud
ited
in
dep
end
ently
.
Aud
its o
f do
nor
exp
end
iture
und
erta
ken
by
the
Ken
ya N
atio
nal
Aud
it O
ffice
, sim
ilar
to a
udits
of d
om
estic
reve
nue
exp
end
iture
, are
, o
n av
erag
e, 1
3 m
ont
hs la
te (P
EFA
200
6). T
he a
udits
are
als
o li
mite
d
to t
rans
actio
n au
diti
ng a
nd d
o n
ot
in a
ll re
spec
ts c
om
ply
with
in
tern
atio
nal g
oo
d p
ract
ice.
In a
dd
itio
n, p
arlia
men
tary
follo
w-u
p o
n au
dits
is w
eak.
Cap
acity
in t
he N
atio
nal A
udit
Offi
ce is
no
t en
oug
h to
mee
t th
e A
udito
r G
ener
al’s
man
dat
e. B
ott
lene
cks
ham
per
effo
rts
to re
duc
e th
e b
ackl
og
. By
law
, aud
it re
po
rts
mus
t b
e fil
ed s
ix t
o s
even
mo
nths
af
ter
the
end
of t
he fi
nanc
ial y
ear,
whi
ch is
no
t en
oug
h tim
e, g
iven
ca
pac
ity.
On
rep
ort
Aid
is c
aptu
red
in g
ove
rnm
ent
finan
cial
sta
tem
ents
inso
far
as it
was
ca
ptu
red
on
acco
unt.
In b
udg
et d
ocu
men
tatio
n, a
id is
rep
ort
ed
on
for
the
year
bef
ore
the
cur
rent
yea
r an
d t
he c
urre
nt y
ear
itsel
f (i.
e. t
wo
yea
rs b
efo
re t
he b
udg
et y
ear)
. Aid
inflo
ws
and
pro
gre
ss
agai
nst
dev
elo
pm
ent
pro
ject
s ar
e re
po
rted
(with
var
ying
co
vera
ge)
in
the
Min
iste
rial P
ublic
Exp
end
iture
Rev
iew
s an
d t
he S
WG
s. F
or
sect
ors
tha
t ha
ve S
WA
ps,
qua
rter
ly re
po
rts
agai
nst
the
op
erat
iona
l p
lans
incl
ude
rep
ort
ing
on
aid
. In
the
Qua
rter
ly B
udg
et R
evie
ws,
ai
d re
venu
es a
re re
po
rted
on
agg
reg
ate
acro
ss g
ove
rnm
ent,
d
isag
gre
gat
ed b
y ai
d m
od
ality
. The
re is
als
o a
tab
le t
hat
spec
ifies
sp
end
ing
und
er t
he d
evel
op
men
t b
udg
et b
y m
inis
try
for
the
year
to
dat
e.
In a
dd
itio
n, a
id a
nd p
rog
ress
ag
ains
t d
evel
op
men
t b
udg
et
spen
din
g is
als
o re
po
rted
by
min
istr
ies
to t
he M
inis
try
of P
lann
ing
an
d D
evel
op
men
t w
ithin
the
Mo
nito
ring
and
Eva
luat
ion
Syst
em.
Whi
le a
id d
isb
urse
men
ts a
nd s
pen
din
g a
re re
po
rted
on
thro
ugh
vario
us m
eans
, cur
rent
ly t
here
is n
o s
our
ce fo
r a
com
pre
hens
ive
pic
ture
– fo
r al
l aid
mo
dal
ities
, by
do
nor,
by
min
istr
y, a
nd fo
r al
l re
leva
nt c
hann
els
of d
isb
urse
men
t.
Ther
e is
als
o a
lack
of n
on-
finan
cial
info
rmat
ion
in re
po
rtin
g o
n th
e d
isb
urse
men
t an
d u
tilis
atio
n o
f aid
. Ap
art
fro
m t
he M
inis
teria
l Pub
lic
Exp
end
iture
Rev
iew
s, S
WG
rep
ort
s an
d S
WA
p re
po
rts,
the
re a
re
very
few
exp
lana
tory
no
tes
acco
mp
anyi
ng re
po
rtin
g o
n ai
d. A
nd t
he
form
er t
wo
inst
rum
ents
are
bo
th w
eak
on
this
. The
imp
lem
enta
tion
of t
he K
enya
Jo
int
Ass
ista
nce
Stra
teg
y, w
ith a
nnua
l rep
ort
s, a
nd t
he
Ken
ya E
xter
nal R
eso
urce
s Po
licy,
with
co
nso
lidat
ed re
po
rtin
g, w
ill
be
help
ful.
The
frag
men
tatio
n o
f res
po
nsib
ilitie
s fo
r p
lann
ing
bet
wee
n th
e M
inis
try
of P
lann
ing
and
Dev
elo
pm
ent
and
the
Min
istr
y o
f Fin
ance
ha
s co
ntrib
uted
to
frag
men
ted
(and
dup
licat
ed) r
epo
rtin
g. W
eak
PFM
and
aid
info
rmat
ion
syst
ems
at m
inis
try
and
cen
tral
leve
ls
also
co
ntrib
ute
to in
com
ple
te re
po
rtin
g o
n ai
d. D
iffer
ent
rep
ort
ing
in
stru
men
ts h
ave
so fa
r fo
cuse
d o
n ai
d in
form
atio
n fo
r d
iffer
ent
reas
ons
, no
ne o
f whi
ch w
as e
xplic
itly
to m
anag
e ai
d b
ette
r an
d
mo
re t
rans
par
ently
. It
is o
nly
now
with
the
Ken
ya J
oin
t A
ssis
tanc
e St
rate
gy
and
the
Ken
ya E
xter
nal R
eso
urce
s Po
licy
that
the
re is
p
rog
ress
on
this
.
Ana
lytic
al c
apac
ity c
ons
trai
nts
affe
ct t
he u
sefu
lnes
s o
f rep
ort
ing
on
aid
(and
go
vern
men
t d
om
estic
sp
end
ing
).
76
sou
tH A
Fric
A
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/w
hy n
ot?
)
On
pla
nTh
ere
is li
ttle
evi
den
ce t
hat
aid
is c
aptu
red
on
wha
t w
oul
d b
e co
nsid
ered
pur
e p
lann
ing
inst
rum
ents
. The
mai
n p
lann
ing
in
stru
men
t in
the
pub
lic s
ecto
r is
the
stra
teg
ic p
lan.
But
the
stra
teg
ic
pla
ns o
f man
y na
tiona
l and
pro
vinc
ial d
epar
tmen
ts t
hat
rece
ive
aid
d
o n
ot
refle
ct a
ny re
fere
nces
to
the
aid
rece
ived
, alth
oug
h so
me
do
, su
ch a
s th
e D
epar
tmen
t o
f Env
ironm
enta
l Affa
irs a
nd T
our
ism
.
Ano
ther
imp
ort
ant
inst
rum
ent
is t
he a
nnua
l bud
get
sub
mis
sio
n,
alth
oug
h th
is is
a c
om
bin
ed p
lann
ing
and
bud
get
ing
to
ol.
At
natio
nal a
nd p
rovi
ncia
l lev
els,
the
sub
mis
sio
n re
qui
res
a ta
ble
tha
t d
etai
ls O
DA
rece
ived
. Ho
wev
er, i
n th
e na
rrat
ives
and
pro
gra
mm
e p
lans
no
refe
renc
e is
mad
e to
aid
.
The
SWA
p in
the
nat
iona
l Dep
artm
ent
of W
ater
Affa
irs a
nd F
ore
stry
d
oes
hav
e a
sep
arat
e st
rate
gic
do
cum
ent
for
its M
asib
amb
ane
pro
gra
mm
e, w
hich
is a
join
t p
lan
for
exte
rnal
and
do
mes
tic re
venu
e.
Ho
wev
er, t
he 2
006/
07 s
trat
egic
pla
n d
oes
no
t m
ake
any
refe
renc
e to
the
pro
gra
mm
e, a
ltho
ugh
the
2006
/07
stra
teg
ic p
lan
and
the
M
asib
amb
ane
do
cum
ents
sha
re t
he s
ame
stra
teg
ic fr
amew
ork
. N
ever
thel
ess,
Mas
ibam
ban
e p
rovi
des
an
exam
ple
of a
dep
artm
ent
that
has
a s
trat
egic
fram
ewo
rk fo
r O
DA
.
Cap
turin
g e
xter
nal f
und
ing
in b
udg
et s
ubm
issi
ons
mee
ts a
m
inim
um re
qui
rem
ent
for
put
ting
aid
on
pla
n, b
ut t
he q
ualit
y an
d
com
ple
tene
ss o
f the
info
rmat
ion
is d
oub
tful
, and
it is
no
t in
teg
rate
d
with
pla
nnin
g fo
r d
om
estic
reve
nue.
The
inte
gra
tion
of e
xter
nal a
nd d
om
estic
fina
ncin
g in
the
M
asib
amb
ane
pro
gra
mm
e se
ts g
oo
d p
ract
ice
stan
dar
ds
for
sect
ors
th
at re
ceiv
e si
gni
fican
t ex
tern
al fu
ndin
g.
On
the
surf
ace,
the
Tre
asur
y R
egul
atio
ns –
whi
ch fr
ame
the
form
at,
cont
ent
and
pro
cess
for
stra
teg
ic p
lans
and
pla
nnin
g –
do
no
t re
qui
re e
xter
nal f
und
ing
to
be
inte
gra
ted
at
the
pla
nnin
g p
hase
.
The
mo
re in
-dep
th e
xpla
natio
n fo
r th
e p
oo
r in
teg
ratio
n o
f aid
on
pla
n ha
s to
do
with
the
mis
conc
eptio
n o
f the
sta
tus
of O
DA
in t
he
Sout
h A
fric
an c
ont
ext.
The
po
licy
– as
exp
ress
ed b
y th
e p
rovi
sio
ns
for
aid
man
agem
ent
in t
he P
ublic
Fin
ance
Man
agem
ent
Act
– is
tha
t ai
d, e
ven
whe
n d
isb
urse
d t
o t
he g
ove
rnm
ent,
is e
xtra
-bud
get
ary
and
to
o u
npre
dic
tab
le (a
nd in
sig
nific
ant?
) to
be
incl
uded
in t
he
MTE
F.
The
aid
man
agem
ent
syst
em is
wel
l co
ord
inat
ed b
y th
e g
ove
rnm
ent
and
am
ong
do
nors
, and
mo
nito
ring
info
rmat
ion
and
str
ateg
ic
ove
rsig
ht a
re c
entr
alis
ed, b
ut it
run
s la
rgel
y se
par
atel
y fr
om
the
co
re p
lann
ing
and
bud
get
ing
sys
tem
. It
is, h
ow
ever
, alig
ned
to
the
go
vern
men
t’s P
rog
ram
me
of A
ctio
n an
d d
epar
tmen
tal
prio
ritie
s. M
ost
of t
he li
nkag
es a
re fo
r us
ing
info
rmat
ion
fro
m t
he
aid
man
agem
ent
syst
em t
o in
form
aid
cho
ices
, rat
her
than
the
re
cip
roca
l use
of a
id in
form
atio
n to
info
rm b
udg
etin
g c
hoic
es. T
his
is t
rue
at n
atio
nal c
oo
rdin
atio
n an
d d
epar
tmen
tal s
trat
egic
leve
ls:
do
nors
co
nsul
t st
rate
gic
pla
ns t
o in
form
the
ir sp
end
ing
cho
ices
, ra
ther
tha
n m
uch
of t
he a
id a
ctiv
ity b
eing
inte
gra
ted
into
the
pla
ns
and
bud
get
s th
emse
lves
.
Ano
ther
imp
edim
ent
is t
he la
ck o
f str
ateg
ic p
lann
ing
cap
acity
in
man
y d
epar
tmen
ts. I
t is
als
o d
oub
tful
tha
t d
epar
tmen
ts d
iscl
ose
full
info
rmat
ion
on
the
sup
po
rt t
hey
rece
ive,
fear
ing
tha
t it
may
affe
ct
thei
r al
loca
tions
neg
ativ
ely,
or
that
the
y w
oul
d h
ave
less
dis
cret
ion
ove
r th
e us
e o
f the
fund
s. D
ono
rs b
enefi
t so
mew
hat
fro
m t
his
wea
k in
teg
ratio
n: t
hey
are
muc
h b
ette
r ab
le t
o s
et t
heir
ow
n ag
end
as,
or
cho
ose
fro
m s
trat
egic
fram
ewo
rks
tho
se a
ctiv
ities
whi
ch a
lign
with
the
ir p
riorit
ies.
The
y ar
e no
t p
ushe
d t
o c
ons
ider
dep
artm
enta
l p
riorit
ies.
On
the
oth
er h
and
, dep
artm
ents
, par
ticul
arly
tho
se w
ho
rece
ive
sig
nific
ant
aid
(eith
er in
ter
ms
of v
olu
me
or
add
ed v
alue
), ar
e no
t p
ushe
d t
o c
ons
ider
the
ir st
rate
gie
s an
d b
udg
etin
g in
re
latio
n to
aid
.
77
sou
tH A
Fric
A
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/w
hy n
ot?
)
On
bud
get
Ther
e is
no
refe
renc
e to
aid
at
natio
nal o
r p
rovi
ncia
l lev
els,
oth
er
than
info
rmat
ion
on
pas
t, c
urre
nt a
nd e
xpec
ted
forw
ard
flo
ws
into
th
e R
eco
nstr
uctio
n an
d D
evel
op
men
t Fu
nd (e
stab
lishe
d a
t na
tiona
l le
vel t
o re
ceiv
e ai
d fl
ow
s) in
ag
gre
gat
e in
so
me
tab
les
in t
he B
udg
et
Rev
iew
. By
div
idin
g t
his
num
ber
by
the
aid
dis
bur
sed
to
cen
tral
g
ove
rnm
ent,
the
200
6 Su
rvey
on
Mo
nito
ring
the
Par
is D
ecla
ratio
n d
eter
min
ed In
dic
ato
r 3.
The
cap
ture
is n
ot
effe
ctiv
e. It
ag
gre
gat
es a
ll flo
ws
into
the
R
eco
nstr
uctio
n an
d D
evel
op
men
t Fu
nd in
to o
ne n
umb
er, w
hich
, fu
rthe
r, d
iffer
s fr
om
the
fund
’s au
dite
d o
utco
mes
pub
lishe
d in
th
e b
udg
et in
form
atio
n. T
he d
iffer
ence
can
onl
y b
e ex
pla
ined
by
trac
ing
the
his
tory
of a
id re
po
rtin
g b
ack
to t
he 2
003
Bud
get
Rev
iew
. Th
ere
a se
par
ate
tab
le s
how
s th
at t
he n
umb
er fo
r fo
reig
n g
rant
s an
d a
ssis
tanc
e un
der
the
Rec
ons
truc
tion
and
Dev
elo
pm
ent
Fund
co
nsis
ts o
f the
fina
ncia
l flo
ws
(eq
ual t
o t
he a
udite
d n
umb
er) p
lus
tech
nica
l co
op
erat
ion
man
aged
by
do
nors
. Thi
s is
als
o t
he o
nly
bud
get
revi
ew t
hat
offe
rs a
brie
f sum
mar
y o
n ai
d fl
ow
s, in
clud
ing
m
entio
ning
a fe
w h
igh
pro
file
pro
ject
s.
The
cap
ture
pro
vid
es n
o e
ffect
ive
info
rmat
ion,
e.g
. who
will
be
usin
g t
he a
id, w
hat
for,
with
wha
t co
nseq
uenc
es, o
r w
hat
kind
of a
id
it is
.
The
2008
MTE
F g
uid
elin
es s
et o
ut t
he e
xpec
ted
issu
es a
nd
info
rmat
ion
req
uire
d fo
r th
e 20
08 b
udg
et s
ubm
issi
ons
, and
for
the
first
tim
e m
entio
n th
at t
he in
teg
ratio
n o
f OD
A in
to p
lans
and
b
udg
ets
is a
n is
sue.
Aid
form
s su
ch a
sm
all p
rop
ort
ion
of t
he b
udg
et (l
ess
than
1%
) th
at it
do
es n
ot
feat
ure
sig
nific
antly
in c
entr
al a
gen
cy p
roce
sses
, in
clud
ing
the
bud
get
offi
ce a
nd p
arlia
men
t. C
entr
al a
gen
cy
pro
cess
es a
re c
ong
este
d a
lread
y, a
nd h
ave
to m
anag
e, a
bso
rb a
nd
use
larg
e am
oun
ts o
f inf
orm
atio
n: a
id is
eas
y to
igno
re. A
ltho
ugh
aid
info
rmat
ion
is p
rovi
ded
in t
he b
udg
et s
ubm
issi
ons
, it
do
es n
ot
get
pub
lishe
d b
ecau
se t
he q
ualit
y is
po
or
and
the
co
vera
ge
too
w
eak
to b
e m
eani
ngfu
l.
Ho
wev
er, a
id c
an b
e a
cons
ider
able
pro
po
rtio
n o
f a d
epar
tmen
t’s
dis
cret
iona
ry fu
nds,
if (i
) so
cial
sec
urity
sp
end
ing
and
oth
er m
ore
rig
id it
ems
are
take
n o
ut o
f the
bud
get
; and
(ii)
it is
tak
en in
to
acco
unt
that
aid
is n
ot
spre
ad a
cro
ss a
ll d
epar
tmen
ts a
nd p
rovi
nces
b
ut fo
cuse
d in
a fe
w. I
t th
en b
eco
mes
imp
ort
ant
for
the
dep
artm
ent
and
the
rele
vant
tre
asur
y an
d le
gis
latu
re t
o g
et a
go
od
pic
ture
of
the
aid
tha
t is
bei
ng d
isb
urse
d a
nd w
hat
for.
Alth
oug
h b
oth
the
Inte
rnat
iona
l Dev
elo
pm
ent
Co
ord
inat
ion
Uni
t an
d t
he P
ublic
Fin
ance
and
Bud
get
Co
ord
inat
ion
units
at
the
Nat
iona
l Tre
asur
y ha
ve p
roce
sses
and
mec
hani
sms
in p
lace
for
thei
r in
div
idua
l pur
po
ses,
the
re is
ver
y lit
tle c
om
mun
icat
ion
bet
wee
n th
ese
units
– b
oth
a s
ymp
tom
and
a c
ont
ribut
ing
cau
se o
f the
se
par
atio
n o
f pla
nnin
g a
nd b
udg
etin
g fo
r ai
d fl
ow
s fr
om
pla
nnin
g
and
bud
get
ing
for
do
mes
tic re
venu
e. W
hile
the
Inte
rnat
iona
l D
evel
op
men
t C
oo
rdin
atio
n U
nit
mak
es u
se o
f the
out
put
s o
f the
o
ther
div
isio
ns (t
he M
TEF)
to
up
dat
e its
str
ateg
ic fr
amew
ork
for
aid
, the
div
isio
ns a
re n
ot
in t
urn
reci
pie
nts
of i
nfo
rmat
ion
fro
m
the
Inte
rnat
iona
l Dev
elo
pm
ent
Co
ord
inat
ion
Uni
t. T
he c
ause
o
f thi
s d
isco
nnec
t is
no
t in
stitu
tiona
l str
uctu
re: t
he In
tern
atio
nal
Dev
elo
pm
ent
Co
ord
inat
ion
Uni
t is
par
t o
f the
Bud
get
Co
ord
inat
ion
Uni
t. R
athe
r, it
lies
in t
he p
rese
nt s
tatu
s ac
cord
ed t
o a
id in
pub
lic
finan
ces.
On
par
liam
ent
Aid
is n
ot
app
rove
d a
s p
art
of t
he b
udg
et, o
r se
par
atel
y, b
y p
arlia
men
t. T
he p
resi
den
t si
gns
off
on
aid
ag
reem
ents
. Par
liam
ent
is in
form
ed e
ither
thr
oug
h th
e N
atio
nal T
reas
ury
or
the
Dep
artm
ent
of F
ore
ign
Affa
irs w
hen
fram
ewo
rk a
gre
emen
ts w
ith d
ono
rs a
re
conc
lud
ed, b
ut it
do
es n
ot
rece
ive
any
info
rmat
ion
on
ind
ivid
ual
pro
gra
mm
es o
r p
roje
cts
with
in t
he a
gre
emen
ts, u
nles
s a
com
mitt
ee
asks
for
it. N
or
do
any
of t
he p
rovi
ncia
l par
liam
ents
.
All
inte
rnat
iona
l ag
reem
ents
hav
e to
be
sub
mitt
ed t
o p
arlia
men
t fo
r no
tice
pur
po
ses.
Aid
is n
ot
app
rove
d a
s p
art
of t
he b
udg
et, o
r se
par
atel
y, b
y p
arlia
men
t.A
ll ai
d, e
ven
whe
n m
anag
ed t
hro
ugh
go
vern
men
t sy
stem
s,
is d
efine
d a
s ex
tra-
bud
get
ary
in t
erm
s o
f the
Pub
lic F
inan
ce
Man
agem
ent
Act
. The
re is
no
GB
S.
78
sou
tH A
Fric
A
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/w
hy n
ot?
)
On
trea
sury
Aid
in S
out
h A
fric
a flo
ws
thro
ugh
two
cha
nnel
s. O
DA
in t
he fo
rm o
f g
rant
s o
r TA
is re
gar
ded
as
a d
ona
tion
to t
he g
ove
rnm
ent.
In p
rinci
ple
, all
gra
nts
have
to
be
pai
d in
to t
he R
eco
nstr
uctio
n an
d D
evel
op
men
t Fu
nd (C
hann
el 1
). In
pra
ctic
e, t
his
is t
he c
ase
for
gra
nts
for
whi
ch t
he g
ove
rnm
ent
take
s re
spo
nsib
ility
for
the
finan
cial
man
agem
ent.
The
acc
oun
ting
offi
cer
(usu
ally
the
dire
cto
r g
ener
al) h
as t
o a
pp
rove
and
is a
cco
unta
ble
for
the
gra
nt, a
nd t
he
chie
f fina
ncia
l offi
cer
(a p
osi
tion
crea
ted
in t
erm
s o
f the
Tre
asur
y R
egul
atio
ns) h
as t
o k
eep
a re
gis
ter
of a
ll g
rant
s.
Mo
st d
ona
tions
tha
t ar
e no
t ca
sh g
rant
s ar
e TA
man
aged
by
the
do
nor.
Som
e ag
reem
ents
mak
e fu
nds
avai
lab
le fo
r th
e ac
tiviti
es
of t
he im
ple
men
ting
ag
ency
, but
the
se fu
nds
are
not
chan
nelle
d
thro
ugh
the
Rec
ons
truc
tion
and
Dev
elo
pm
ent
Fund
: the
do
nor
take
s re
spo
nsib
ility
for
thei
r fin
anci
al m
anag
emen
t. T
he a
cco
untin
g
offi
cer
still
has
resp
ons
ibili
ty fo
r th
e p
roje
ct it
self
– e.
g. f
or
pro
ject
ef
fect
iven
ess.
In t
hese
cas
es (C
hann
el 3
), th
e d
ona
tion
is m
anag
ed
thro
ugh
a th
ird p
arty
, suc
h as
eith
er a
PIU
, or
a tr
ust
whi
ch m
anag
es
all d
ono
r fu
nds
on
beh
alf o
f the
dep
artm
ent
(e.g
. in
the
Dep
artm
ent
of E
duc
atio
n in
Lim
po
po
), o
r b
y th
e d
ono
r ag
ency
itse
lf.
Acc
ord
ing
to
the
200
6 Su
rvey
on
Mo
nito
ring
the
Par
is D
ecla
ratio
n,
36%
of a
id d
isb
urse
d t
o g
ener
al g
ove
rnm
ent
uses
Cha
nnel
1. O
f the
th
ree
larg
est
do
nors
, DFI
D a
nd U
SAID
use
Cha
nnel
3 e
xclu
sive
ly,
whi
le t
he E
C u
ses
bo
th C
hann
el 1
and
3, w
ith 4
2% o
f its
aid
go
ing
th
roug
h C
hann
el 1
.
Do
nors
who
use
the
Rec
ons
truc
tion
and
Dev
elo
pm
ent
Fund
m
echa
nism
co
mp
lain
of l
ong
del
ays
in g
ettin
g p
roje
cts
off
the
gro
und
, co
mp
rom
isin
g t
heir
stra
teg
ic re
leva
nce.
Thi
s is
no
t ne
cess
arily
bec
ause
aid
dis
bur
sem
ents
fro
m t
he c
entr
e to
the
line
ar
e un
pre
dic
tab
le: t
he p
roce
ss g
ets
del
ayed
in t
he b
urea
ucra
cy,
oft
en a
t re
cip
ient
leve
l. If
follo
wed
co
rrec
tly, t
he p
roce
ss t
akes
ap
pro
xim
atel
y fiv
e w
ork
ing
day
s. B
ut d
evel
op
men
t p
artn
ers
do
no
t al
way
s d
epo
sit
fund
s o
n tim
e o
r no
tify
the
Inte
rnat
iona
l D
evel
op
men
t C
oo
rdin
atio
n U
nit
or
reci
pie
nt d
epar
tmen
t o
f the
tr
ansf
er, o
r th
e re
cip
ient
dep
artm
ent
do
es n
ot
req
uest
the
fund
s o
n tim
e. T
he R
eco
nstr
uctio
n an
d D
evel
op
men
t Fu
nd a
nnua
l rep
ort
an
d t
he D
evel
op
men
t C
oo
per
atio
n In
form
atio
n Sy
stem
dat
abas
e (a
t p
roje
ct le
vel)
giv
e an
ove
rvie
w o
f fun
ds
rece
ived
and
dis
bur
sed
. H
ow
ever
, it
is n
ot
easy
to
tra
ce w
hen
the
find
s w
ere
sup
po
sed
to
b
e d
isb
urse
d (u
nles
s in
div
idua
l pro
ject
ag
reem
ents
and
op
erat
iona
l p
lans
are
tra
ced
), w
hich
leav
es t
he a
cco
unta
bili
ty g
rey.
Whe
n d
ono
rs a
nd/o
r im
ple
men
tatio
n ag
enci
es c
hoo
se n
ot
to
use
the
Rec
ons
truc
tion
and
Dev
elo
pm
ent
Fund
mec
hani
sm
(whi
ch a
uto
mat
ical
ly m
eans
dis
bur
sem
ent
thro
ugh
cent
ral
trea
sury
co
ntro
lled
mec
hani
sms
and
fina
ncia
l man
agem
ent
by
the
imp
lem
entin
g a
gen
cy),
it is
for
one
or
a co
mb
inat
ion
of t
he
follo
win
g re
aso
ns: (
i) Th
ere
are
real
or
per
ceiv
ed d
elay
s as
soci
ated
w
ith t
he p
roce
dur
es t
hat
need
to
be
follo
wed
for
dis
bur
sem
ent
thro
ugh
the
Rec
ons
truc
tion
and
Dev
elo
pm
ent
Fund
. (ii)
The
re
cip
ient
dep
artm
ent’s
fina
ncia
l man
agem
ent
cap
acity
is t
oo
wea
k.
(iii)
The
do
nor
is c
ons
trai
ned
by
its o
wn
rule
s an
d re
gul
atio
ns t
o u
se
go
vern
men
t sy
stem
s. (i
v) T
he re
cip
ient
dep
artm
ent
pre
fers
a t
hird
p
arty
to
man
age
the
pro
ject
ove
rall,
bec
ause
it p
erce
ives
it t
o o
ffer
mo
re v
alue
or
per
ceiv
es it
s o
wn
ove
rall
man
agem
ent
cap
acity
to
be
too
wea
k.
Als
o, i
n p
rinci
ple
the
pro
ced
ures
for
tran
sfer
ring
do
nor
fund
s fr
om
th
e R
eco
nstr
uctio
n an
d D
evel
op
men
t Fu
nd t
o t
he im
ple
men
ting
ag
ency
, sho
uld
tak
e ab
out
a w
eek.
But
the
Rec
ons
truc
tion
and
D
evel
op
men
t Fun
d m
echa
nism
ent
ails
hav
ing
pro
ject
man
ager
s th
at
are
not
nece
ssar
ily d
edic
ated
sta
ff, b
ut m
anag
e th
e d
ono
r-fu
nded
p
roje
ct (o
r d
ono
r-fu
nded
act
iviti
es w
ithin
a la
rger
pro
gra
mm
e)
alo
ng w
ith a
noth
er p
ort
folio
of w
ork
. Thi
s in
itse
lf m
ay c
ause
d
elay
s. A
ded
icat
ed u
nit
for
man
agin
g fu
nds
com
es w
ith a
dd
itio
nal
imp
lem
enta
tion
cap
acity
.
79
sou
tH A
Fric
A
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/w
hy n
ot?
)
On
acco
unt
All
fund
s p
aid
into
the
Rec
ons
truc
tion
and
Dev
elo
pm
ent
Fund
ac
coun
t ar
e m
anag
ed t
hro
ugh
go
vern
men
t sy
stem
s, fo
llow
ing
st
and
ard
ised
acc
oun
ting
pro
ced
ures
and
acc
ord
ing
to
co
mm
on
acco
untin
g s
tand
ard
s.
Fund
s th
at a
re m
anag
ed t
hro
ugh
Cha
nnel
3, h
ow
ever
, are
no
t o
n ac
coun
t.
Fund
s th
at a
re n
ot
dis
bur
sed
thr
oug
h C
hann
el 1
, are
acc
oun
ted
for
by
third
par
ty o
r d
ono
r sy
stem
s.
Bec
ause
Ger
man
Tec
hnic
al C
oo
per
atio
n re
po
rted
ly d
oes
no
t us
e C
hann
el 1
for
dis
bur
sem
ent
and
exe
cutio
n, b
ut d
oes
for
finan
cial
re
po
rtin
g (w
hich
wo
uld
imp
ly u
sing
go
vern
men
t’s a
cco
untin
g
syst
ems)
, the
per
cent
age
is h
ighe
r fo
r o
n ac
coun
t th
an fo
r o
n tr
easu
ry, i
.e. 4
1%.
Gen
eral
ly d
ono
rs w
ill u
se g
ove
rnm
ent
acco
untin
g s
yste
ms
in
dep
artm
ents
tha
t ha
ve t
he c
apac
ity fo
r th
e fin
anci
al m
anag
emen
t o
f the
fund
s. U
sing
go
vern
men
t sy
stem
s is
effe
ctiv
e in
sofa
r as
it
pro
vid
es t
imel
y an
d c
red
ible
info
rmat
ion
on
spen
din
g. H
ow
ever
, no
t us
ing
go
vern
men
t sy
stem
s (o
r th
e R
eco
nstr
uctio
n an
d D
evel
op
men
t Fu
nd c
hann
el) b
ecau
se o
f per
ceiv
ed w
eakn
esse
s in
PFM
cap
acity
, ha
s th
e sa
me
dra
wb
acks
as
else
whe
re. F
irst,
par
alle
l del
iver
y sy
stem
s th
at p
rovi
de
serv
ices
in t
he s
hort
ter
m, b
ut d
o n
ot
offe
r su
stai
nab
le s
olu
tions
. And
sec
ond
, whe
ther
fund
s ar
e o
n ac
coun
t o
r no
t (b
ut m
ore
for
tho
se w
hich
are
no
t o
n ac
coun
t), fi
nanc
ial r
epo
rts
on
how
fund
s w
ere
used
rar
ely
reac
h d
om
estic
sta
keho
lder
s b
eyo
nd
the
agen
cy c
onc
erne
d a
nd p
erha
ps
the
Inte
rnat
iona
l Dev
elo
pm
ent
Co
ord
inat
ion
Uni
t, c
om
pro
mis
ing
acc
oun
tab
ility
.
Whe
ther
fund
s ar
e o
n ac
coun
t o
r no
t is
a fu
nctio
n o
f whe
ther
C
hann
el 1
or
Cha
nnel
3 is
use
d fo
r d
isb
urse
men
t.
Perh
aps
the
que
stio
n is
no
t w
hy p
roje
cts
are
not
on
acco
unt,
but
w
hy b
eing
on
acco
unt
or
not
mak
es t
hem
effe
ctiv
e o
r in
effe
ctiv
e?
In t
he S
out
h A
fric
an c
ont
ext,
arg
uab
ly, a
pro
ject
’s ef
fect
iven
ess
is
less
to
do
with
imm
edia
te o
r d
irect
out
put
s an
d o
utco
mes
tha
n w
ith w
heth
er it
bui
lds
the
cap
acity
of t
he g
ove
rnm
ent
to u
tilis
e its
o
wn
do
mes
tic re
venu
e ef
fect
ivel
y. A
car
eful
jud
gem
ent
mus
t b
e m
ade
abo
ut t
he t
rad
e-o
ffs b
etw
een
usin
g g
ove
rnm
ent
syst
ems
or
alte
rnat
ive
PFM
sys
tem
s. M
echa
nism
s su
ch a
s PI
Us
wea
ken
long
-te
rm c
apac
ity b
uild
ing
for
PFM
in c
oun
trie
s th
at a
re a
id d
epen
den
t an
d h
ave
wea
k in
stitu
tions
. In
Sout
h A
fric
a, t
he is
sue
is n
ot
so m
uch
PFM
sys
tem
s b
ut s
ervi
ce d
eliv
ery
syst
ems.
If a
pro
ject
is m
anag
ed
thro
ugh
a PI
U b
ut d
eliv
ers
serv
ices
inno
vativ
ely
and
bui
lds
del
iver
y ca
pac
ity in
a g
ove
rnm
ent
dep
artm
ent,
pro
vid
es g
oo
d p
roje
ct
and
fina
ncia
l man
agem
ent
and
mo
re t
imel
y d
isb
urse
men
t an
d
imp
lem
enta
tion
mec
hani
sms,
wo
uld
it n
ot
be
ben
efici
al t
o u
se
a PI
U?
The
issu
e is
per
hap
s no
t ho
w p
roje
cts
and
fund
flo
ws
are
man
aged
, but
ho
w P
IUs
go
ab
out
the
ir su
bst
antiv
e b
usin
ess:
in
par
alle
l or
wo
rkin
g t
hro
ugh
dep
artm
enta
l sta
ff an
d in
stitu
tions
.
This
dem
ons
trat
es w
hy t
he P
aris
Dec
lara
tion
prin
cip
les
sho
uld
no
t b
e ap
plie
d b
lind
ly t
o S
out
h A
fric
a. T
he v
alue
s an
d in
tent
beh
ind
th
e d
irect
ives
are
mo
re im
po
rtan
t th
an p
olic
y p
resc
riptio
ns. T
he re
al
que
stio
ns a
re: W
hat
add
ed v
alue
do
do
nors
brin
g?
Wha
t is
the
mo
st
effe
ctiv
e w
ay t
o e
xtra
ct t
hat
valu
e ef
ficie
ntly
?
On
aud
itTh
e us
e o
f fun
ds
that
are
dis
bur
sed
thr
oug
h C
hann
el 1
mus
t b
y la
w (t
he R
eco
nstr
uctio
n an
d D
evel
op
men
t Fu
nd A
ct) b
e au
dite
d b
y th
e A
udito
r G
ener
al. T
his
incl
udes
bo
th t
he fi
nanc
ial r
eco
rds
and
st
atem
ents
of t
he R
eco
nstr
uctio
n an
d D
evel
op
men
t Fu
nd, a
nd t
he
finan
cial
reco
rds
and
sta
tem
ents
of t
he s
pen
din
g a
gen
cy t
o w
hich
fu
nds
have
bee
n d
isb
urse
d.
Cha
nnel
3 fu
nds
may
be
aud
ited
by
the
Aud
itor
Gen
eral
in t
erm
s o
f the
Aud
itor
Gen
eral
’s m
and
ate,
but
can
als
o b
e au
dite
d b
y th
ird
par
ty a
udito
rs o
r th
e d
ono
r’s a
udito
rs. T
his
is d
eter
min
ed b
y th
e fu
ndin
g a
gre
emen
t re
gul
atin
g t
he d
ono
r’s a
ctiv
ities
in S
out
h A
fric
a,
and
ind
ivid
ual p
roje
ct a
gre
emen
ts.
All
Cha
nnel
1 fu
nds
are
aud
ited
as
req
uire
d. T
he 2
006
Surv
ey o
n M
oni
torin
g t
he P
aris
Dec
lara
tion
sug
ges
ts t
hat
som
e C
hann
el 3
fu
nds
are
also
aud
ited
: 44%
of f
und
s d
isb
urse
d a
s O
DA
to
gen
eral
g
ove
rnm
ent
are
aud
ited
thr
oug
h co
untr
y au
dit
syst
ems.
Whe
ther
fund
s ar
e o
n au
dit
is la
rgel
y a
func
tion
of w
heth
er C
hann
el
1 o
r C
hann
el 3
is s
elec
ted
, and
whe
ther
, if C
hann
el 3
, the
do
nor
or
dep
artm
ent
cho
ose
s to
be
aud
ited
by
the
Aud
itor
Gen
eral
or
not.
Dep
artm
ents
and
do
nors
may
cho
ose
to
hav
e fu
nds
aud
ited
by
the
Aud
itor
Gen
eral
bec
ause
of t
he p
oss
ible
eco
nom
ies
of s
cale
if
the
aud
it is
do
ne b
y a
cred
ible
aud
itor
alre
ady
fam
iliar
with
d
epar
tmen
tal o
rgan
isat
ion
and
wea
knes
ses.
80
sou
tH A
Fric
A
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/w
hy n
ot?
)
On
rep
ort
The
Rec
ons
truc
tion
and
Dev
elo
pm
ent
Fund
Act
and
the
Tre
asur
y R
egul
atio
ns re
qui
re t
hat
fore
ign
assi
stan
ce re
ceiv
ed in
cas
h an
d
in k
ind
is re
po
rted
in d
epar
tmen
ts’ a
nnua
l rep
ort
s, w
heth
er it
is
rece
ived
thr
oug
h th
e R
eco
nstr
uctio
n an
d D
evel
op
men
t Fu
nd o
r no
t. R
epo
rts
refe
r to
fore
ign
assi
stan
ce re
ceiv
ed, a
nd t
he a
ttac
hed
an
nual
fina
ncia
l sta
tem
ents
reco
rd t
he a
ssis
tanc
e se
par
atel
y at
ag
gre
gat
e an
d p
roje
ct le
vels
. Co
nsul
tanc
ies
are
rep
ort
ed s
epar
atel
y,
no m
atte
r w
hat
the
sour
ce o
f the
fund
ing
.
The
Rec
ons
truc
tion
and
Dev
elo
pm
ent
Fund
als
o p
ublis
hes
an
annu
al re
po
rt, w
hich
incl
udes
an
ove
rvie
w o
f the
mai
n re
cip
ient
s o
f aid
and
wha
t it
is fo
r, an
d a
det
aile
d s
tate
men
t o
n tr
ansf
ers
to
reci
pie
nt a
gen
cies
. The
fund
is n
ot
ob
ligat
ed t
o re
po
rt o
n ho
w
tho
se t
rans
fers
wer
e us
ed b
y re
cip
ient
ag
enci
es: b
y la
w t
hat
is t
he
ob
ligat
ion
of t
he a
cco
untin
g o
ffice
rs o
f the
ag
enci
es.
Fina
ncia
l and
per
form
ance
rep
ort
s ar
e p
rovi
ded
to
do
nors
, eith
er b
y d
epar
tmen
ts o
r th
ird p
arty
imp
lem
entin
g a
gen
ts.
Sub
stan
tial p
roje
ct o
r p
rog
ram
me
rep
ort
ing
is n
ot
easi
ly a
cces
sib
le
for
stak
eho
lder
s: t
he o
nly
evid
ence
of t
his
is t
he M
asib
amb
ane
rep
ort
s (w
ater
sec
tor)
. Whi
le t
he In
tern
atio
nal D
evel
op
men
t C
oo
rdin
atio
n U
nit
is g
ener
ally
par
t o
f the
rep
ort
ing
pro
cess
, the
re
po
rts
are
not
com
mo
nly
avai
lab
le.
Dep
artm
ents
’ ann
ual fi
nanc
ial s
tate
men
ts p
rovi
de
a hi
gh-
leve
l co
mp
rehe
nsiv
e o
verv
iew
of O
DA
rece
ived
. Ho
wev
er, a
nnua
l rep
ort
s th
emse
lves
do
no
t re
po
rt o
n th
e ef
fect
iven
ess
and
effi
cien
cy o
f aid
sp
end
ing
, as
they
do
for
do
mes
tic re
venu
e, h
ow
ever
inco
mp
lete
ly.
No
ne o
f the
ann
ual r
epo
rts
scru
tinis
ed c
om
plie
d in
full
with
Tre
asur
y R
egul
atio
ns re
qui
rem
ents
, and
are
thu
s no
t as
effe
ctiv
e as
the
y m
ight
be.
Par
liam
enta
ry c
om
mitt
ees
scru
tinis
e an
nual
rep
ort
s.
Tran
spar
ency
is la
ckin
g a
roun
d t
he re
view
and
eva
luat
ion
rep
ort
s th
at a
re p
rep
ared
for
do
nors
(in
prin
cip
le fo
r b
oth
do
nors
and
the
g
ove
rnm
ent)
. The
se a
re n
ot
com
mo
nly
pub
licly
ava
ilab
le. H
ow
ever
, th
e In
tern
atio
nal D
evel
op
men
t C
oo
rdin
atio
n U
nit
has
laun
ched
a
stud
y to
est
ablis
h a
dat
abas
e w
hich
will
co
ntai
n al
l rev
iew
s,
eval
uatio
ns a
nd g
oo
d p
ract
ices
.
The
annu
al re
po
rt a
nd fi
nanc
ial s
tate
men
t fo
rmat
s ar
e p
resc
ribed
b
y th
e Tr
easu
ry R
egul
atio
ns. T
hat
dep
artm
ents
– e
ven
the
Nat
iona
l Tr
easu
ry –
und
er-r
epo
rt a
gai
nst
thes
e fo
rmal
req
uire
men
ts is
ac
cep
ted
pra
ctic
e. P
arlia
men
tary
co
mm
ittee
s d
o n
ot
dem
and
fu
ller
rep
ort
ing
. Oth
er in
stitu
tions
suc
h as
the
Inte
rnat
iona
l D
evel
op
men
t C
oo
rdin
atio
n U
nit
coul
d a
lso
pla
y a
role
, as
it is
w
ithin
the
resp
ons
ibili
ties
of t
he N
atio
nal T
reas
ury
to e
nfo
rce
the
Pub
lic F
inan
ce M
anag
emen
t A
ct a
nd t
he T
reas
ury
Reg
ulat
ions
. E
nfo
rcin
g t
hem
mo
re fu
lly is
par
ticul
arly
imp
ort
ant
for
dep
artm
ents
an
d p
rovi
nces
whe
re fo
reig
n ai
d is
mo
re s
igni
fican
t, e
ither
as
a p
erce
ntag
e o
f to
tal b
udg
et o
r in
ter
ms
of i
ts a
dd
ed v
alue
.
Rea
sons
for
und
er-r
epo
rtin
g o
n ai
d:
Som
e d
epar
tmen
ts a
rgue
tha
t fu
nds
that
do
no
t co
me
thro
ugh
the
Rec
ons
truc
tion
and
Dev
elo
pm
ent
Fund
are
no
t th
eir
finan
cial
re
spo
nsib
ility
and
the
refo
re t
hey
do
no
t re
po
rt o
n th
em.
Dep
artm
ents
do
no
t ne
cess
arily
hav
e th
e co
rrec
t am
oun
t to
rep
ort
if
the
aid
is in
kin
d, a
s th
ey h
ave
to g
et t
he in
form
atio
n fr
om
the
d
evel
op
men
t p
artn
er.
Unl
ess
the
fund
s flo
w t
hro
ugh
the
go
vern
men
t sy
stem
, the
y ar
e us
ually
rep
ort
ed in
the
cur
renc
y o
f the
do
nor.
Dep
artm
ents
the
n ha
ve d
ifficu
lty in
rep
ort
ing
in r
and
as
they
are
no
t su
re w
hich
ex
chan
ge
rate
to
use
.
81
tAn
ZA
niA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/w
hy n
ot?
)
On
pla
nM
kuku
ta: A
id in
flow
s ar
e in
clud
ed in
the
fina
ncin
g fr
amew
ork
for
this
to
p-le
vel s
trat
egic
pla
n, T
anza
nia’
s PR
SP, w
ith a
bre
akd
ow
n b
etw
een
loan
s an
d g
rant
s an
d a
id m
od
aliti
es. T
he ro
le o
f do
nors
is d
iscu
ssed
as
wel
l.
Sect
or s
trat
egic
pla
ns: S
imila
rly, a
id is
incl
uded
in t
he fi
nanc
ing
fr
amew
ork
for
sect
or
stra
teg
ic p
lans
.
Cro
ss-s
ecto
r MTE
F: T
he c
ross
-sec
tor
MTE
F is
pub
lishe
d b
efo
re
the
bud
get
and
is t
he re
sult
of t
he t
op
-do
wn
pla
nnin
g fo
r th
e fis
cal
fram
ewo
rk. A
id is
incl
uded
in t
his
at a
gg
reg
ate
leve
l on
the
reve
nue
sid
e (b
y m
od
ality
and
bud
get
). Ta
nzan
ia is
intr
od
ucin
g s
cena
rio
pla
nnin
g in
ter
ms
of t
he J
oin
t A
ssis
tanc
e St
rate
gy
for
Tanz
ania
: in
MTE
Fs a
nd B
udg
et D
iges
ts in
futu
re, d
iffer
ent
leve
ls o
f aid
will
like
ly
corr
esp
ond
with
diff
eren
t sc
enar
ios.
The
deg
ree
to w
hich
thi
s w
ill
be
cap
ture
d o
n p
lan
and
fed
thr
oug
h to
bud
get
allo
catio
ns is
stil
l un
clea
r.
Sect
or P
ublic
Exp
end
iture
Rev
iew
s (P
ERs)
: PE
Rs
are
the
revi
ew a
nd
forw
ard
pla
nnin
g m
echa
nism
use
d in
Tan
zani
a, b
ridg
ing
str
ateg
ic
pla
ns, r
evie
ws
of p
ast
per
form
ance
, and
forw
ard
co
stin
g a
nd
allo
catio
ns. A
id in
flow
s ar
e d
iscu
ssed
to
a s
igni
fican
t le
vel o
f det
ail
in s
ecto
r PE
Rs.
Thi
s d
iscu
ssio
n in
clud
es n
ot
onl
y a
refle
ctio
n o
f the
d
egre
e o
f do
nor
inflo
ws,
but
als
o is
sues
aro
und
the
man
agem
ent
of
do
nor
flow
s (e
.g. l
ate
dis
bur
sem
ents
, hig
h re
po
rtin
g re
qui
rem
ents
, no
n-co
mp
lianc
e w
ith re
po
rtin
g re
qui
rem
ents
). D
ono
r fu
nds
are
mo
stly
refle
cted
on
the
inco
me
sid
e o
f bo
th re
curr
ent
and
d
evel
op
men
t b
udg
ets
(alm
ost
exc
lusi
vely
the
dev
elo
pm
ent
bud
get
), b
y d
ono
r. Th
e ex
pen
ditu
re s
ide
is d
iscu
ssed
in a
co
nso
lidat
ed m
anne
r. E
arm
arke
d d
ono
r fu
nds
are
not
dis
cuss
ed,
exce
pt
in p
roje
ct d
iscu
ssio
ns.
Bud
get
sub
mis
sion
s as
set
out
in th
e B
udg
et G
uid
elin
es: A
id o
ther
th
an b
udg
et s
upp
ort
is in
clud
ed in
the
bud
get
sub
mis
sio
ns t
o a
g
reat
leve
l of d
etai
l. M
DA
s an
d lo
cal a
utho
ritie
s ne
ed t
o p
rovi
de
do
nor f
und
ing
info
rmat
ion
by
pro
ject
thro
ugh
the
pro
ject
tab
les
and
th
e p
roje
ct d
ata
form
. Aid
is a
lso
dis
cuss
ed o
n ag
gre
gat
e ac
ross
all
MD
As
and
loca
l aut
horit
ies
in V
olu
me
1 o
f the
Gui
del
ines
.
Ove
rall,
aid
is in
clud
ed o
n p
lan
in T
anza
nia.
Tha
t is
par
ticul
arly
tr
ue fo
r se
cto
rs t
hat
have
SW
Ap
s in
pla
ce (s
uch
as e
duc
atio
n,
heal
th a
nd w
ater
), w
here
tec
hnic
al a
naly
sis
cap
acity
has
bee
n d
evel
op
ed a
nd is
sup
po
rted
.
Ho
wev
er, t
he s
epar
atio
n b
etw
een
diff
eren
t ai
d m
od
aliti
es
(i.e.
po
ole
d fu
ndin
g v
ersu
s se
cto
r b
udg
et s
upp
ort
) and
the
ir ac
com
pan
ying
diff
eren
t st
ruct
ures
mak
es t
he p
roce
ss a
nd t
he
out
com
e le
ss s
trea
mlin
ed t
han
it co
uld
be.
In t
he e
duc
atio
n se
cto
r, th
e p
oo
led
-fun
din
g a
rran
gem
ent
dev
elo
pm
ent
par
tner
s ha
ve fe
lt th
at t
he P
ER
pro
cess
, thr
oug
h w
hich
the
y fe
ed in
to t
he b
udg
et p
roce
ss, r
ecei
ves
insu
ffici
ent
atte
ntio
n fr
om
go
vern
men
t o
ffici
als,
whi
ch lo
wer
s th
e q
ualit
y o
f the
in
teg
ratio
n o
f aid
.
In t
he P
ER
s th
emse
lves
, on
the
recu
rren
t si
de,
whe
re d
ono
r fu
nds
are
also
incl
uded
, sp
ecifi
cally
in s
ecto
r su
pp
ort
and
p
oo
led
-fun
din
g a
rran
gem
ents
, do
nor
fund
reve
nues
are
no
t co
vere
d e
xplic
itly
eno
ugh:
it is
diffi
cult
to t
rack
whe
re t
he fu
nds
com
e fr
om
, ho
w t
hey
are
dis
bur
sed
and
whe
re p
rob
lem
s ar
e. It
is
ver
y d
ifficu
lt, if
no
t im
po
ssib
le, t
o re
conc
ile a
mo
unts
acr
oss
d
ocu
men
ts a
nd e
ven
with
in d
ocu
men
ts b
etw
een
recu
rren
t an
d d
evel
op
men
t ex
pen
ditu
re, a
nd lo
cal a
nd fo
reig
n fin
ance
d
exp
end
iture
. It
is n
ot
clea
r fr
om
the
do
cum
ents
whe
ther
the
sp
lit b
etw
een
recu
rren
t an
d d
evel
op
men
t re
fers
to
wha
t is
on
the
recu
rren
t and
wha
t is
on
the
dev
elo
pm
ent b
udg
et (b
ecau
se
it d
iffer
s fr
om
the
bud
get
bo
oks
the
mse
lves
), o
r w
heth
er it
re
fers
to
an
eco
nom
ic c
lass
ifica
tion
acro
ss t
he b
udg
ets.
Tanz
ania
n p
lann
ing
and
bud
get
ing
str
uctu
res
req
uire
the
par
ticip
atio
n o
f do
nors
and
oth
er s
take
hold
ers
in s
ecto
r an
d t
hem
atic
wo
rkin
g
gro
ups.
The
wid
e us
e o
f SW
Ap
s su
pp
ort
s th
e in
clus
ion
of a
id in
a m
ore
m
eani
ngfu
l way
on
pla
n. T
he q
ualit
y va
ries,
dep
end
ing
on
how
wel
l st
ruct
ures
wo
rk a
nd t
he a
naly
tical
cap
acity
in a
sec
tor.
Ho
w w
ell a
ll ai
d is
orig
inal
ly in
clud
ed o
n b
udg
et d
epen
ds
on
the
relia
bili
ty o
f do
nor
aid
pro
ject
ions
: in
qua
rter
ly b
udg
et e
xecu
tion
rep
ort
s ac
ross
mo
dal
ities
, no
t al
l aid
tha
t is
act
ually
dis
bur
sed
was
on
pla
n an
d
bud
get
, and
no
t al
l aid
tha
t w
as o
n p
lan
and
bud
get
was
dis
bur
sed
.
Cen
tral
gui
dan
ce is
lack
ing
on
how
aid
sho
uld
be
incl
uded
in P
ER
s, a
nd
on
how
to
co
nstr
uct
tab
les
that
mak
e th
e d
istr
ibut
ion
of a
id a
cro
ss a
nd
with
in b
udg
ets
clea
r.
The
req
uire
men
ts fo
r th
e in
clus
ion
of i
nfo
rmat
ion
on
aid
in t
he b
udg
et
sub
mis
sio
ns a
re m
et t
o a
mea
ning
ful l
evel
of d
etai
l: ho
wev
er, t
he
info
rmat
ion
can
onl
y b
e us
ed e
ffect
ivel
y if
ther
e is
eno
ugh
cap
acity
to
an
alys
e it
in t
he b
udg
et a
lloca
tion
pro
cess
.
82
tAn
ZA
niA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/w
hy n
ot?
)
On
bud
get
Acc
ord
ing
to
the
200
6 Su
rvey
on
Mo
nito
ring
the
Par
is D
ecla
ratio
n,
90%
of a
id d
isb
urse
d is
cap
ture
d o
n b
udg
et. H
ow
ever
, a re
cent
st
udy
foun
d t
hat
88%
of a
id in
the
go
vern
men
t’s a
id fl
ow
s d
atab
ase
is c
aptu
red
on
bud
get
.3
Bud
get
Dig
est (
mac
ro-le
vel s
upp
ortin
g d
ocum
ent f
or b
udg
et):
Dev
elo
pm
ent
inflo
ws
to t
he g
ove
rnm
ent
are
refle
cted
on
the
reso
urce
sid
e o
f the
fisc
al fr
amew
ork
, bro
ken
do
wn
by
pro
ject
loan
s an
d g
rant
s (g
roup
ed),
pro
gra
mm
e lo
ans
and
gra
nts
(gro
uped
), b
aske
t su
pp
ort
loan
s an
d b
aske
t su
pp
ort
gra
nts
(no
t g
roup
ed),
HIP
C re
lief,
and
the
Mul
tilat
eral
Deb
t R
elie
f Ini
tiativ
e. P
roje
ct-r
elat
ed
do
nor
inflo
ws
are
also
refle
cted
on
the
exp
end
iture
sid
e, in
the
d
evel
op
men
t b
udg
et, w
here
pro
ject
exp
end
iture
is b
roke
n d
ow
n in
to d
om
estic
ally
fund
ed a
nd fo
reig
n fu
nded
. The
fisc
al fr
amew
ork
is
giv
en fr
om
14
year
s b
efo
re t
he b
udg
et y
ear
(200
6/07
) to
a s
eco
nd
out
er p
roje
ctio
n ye
ar (2
008/
09).
Min
istr
y an
d re
gio
nal b
udg
ets:
Aid
is n
ot
refle
cted
in t
he re
curr
ent
min
istr
y b
udg
et o
r re
gio
nal b
udg
et e
stim
ates
. Thi
s is
to
be
exp
ecte
d. G
ener
al a
nd s
ecto
r b
udg
et s
upp
ort
reg
iste
rs o
n th
e re
venu
e si
de
of t
he o
vera
ll fis
cal f
ram
ewo
rk.
Aid
is in
clud
ed in
the
dev
elo
pm
ent
bud
get
, by
pro
ject
and
by
do
nor,
and
with
a n
ote
ab
out
whe
ther
it is
a g
rant
or
a lo
an.
Tanz
ania
is in
tro
duc
ing
sce
nario
pla
nnin
g in
its
pla
nnin
g a
nd
bud
get
ing
for
aid
. At
this
sta
ge
it is
unc
lear
to
wha
t d
egre
e th
is
wo
uld
influ
ence
ho
w a
id is
cap
ture
d o
n b
udg
et.
Ove
rall,
gen
eral
and
pro
gra
mm
e b
udg
et s
upp
ort
and
b
aske
t fu
nds
are
cap
ture
d w
ell,
TA a
nd p
roje
ct s
upp
ort
less
w
ell.
Alm
ost
all
sup
po
rt t
o c
entr
al g
ove
rnm
ent
is c
aptu
red
, b
ut c
aptu
re o
f pro
ject
sup
po
rt d
irect
ly t
o d
istr
icts
is n
ot
as
com
ple
te.
It is
diffi
cult
to re
conc
ile t
he o
vera
ll le
vel o
f aid
cap
ture
d in
th
e fis
cal f
ram
ewo
rk w
ith a
id a
s re
flect
ed in
the
low
er le
vels
o
f exp
end
iture
(min
istr
ies
and
reg
ions
). It
see
ms
that
fore
ign
finan
ced
dev
elo
pm
ent
exp
end
iture
eq
uals
pro
gra
mm
e b
udg
et s
upp
ort
, bas
ket
fund
s an
d p
roje
ct e
xpen
ditu
re. H
ow
p
rog
ram
me
bud
get
sup
po
rt is
refle
cted
is u
ncer
tain
.
Ther
e is
mo
re d
etai
l on
aid
in t
he d
evel
op
men
t b
udg
et
on
the
exp
end
iture
sid
e: it
can
be
iso
late
d b
y m
inis
try
and
m
ain
pro
gra
mm
e. H
ow
ever
, the
cla
ssifi
catio
n st
ruct
ure
of
the
dev
elo
pm
ent
bud
get
do
es n
ot
allo
w fo
r co
nsis
tent
cl
assi
ficat
ion
with
the
recu
rren
t b
udg
et, l
imiti
ng a
naly
sis.
It is
b
y p
roje
ct o
nly.
The
shift
to
bud
get
sup
po
rt s
ince
200
1 ha
s im
pro
ved
cap
ture
in t
he
bud
get
at
agg
reg
ate
leve
l. Th
e cr
eatio
n o
f bas
ket
fund
s fo
r se
cto
r w
ide
refo
rm p
rog
ram
mes
(hea
lth, e
duc
atio
n, lo
cal g
ove
rnm
ent
refo
rm
and
pub
lic s
ecto
r re
form
) has
had
a s
imila
r ef
fect
. The
Jo
int
Ass
ista
nce
Stra
teg
y fo
r Ta
nzan
ia is
like
ly, o
ver
time,
to
incr
ease
the
pro
po
rtio
n o
f fu
nds
that
flo
w t
hro
ugh
bud
get
sup
po
rt. T
he c
reat
ion
by
the
Min
istr
y o
f Fin
ance
in 2
001
of a
dat
abas
e o
n ai
d in
flow
s ha
s in
crea
sed
bud
get
co
vera
ge
of s
upp
ort
. A 2
002
req
uest
to
do
nors
to
eith
er s
witc
h to
g
ove
rnm
ent
syst
ems
or
imp
rove
rep
ort
ing
on
pro
ject
sup
po
rt in
crea
sed
th
e co
vera
ge
in t
he b
udg
et o
f pro
ject
sup
po
rt. T
he d
ono
r-g
ove
rnm
ent
par
tner
ship
in T
anza
nia
has
bee
n le
ss a
ble
to
ag
ree
on
inst
itutio
nal
arra
ngem
ents
for
man
agin
g T
A, c
om
par
ed t
o o
ther
aid
mo
dal
ities
. Th
e g
ove
rnm
ent
has
clea
rly s
tate
d it
s at
titud
e at
Co
nsul
tativ
e G
roup
m
eetin
gs
and
on
oth
er o
ccas
ions
: bud
get
sup
po
rt is
the
pre
ferr
ed
mo
dal
ity. T
he g
ove
rnm
ent’s
att
itud
e to
sec
tor
bas
ket
fund
ing
is a
bit
amb
iguo
us: w
hile
the
Min
istr
y o
f Fin
ance
str
ong
ly a
rgue
s fo
r b
udg
et
sup
po
rt, l
ine
min
istr
ies
are
mo
re in
tere
sted
in s
ecto
r b
aske
t fu
ndin
g, a
nd
even
pro
ject
sup
po
rt –
the
se m
od
aliti
es p
rovi
de
line
min
istr
ies
with
mo
re
influ
ence
tha
n b
udg
et s
upp
ort
.
Why
is t
here
diffi
culty
in re
conc
iling
aid
dat
a? T
he fo
rmat
s o
f rec
urre
nt
estim
ates
do
no
t le
nd t
hem
selv
es t
o re
flect
ing
whe
re e
arm
arke
d a
id
sup
po
rt g
oes
in t
he b
udg
et. O
ne w
oul
d n
ot
exp
ect
bud
get
sup
po
rt t
o
be
refle
cted
on
the
exp
end
iture
sid
e ne
cess
arily
(alth
oug
h it
may
ass
ist
the
go
vern
men
t to
get
mo
re b
udg
et s
upp
ort
if it
can
be
trac
ked
), b
ut t
he
pro
gra
mm
e b
udg
et s
upp
ort
tha
t se
ems
to b
e in
the
recu
rren
t b
udg
et
dis
app
ears
bet
wee
n th
e fis
cal f
ram
ewo
rk a
nd t
he a
lloca
tions
.
One
of t
he re
aso
ns fo
r p
roje
ct a
id in
par
ticul
ar n
ot
bei
ng c
aptu
red
on
bud
get
– d
esp
ite b
eing
on
pla
n –
is b
ecau
se t
he M
inis
try
of F
inan
ce
do
es n
ot
put
pro
ject
s w
hich
fall
out
sid
e o
f MD
A c
eilin
gs
on
bud
get
.4
On
par
liam
ent
Aid
tha
t is
refle
cted
in t
he b
udg
et e
stim
ates
is v
ote
d b
y p
arlia
men
t.
In T
anza
nia,
a d
istin
ctio
n is
mad
e b
etw
een
the
bud
get
est
imat
es
and
the
ap
pro
ved
bud
get
, alth
oug
h th
ere
is r
arel
y an
y d
iffer
ence
b
etw
een
the
two
.
Parli
amen
t is
invo
lved
in t
he b
udg
et p
roce
ss in
tw
o p
hase
s.
Bef
ore
the
bud
get
est
imat
es a
re fi
nalis
ed, m
inis
teria
l bud
get
p
rop
osa
ls a
re p
rese
nted
to
par
liam
enta
ry c
om
mitt
ees
for
dis
cuss
ion.
The
Min
istr
y o
f Fin
ance
can
inco
rpo
rate
co
mm
ents
fr
om
the
co
mm
ittee
s in
the
fina
l est
imat
es. I
n th
e se
cond
p
hase
, par
liam
ent
vote
s th
e es
timat
es in
to a
pp
rop
riatio
ns. I
n p
rinci
ple
, par
liam
ent
is a
lso
invo
lved
in t
he P
ER
pro
cess
, but
th
is in
volv
emen
t is
stil
l rel
ativ
ely
wea
k.
The
rela
tivel
y w
eak
invo
lvem
ent
of p
arlia
men
t ca
n b
e at
trib
uted
to
po
litic
al fa
cto
rs (s
tro
ng p
arty
co
hesi
on
acro
ss t
he e
xecu
tive
and
p
arlia
men
t) a
nd c
apac
ity fa
cto
rs.
83
tAn
ZA
niA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/w
hy n
ot?
)
On
trea
sury
Acc
ord
ing
to
the
200
6 Su
rvey
on
Mo
nito
ring
the
Par
is D
ecla
ratio
n,
76%
of d
isb
urse
d a
id is
on
trea
sury
(i.e
. dis
bur
sed
thr
oug
h th
e ex
cheq
uer
syst
em) –
in o
ther
wo
rds,
Cha
nnel
1. T
he re
mai
nder
us
es C
hann
el 3
(par
alle
l im
ple
men
tatio
n m
echa
nism
s, s
uch
as
pro
ject
imp
lem
enta
tion
units
– t
here
are
56
such
uni
ts in
Tan
zani
a).
The
liter
atur
e p
rovi
des
onl
y o
ne c
ase
of e
vid
ence
of t
he p
oss
ible
us
e o
f Cha
nnel
2 (d
isb
urse
men
ts d
irect
ly t
o t
he b
ank
acco
unts
o
f go
vern
men
t un
its) i
n th
e fo
rm o
f dire
ct d
isb
urse
men
ts t
o lo
cal
auth
orit
ies
and
go
vern
men
t un
its.5 It
is u
ncle
ar, h
ow
ever
, whe
ther
th
ese
are
to p
roje
ct m
anag
emen
t un
its a
t th
ese
leve
ls (o
r so
me
par
alle
l str
uctu
re) w
hich
op
erat
e co
mm
erci
al b
ank
acco
unts
, or
into
ac
coun
ts o
per
ated
by
the
units
.
The
Bud
get
Exe
cutio
n re
po
rts
sho
w d
iffer
ence
s in
the
p
red
icta
bili
ty o
f dis
bur
sem
ent
by
aid
mo
dal
ity. A
ltho
ugh
the
scal
e o
f the
diff
eren
ces
is n
ot
cons
iste
nt t
hro
ugh
the
year
(bud
get
sup
po
rt is
fro
nt-lo
aded
, fo
r ex
amp
le),
the
bes
t p
erfo
rman
ce is
usu
ally
on
the
mul
tilat
eral
deb
t re
lief f
und
(93%
in
qua
rter
1 o
f 200
6/07
), an
d t
he s
eco
nd b
est
per
form
ance
(im
pro
ving
fro
m lo
ws
in t
he fi
rst
half
of t
he d
ecad
e) is
on
bud
get
sup
po
rt fu
nds
(73%
in q
uart
er 1
of 2
006/
07).
The
wo
rst
per
form
ance
is o
n b
aske
t su
pp
ort
fund
s (2
6% in
qua
rter
1 o
f 20
06/0
7), w
ith p
roje
cts
also
bei
ng p
oo
r (5
0% in
qua
rter
1 o
f 20
06/0
7).
The
PER
s re
view
ed re
fer
to t
he e
ffici
ency
of d
iffer
ent
dis
bur
sem
ent
mec
hani
sms.
The
re a
re d
elay
s if
fund
s ar
e d
isb
urse
d t
hro
ugh
the
exch
eque
r sy
stem
. The
re is
, ho
wev
er,
evid
ence
in o
lder
do
cum
ents
tha
t d
elay
s ha
ve b
een
exp
erie
nced
in t
he t
rans
fer
of f
und
s fr
om
sec
tor
min
istr
ies
to
dis
tric
ts, w
here
pro
ject
s ar
e im
ple
men
ted
.
The
incr
ease
d u
se o
f bud
get
sup
po
rt, p
rog
ram
me
bud
get
sup
po
rt
and
bas
ket
fund
mo
dal
ities
has
incr
ease
d t
he p
erce
ntag
e o
f aid
tha
t is
d
isb
urse
d t
hro
ugh
the
trea
sury
sys
tem
. Pro
ject
aid
dis
bur
sem
ent
is s
till
qui
te lo
w, a
ltho
ugh
the
chan
nel h
as b
een
crea
ted
to
dis
bur
se p
roje
ct a
id
thro
ugh
the
trea
sury
sys
tem
.
Del
ays
in t
he d
isb
urse
men
t o
f do
nor
fund
s th
roug
h th
e ex
cheq
uer
syst
em o
ften
hav
e to
do
with
no
n-p
erfo
rman
ce o
f pro
gra
mm
es a
nd
pro
ject
s ag
ains
t re
po
rtin
g re
qui
rem
ents
or
per
form
ance
tar
get
s, r
athe
r th
an d
elay
s in
here
nt in
the
sys
tem
. Acr
oss
rep
ort
s, t
here
is a
lso
evi
den
ce
of d
elay
s in
do
nor
pro
ced
ures
on
dis
bur
sem
ents
.
The
del
ays
in p
roje
ct d
isb
urse
men
ts a
re a
ttrib
uted
bo
th t
o d
elay
s in
ca
ptu
ring
dis
bur
sem
ents
(bec
ause
the
y o
ccur
out
sid
e o
f the
exc
heq
uer
syst
em) a
nd t
o a
ctua
l del
ays
in d
isb
urse
men
ts o
n ac
coun
t o
f rep
ort
ing
re
qui
rem
ents
, slo
wer
sp
end
ing
tha
n ex
pec
ted
, or
do
nor
del
ays.
As
the
per
cent
age
of d
ono
r fu
nds
dis
bur
sed
thr
oug
h th
e ex
cheq
uer
syst
em h
as
incr
ease
d, m
ore
fund
s ar
e flo
win
g t
hro
ugh
sect
or
min
istr
ies
to p
roje
cts
at
loca
l lev
el, r
athe
r th
an d
irect
ly fr
om
the
do
nors
to
the
dis
tric
ts. M
inis
trie
s ar
e so
met
imes
slo
w t
o d
isb
urse
.
On
acco
unt
Acc
ord
ing
to th
e 20
06 S
urve
y on
Mon
itorin
g th
e Pa
ris D
ecla
ratio
n,
60.1
% o
f dis
bur
sed
aid
to th
e g
over
nmen
t sec
tor u
sed
cou
ntry
ac
coun
ting
sys
tem
s. T
his
mea
ns th
at n
ot a
ll C
hann
el 1
fund
s us
e co
untr
y ac
coun
ting
sys
tem
s. S
ome
don
ors,
how
ever
, do
not
dis
bur
se th
roug
h th
e ex
cheq
uer,
but
use
cou
ntry
acc
ount
ing
sys
tem
s (i.
e. C
hann
el 3
fund
s, a
ccou
nted
for t
hrou
gh
coun
try
acco
untin
g
syst
ems)
. 53%
of a
id o
n th
e ai
d fl
ow d
atab
ase
is re
gis
tere
d o
n g
over
nmen
t acc
ount
s, w
heth
er th
roug
h th
e C
Cha
nnel
(cas
h d
isb
urse
d th
roug
h th
e B
ank
of T
anza
nia
– i.e
. Cha
nnel
1) o
r the
D
Cha
nnel
(the
con
trib
utio
n of
goo
ds
and
ser
vice
s p
aid
for t
hrou
gh
par
alle
l arr
ang
emen
ts –
i.e.
Cha
nnel
3).6 S
ee th
e ta
ble
bel
ow.7
Gap
bet
wee
n na
tion
al b
udg
et e
stim
ates
, dis
bur
sem
ents
and
aid
flow
s
Ove
rall
do
cum
ents
, suc
h as
the
Pub
lic E
xpen
ditu
re a
nd
Fina
ncia
l Acc
oun
tab
ility
Rep
ort
, no
te im
pro
vem
ents
in t
he
qua
lity
of c
aptu
re o
f fina
ncia
l inf
orm
atio
n o
ver
the
last
yea
rs,
but
als
o t
hat
ther
e ar
e st
ill s
hort
com
ing
s in
the
acc
urac
y an
d
timel
ines
s o
f fina
ncia
l rep
ort
ing
.
The
intr
od
uctio
n o
f an
IFM
IS a
nd it
s ro
ll-o
ut t
o t
he lo
cal l
evel
of
go
vern
men
t ha
s in
crea
sed
co
nfid
ence
in t
he re
liab
ility
of c
aptu
re in
g
ove
rnm
ent’s
acc
oun
ting
sys
tem
s.
The
dis
crep
ancy
in t
he u
se b
y C
hann
el 1
and
Cha
nnel
3 fu
nds
of c
oun
try
acco
untin
g s
yste
ms
is e
xpla
ined
by
two
mec
hani
sms.
Firs
t, b
aske
t su
pp
ort
fund
s (s
uch
as t
he h
ealth
fund
) are
dis
bur
sed
thr
oug
h th
e ex
cheq
uer
syst
em (t
hro
ugh
an a
cco
unt
at t
he B
ank
of T
anza
nia)
, but
th
eir
use
is a
cco
unte
d fo
r o
utsi
de
of t
he g
ove
rnm
ent
syst
ems
thro
ugh
par
alle
l ad
min
istr
ativ
e ar
rang
emen
ts (a
ltho
ugh
thes
e ar
e so
mew
hat
emb
edd
ed in
min
istr
ies)
. Sec
ond
, at
the
sam
e tim
e, t
here
is a
mec
hani
sm
for
fund
s to
be
dis
bur
sed
out
sid
e o
f the
exc
heq
uer
syst
em, b
ut s
till t
o b
e ac
coun
ted
for
in t
he a
cco
unt
of t
he g
ove
rnm
ent
of T
anza
nia.
Do
nor
reg
ulat
ions
als
o e
xpla
in s
om
e o
f the
dis
crep
ancy
: so
me
do
nors
ca
nno
t d
isb
urse
cas
h th
roug
h g
ove
rnm
ent
syst
ems
(and
the
refo
re
use
Cha
nnel
3),
but
stil
l use
go
vern
men
t ac
coun
ting
sys
tem
s vi
a th
e m
echa
nism
des
crib
ed a
bo
ve.
Cen
tral
and
sat
ellit
e tr
easu
ries
are
resp
ons
ible
for
pay
men
ts, a
nd t
he
Offi
ce o
f the
Acc
oun
tant
Gen
eral
for
mai
ntai
ning
the
acc
oun
ts. H
ow
ever
, se
cto
r m
inis
trie
s an
d lo
cal g
ove
rnm
ents
are
resp
ons
ible
for
the
rest
o
f the
ste
ps
in t
he e
xpen
ditu
re m
anag
emen
t p
roce
ss. T
his
mea
ns
that
tre
asur
ies
and
the
Acc
oun
tant
Gen
eral
rely
on
the
timel
ines
s an
d
do
cum
enta
tion
of o
ther
par
ties.
It is
the
refo
re u
nder
stan
dab
le, g
iven
the
un
even
ness
of c
apac
ity, t
hat
ther
e ar
e st
ill p
rob
lem
s w
ith t
he t
imel
ines
s an
d a
ccur
acy
of a
cco
unts
. The
furt
her
roll-
out
of t
he IF
MIS
and
rela
ted
ca
pac
ity b
uild
ing
will
hel
p a
dd
ress
the
se is
sues
.
84
tAn
ZA
niA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/w
hy n
ot?
)
On
aud
itA
pp
roxi
mat
ely
the
sam
e p
erce
ntag
e o
f aid
tha
t is
on
acco
unt
is
aud
ited
thr
oug
h g
ove
rnm
ent
syst
ems.
Ag
ain,
ho
wev
er, t
here
are
m
ore
fund
s o
n ac
coun
t th
an a
udite
d, a
nd v
ice
vers
a, re
flect
ing
d
ono
r re
gul
atio
ns a
nd t
rust
in t
he g
ove
rnm
ent’s
aud
it sy
stem
s.
The
Pub
lic E
xpen
ditu
re a
nd F
inan
cial
Acc
oun
tab
ility
Rep
ort
no
tes
rece
nt im
pro
vem
ents
in t
he t
imel
ines
s o
f go
vern
men
t au
diti
ng p
roce
sses
, but
stil
l no
tes
conc
erns
with
the
co
mp
lianc
e o
f aud
it p
roce
dur
es a
nd s
cop
e w
ith in
tern
atio
nal
stan
dar
ds,
and
with
the
wea
knes
s o
f par
liam
enta
ry p
roce
sses
in
follo
win
g t
hro
ugh
on
aud
it fin
din
gs.
The
rela
tivel
y lo
w le
vel o
f aid
on
aud
it, c
om
par
ed t
o w
hat
is o
n b
udg
et
and
on
par
liam
ent,
refle
cts
do
nor
reg
ulat
ions
and
low
er t
rust
in t
he
qua
lity
of a
udits
for
the
time
bei
ng. F
or
exam
ple
, the
re a
re a
few
do
nors
fo
r w
hom
mo
re fu
nds
are
on
acco
unt
than
on
aud
it: p
artic
ular
ly D
enm
ark.
The
dis
crep
ancy
bet
wee
n w
hat
is o
n C
hann
el 1
and
Cha
nnel
3, w
hat
is o
n ac
coun
t o
r no
t, a
nd w
hat
is o
n au
dit
or
not,
als
o li
es in
bas
ket
fund
s: t
he h
ealth
bas
ket
fund
ag
reem
ent,
for
exam
ple
, req
uire
s fu
nds
to b
e d
isb
urse
d t
hro
ugh
the
exch
eque
r, ac
coun
ted
for
by
the
bas
ket
fund
ad
min
istr
ativ
e ar
rang
emen
ts, a
nd a
udite
d b
y th
e g
ove
rnm
ent
(co
ntra
ctin
g in
dep
end
ent
aud
itors
if re
qui
red
). Th
is a
cco
unts
for
bas
ket
fund
co
untr
ies
(Sw
eden
, the
Net
herla
nds,
the
UK
, etc
.) ha
ving
mo
re
fund
s o
n au
dit
than
on
acco
unt.
The
conc
erns
ab
out
go
vern
men
t au
dits
are
tha
t th
e le
gal
req
uire
men
ts
are
lag
gin
g b
ehin
d in
tern
atio
nal g
oo
d p
ract
ices
and
tha
t th
e A
udito
r G
ener
al’s
offi
ce la
cks
cap
acity
. Bo
th t
hese
asp
ects
are
rece
ivin
g a
tten
tion
in re
form
pro
gra
mm
es.
On
rep
ort
Stra
teg
ic p
erfo
rman
ce re
po
rtin
g: T
he M
kuku
ta fr
amew
ork
incl
udes
m
oni
torin
g o
f ind
icat
ors
and
an
annu
al im
ple
men
tatio
n re
po
rt,
mo
stly
at
the
activ
ity a
nd o
utp
ut le
vels
. Whi
le a
id fi
nanc
ial fl
ow
s o
n th
e re
venu
e si
de
do
no
t re
ceiv
e m
uch
atte
ntio
n, t
he fr
amew
ork
d
oes
pro
vid
e re
gul
ar u
pd
ates
on
wha
t ai
d is
use
d fo
r.
Bud
get
exe
cutio
n an
d fi
nanc
ial r
epo
rts:
Aid
is re
flect
ed a
t ag
gre
gat
e le
vel i
n g
ove
rnm
ent
finan
cial
sta
tem
ents
by
typ
e o
f flo
w
and
mo
dal
ity. I
t is
dis
cuss
ed in
so
me
det
ail i
n th
e q
uart
erly
bud
get
ex
ecut
ion
rep
ort
s p
ublis
hed
by
the
Min
istr
y o
f Fin
ance
, mo
re s
o
on
the
reve
nue
sid
e (w
here
the
pre
dic
tab
ility
of d
isb
urse
men
ts b
y m
od
ality
rece
ives
gre
at a
tten
tion)
tha
n o
n th
e ex
pen
ditu
re s
ide.
Th
ere
are
agg
reg
ate
reve
nue
and
exp
end
iture
per
form
ance
tab
les,
w
here
aid
dis
bur
sem
ent
per
form
ance
is re
flect
ed b
y m
od
ality
.
At
sect
or
leve
l: Th
e re
po
rtin
g m
echa
nism
s o
f SW
Ap
s in
clud
e an
in-
dep
th d
iscu
ssio
n o
f aid
flo
ws
on
the
reve
nue
and
exp
end
iture
sid
es.
The
annu
al P
ER
exe
rcis
e al
so in
clud
es a
sig
nific
ant
revi
ew a
spec
t,
par
ticul
arly
of fi
nanc
ial fl
ow
s. A
id, h
ow
ever
, is
mo
stly
revi
ewed
on
the
reve
nue
sid
e: t
he e
xpen
ditu
re d
iscu
ssio
ns d
o n
ot
dis
ting
uish
b
etw
een
aid
and
loca
lly fi
nanc
ed e
xpen
ditu
re.
By
aid
mo
dal
ity: T
he a
nnua
l GB
S re
view
rep
ort
pro
vid
es a
n o
verv
iew
o
f aid
inflo
ws
and
rep
ort
s ag
ains
t th
e p
erfo
rman
ce a
sses
smen
t fr
amew
ork
on
the
use
of a
id, p
artic
ular
ly G
BS.
Pro
ject
-leve
l re
po
rtin
g is
no
t p
ublic
and
dep
end
s o
n th
e p
artic
ular
pro
ject
ar
rang
emen
ts.
The
2006
GB
S re
view
and
oth
er re
po
rts
note
tha
t re
po
rtin
g
is w
eake
r th
an it
co
uld
be:
the
re is
no
rep
ort
ing
ag
ains
t th
e St
rate
gic
Bud
get
Allo
catio
n Sy
stem
(the
sys
tem
use
d in
bud
get
p
rep
arat
ion
to li
nk a
lloca
tions
to
PR
SP, M
TEF
and
bud
get
cl
uste
rs).
Tanz
ania
, lik
e m
any
coun
trie
s w
hich
rece
ive
hig
h le
vels
of a
id, h
as a
ver
y co
mp
lex
rep
ort
ing
arc
hite
ctur
e. T
he re
aso
ns fo
r th
is a
re: (
i) th
e va
riety
o
f rep
ort
ing
req
uire
men
ts t
hat
go
with
a h
igh
num
ber
of d
iffer
ent
do
nor
agre
emen
ts; (
ii) t
he p
rolif
erat
ion
of p
lann
ing
inst
rum
ents
; and
(iii)
the
lack
o
f str
eam
linin
g o
f the
co
untr
y’s
ow
n fin
anci
al a
nd p
olic
y p
erfo
rman
ce
rep
ort
ing
mec
hani
sms.
The
latt
er is
als
o re
late
d t
o in
stitu
tiona
l co
mp
lexi
ty a
roun
d t
he m
anag
emen
t o
f exp
end
iture
(diff
eren
t ag
enci
es,
diff
eren
t b
udg
ets)
. The
str
eam
linin
g o
f do
nor
mec
hani
sms,
if t
rans
late
d
into
rep
ort
ing
mec
hani
sms,
will
ass
ist
in re
duc
ing
the
co
mp
lexi
ty. T
he
inte
gra
tion
of d
ialo
gue
mec
hani
sms
is s
tart
ing
to
bea
r fr
uit
inso
far
as t
he
2006
GB
S re
view
dre
w o
n M
kuku
ta re
po
rtin
g a
s an
alyt
ical
inp
ut.
The
rela
tivel
y w
eak
refle
ctio
n o
f aid
on
the
exp
end
iture
sid
e o
f bud
get
ex
ecut
ion
rep
ort
s is
sym
pto
mat
ic o
f the
wea
ker
refle
ctio
n o
f exp
end
iture
o
vera
ll. E
xpen
ditu
re is
no
t re
flect
ed b
y m
inis
try
or
by
reg
ion,
onl
y b
y ag
gre
gat
e ec
ono
mic
cat
ego
ry o
f sp
end
ing
.
Ther
e ar
e a
num
ber
of e
xpen
ditu
re, a
cco
untin
g a
nd re
po
rtin
g d
atab
ases
in
use
. Ho
wev
er, w
eak
linka
ges
bet
wee
n th
em d
etra
ct fr
om
the
co
mp
rehe
nsiv
enes
s an
d q
ualit
y o
f rep
ort
ing
. The
re a
re, f
or
exam
ple
, w
eak
links
bet
wee
n th
e St
rate
gic
Bud
get
and
Allo
catio
n Sy
stem
, Pl
anR
EP
(reco
rdin
g a
nd re
po
rtin
g o
n d
evel
op
men
t p
roje
cts)
, the
IFM
IS
and
RIM
KU
(rep
ort
ing
on
Mku
kuta
ind
icat
ors
) dat
abas
es. T
his
mak
es
it d
ifficu
lt to
link
fina
ncia
l rep
ort
s to
no
n-fin
anci
al (a
ctiv
ity, o
utp
uts,
o
utco
mes
and
imp
acts
) rep
ort
s.
85
Gr
ou
p B
cA
se
st
ud
y c
ou
nt
rie
s
GH
An
A
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/w
hy n
ot?
)
On
pla
nA
t o
vera
ll st
rate
gic
leve
l, th
e G
row
th a
nd P
ove
rty
Red
uctio
n St
rate
gy
(200
6–09
) inc
lud
es e
xter
nally
fina
nced
gra
nts
and
cre
dits
, in
clud
ing
bud
get
sup
po
rt g
rant
s/cr
edits
and
pro
gra
mm
e/p
roje
ct
gra
nts/
cred
its.
Und
er t
he a
usp
ices
of t
he M
TEF,
eac
h M
DA
pro
duc
es o
ne o
r m
ore
b
udg
et v
olu
mes
set
ting
out
the
det
ails
of b
udg
et e
stim
ates
for
pro
gra
mm
es a
nd a
ctiv
ities
for
the
com
ing
bud
get
yea
r. Fo
r ea
ch
activ
ity, t
he e
stim
ates
set
out
the
typ
e o
f fun
ds
(e.g
. Co
nso
lidat
ed
Fund
, Int
erna
lly G
ener
ated
Fun
ds,
HIP
C fu
nds,
or
dev
elo
pm
ent
par
tner
fund
s), t
he s
our
ce o
f fun
ds
(incl
udin
g id
entif
ying
ind
ivid
ual
do
nors
), au
tho
risat
ion
(e.g
. nat
iona
l bud
get
), re
spo
nsib
le in
stitu
tion
and
sec
tor.
Do
nor
fund
s no
rmal
ly a
pp
ly t
o t
he it
ems
for
serv
ices
(i.e
. g
oo
ds
and
ser
vice
s) a
nd in
vest
men
t. T
hese
MD
A M
TEF
volu
mes
ar
e th
e m
ain
link
bet
wee
n b
udg
ets
and
pla
ns. I
n ad
diti
on,
for
the
mo
st im
po
rtan
t sp
end
ing
sec
tors
in w
hich
ext
erna
l fina
ncin
g is
m
ost
pro
min
ent,
ann
ual j
oin
t g
ove
rnm
ent–
do
nor
sect
or
revi
ews
are
held
: bro
ad s
trat
egic
inve
stm
ent
pla
ns fo
r th
e se
cto
r ar
e d
iscu
ssed
, cu
lmin
atin
g in
the
pre
par
atio
n o
f the
sec
tor’s
ann
ual P
rog
ram
me
of
Wo
rk.
Whi
le t
he G
row
th a
nd P
ove
rty
Red
uctio
n St
rate
gy
is in
tend
ed t
o b
e th
e g
uid
ing
prin
cip
le fo
r p
riorit
isin
g in
vest
men
ts, i
t is
insu
ffici
ently
d
etai
led
to
ena
ble
MD
As
to p
lan
thei
r d
etai
led
inve
stm
ent
pro
gra
mm
es.8
Als
o, i
n p
ract
ice
it is
no
t cl
ear
how
op
erat
iona
l the
MTE
F is
, in
par
ticul
ar t
he d
egre
e o
f att
entio
n g
iven
to
the
se d
etai
led
est
imat
es
whe
n sc
rutin
isin
g, a
pp
rovi
ng a
nd im
ple
men
ting
bud
get
s. T
here
se
ems
to b
e a
dis
conn
ect
bet
wee
n p
lans
in t
he M
TEF
and
the
im
ple
men
tatio
n o
f the
bud
get
. Fo
r ex
amp
le, t
here
are
sig
nific
ant
varia
nces
bet
wee
n b
udg
et p
lans
and
act
ual b
udg
et o
ut-t
urns
. Th
is re
flect
s, in
par
t, t
he la
rge
num
ber
of a
ctiv
ities
, whi
ch m
akes
it
diffi
cult
for
par
liam
ent
and
oth
ers
to s
ee h
ow
reso
urce
s ar
e al
loca
ted
to
war
ds
spec
ific
go
vern
men
t p
olic
ies.
Unt
il re
cent
ly,
the
MTE
F co
ntai
ned
mo
re t
han
17,0
00 a
ctiv
ities
. The
num
ber
of
activ
ities
was
red
uced
to
45
stan
dar
d a
ctiv
ities
in p
rep
arat
ion
for
the
2007
bud
get
(and
200
7–09
MTE
F).
The
dis
conn
ect
bet
wee
n M
TEF
activ
ities
and
the
bud
get
refle
cts
bo
th: (
i) ca
pac
ity c
ons
trai
nts
at M
DA
s to
pla
n an
d fu
lly c
ost
out
ac
tiviti
es; a
nd (i
i) th
e cu
rren
t la
ck o
f a m
echa
nism
to
link
act
iviti
es t
o
reso
urce
s d
urin
g b
udg
et im
ple
men
tatio
n.
The
curr
ent
clas
sific
atio
n sy
stem
use
d b
y th
e C
ont
rolle
r an
d
Acc
oun
tant
Gen
eral
do
es n
ot
incl
ude
the
activ
ity c
od
es, s
o
rep
ort
ing
on
exp
end
iture
s m
ay n
ot
be
do
ne o
n th
e b
asis
of p
lann
ed
activ
ities
.
Ind
ivid
ual M
DA
MTE
F es
timat
es a
re n
ot
com
pre
hens
ive,
sin
ce
the
item
s fo
r p
erso
nal e
mo
lum
ents
and
ad
min
istr
atio
ns (b
oth
p
erso
nnel
-rel
ated
), b
eing
less
dis
cret
iona
ry in
pra
ctic
e, a
re n
ot
prio
ritis
ed in
line
with
the
item
s fo
r se
rvic
es (g
oo
ds
and
ser
vice
s)
and
inve
stm
ent.
86
GH
An
A
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/w
hy n
ot?
)
On
bud
get
The
go
vern
men
t o
f Gha
na’s
annu
al b
udg
et d
ocu
men
tatio
n co
mp
rises
: (i)
the
Ap
pro
pria
tions
Act
; (ii)
the
Bud
get
Sta
tem
ent
(co
ntai
ning
the
min
iste
r’s b
udg
et s
pee
ch);
and
(iii)
sep
arat
e vo
lum
es
cont
aini
ng t
he a
nnua
l est
imat
es fo
r ea
ch M
DA
. (Se
e “O
n p
lan”
ab
ove
.)
The
Ap
pro
pria
tions
Act
and
the
ap
pen
dic
es in
the
Bud
get
St
atem
ent
incl
ude
tab
les
sho
win
g p
lann
ed d
isb
urse
men
ts o
f lo
ans
and
gra
nts.
Sp
ecifi
cally
, the
bud
get
do
cum
ents
incl
ude
det
ails
of:
(i) b
udg
et s
upp
ort
; and
(ii)
exte
rnal
pro
ject
ass
ista
nce
for
pro
ject
s ad
min
iste
red
thr
oug
h PI
Us
in M
DA
s.
In t
he 2
007
Bud
get
Sta
tem
ent,
pla
nned
dis
bur
sem
ents
of e
xter
nally
fin
ance
d e
xpen
ditu
res
for
the
com
ing
bud
get
yea
r ar
e sh
ow
n b
y ec
ono
mic
item
(per
sona
l em
olu
men
ts, a
dm
inis
trat
ion,
ser
vice
s an
d
inve
stm
ent)
and
by
MD
A. F
or
the
two
forw
ard
MTE
F ye
ars
(and
fo
r th
e co
min
g b
udg
et y
ear)
, pla
nned
dis
bur
sem
ents
of e
xter
nal
finan
ce a
re s
how
n b
y M
DA
. As
ind
icat
ed u
nder
“O
n p
lan”
, in
the
mo
re d
etai
led
MD
A v
olu
mes
of e
stim
ates
, the
typ
e (g
rant
/cre
dit)
, so
urce
(whi
ch d
ono
r) a
nd a
mo
unt
of fi
nanc
ing
is id
entifi
ed fo
r ea
ch
activ
ity.
The
bud
get
do
cum
ents
do
no
t in
clud
e: (i
) ext
erna
lly fi
nanc
ed
pro
ject
reso
urce
s th
roug
h w
hich
do
nors
pro
vid
e fin
ance
dire
ctly
to
no
n-p
ublic
sec
tor
inst
itutio
ns, s
uch
as lo
cal o
r ex
tern
al N
GO
s; a
nd
(ii) a
id-in
-kin
d, w
hich
is t
houg
ht t
o b
e re
lativ
ely
smal
l.
Ther
e ar
e d
iffer
ence
s b
etw
een
the
figur
es fo
r p
roje
cted
ext
erna
l re
sour
ce fl
ow
s in
the
bud
get
and
for
pro
ject
ed d
isb
urse
men
ts in
d
ono
r re
po
rts.
Thi
s is
par
tly b
ecau
se t
he b
udg
et d
oes
no
t ca
ptu
re
som
e ca
teg
orie
s o
f res
our
ces
(e.g
. dire
ct fi
nanc
ing
). (S
ee “
Evi
den
ce
of c
aptu
re”.
) Ext
erna
l dis
bur
sem
ents
incl
uded
in t
he 2
007
Bud
get
St
atem
ent
(for
2006
act
ual d
isb
urse
men
ts) w
ere
in a
gg
reg
ate
94%
of t
he d
ata
pre
sent
ed in
Jun
e to
the
Co
nsul
tativ
e G
roup
on
dis
bur
sem
ents
for
2006
. Fo
r 20
07, t
he a
gg
reg
ate
gap
bet
wee
n th
e g
ove
rnm
ent
and
do
nors
on
pla
nned
dis
bur
sem
ents
is 9
3%.
In s
om
e ca
ses
ther
e ar
e al
so d
iffer
ence
s in
dis
agg
reg
ated
dat
a b
etw
een
go
vern
men
t (a
s re
flect
ed in
the
bud
get
) and
do
nor
sour
ces,
in p
art
bec
ause
of d
iffer
ence
s in
exp
ecte
d d
isb
urse
men
t tim
ing
s an
d in
the
sta
tus
of s
om
e p
roje
ctio
ns (d
iffer
ence
s in
wha
t in
form
atio
n is
co
nsid
ered
offi
cial
pro
ject
ions
). Th
e M
inis
try
of
Fina
nce
and
Eco
nom
ic P
lann
ing
is u
nder
taki
ng fu
rthe
r an
alys
is o
f th
ese
diff
eren
ces
and
the
exp
lana
tion
for
them
.
Ther
e ha
s b
een
cons
ider
able
pro
gre
ss in
rece
nt y
ears
in c
ove
rag
e o
f do
nor
flow
s in
the
bud
get
, due
to
co
nsid
erab
le w
ork
by
the
join
t g
ove
rnm
ent–
do
nor
inst
itutio
nal i
nfra
stru
ctur
e, s
pec
ifica
lly t
he
Mul
ti-D
ono
r B
udg
et S
upp
ort
Gro
up, a
nd a
s p
art
of C
ons
ulta
tive
Gro
up p
rep
arat
ions
. In
par
ticul
ar, t
he d
iscu
ssio
ns o
n p
lann
ed
dis
bur
sem
ents
of e
xter
nal r
eso
urce
s ha
ve le
d t
o im
pro
vem
ents
in
the
avai
lab
ility
of r
elia
ble
info
rmat
ion
on
exte
rnal
fina
nce
in t
he
bud
get
.
On
par
liam
ent
Parli
amen
t ap
pro
ves
the
Ap
pro
pria
tions
Bill
, whi
ch t
hen
bec
om
es
the
Ap
pro
pria
tions
Act
. The
Ap
pro
pria
tions
Act
co
nsis
ts o
f: (i)
a
sum
mar
y ta
ble
sho
win
g a
pp
rop
riate
d e
xpen
ditu
res
by
MD
A a
nd
sour
ce o
f fun
ds
(Co
nso
lidat
ed F
und
, int
erna
lly g
ener
ated
fund
s,
HIP
C fu
nds,
and
do
nor
fund
s); a
nd (i
i) d
etai
led
exp
end
iture
s b
y he
ad/s
ub-h
ead
and
eco
nom
ic it
em, a
cro
ss t
he fo
ur s
our
ces
of
fund
s.
The
Bud
get
Sta
tem
ent
and
the
sup
ple
men
tary
MD
A v
olu
mes
are
p
rese
nted
to
par
liam
ent
as p
art
of t
he b
udg
et s
ubm
issi
on,
but
ar
e no
t ap
pro
pria
ted
as
such
(i.e
. the
y d
o n
ot
form
par
t o
f the
A
pp
rop
riatio
ns A
ct).
Parli
amen
t’s re
view
of b
udg
eted
ext
erna
l allo
catio
ns in
ag
gre
gat
e re
pre
sent
s a
max
imum
of 9
3% o
f to
tal p
roje
cted
aid
dis
bur
sem
ents
as
rep
ort
ed b
y d
ono
rs.9 B
ut, a
s p
arlia
men
t te
nds
to fo
cus
mai
nly
on
Co
nso
lidat
ed F
und
reso
urce
s, t
here
is li
kely
to
be
less
det
aile
d
scru
tiny
of o
ther
ext
erna
l res
our
ces.
Whi
le t
he p
arlia
men
tary
Fin
ance
Co
mm
ittee
tak
es a
n ac
tive
role
in
scr
utin
isin
g t
he b
udg
et p
rop
osa
ls, i
ts w
ork
is h
amp
ered
by:
(i)
exce
ssiv
e d
etai
l in
bud
get
info
rmat
ion
for
ind
ivid
ual M
DA
s;10
and
(ii
) the
lack
of r
ecen
t tr
end
s (p
revi
ous
yea
r’s a
ctua
l, cu
rren
t ye
ar’s
estim
ates
, and
nex
t ye
ar’s
bud
get
pla
ns) s
how
n si
de
by
sid
e fo
r M
DA
exp
end
iture
s.
Imp
rovi
ng t
he p
rese
ntat
ion
of t
he b
udg
et, (
par
ticul
arly
of M
DA
b
udg
ets)
– t
o s
how
the
link
s b
etw
een
stra
teg
ic p
olic
ies
and
bud
get
al
loca
tions
and
ho
w t
hese
exp
end
iture
s ha
ve e
volv
ed o
ver
time
– w
oul
d g
o a
long
way
to
war
ds
imp
rovi
ng t
he q
ualit
y o
f par
liam
enta
ry
deb
ate.
On
trea
sury
The
follo
win
g a
re p
aid
into
the
Co
nso
lidat
ed F
und
and
dis
bur
sed
to
get
her
with
do
mes
tic re
venu
es: (
i) b
udg
et s
upp
ort
gra
nts;
and
(ii)
bud
get
sup
po
rt c
red
its.
Pro
ject
fund
s (g
rant
s an
d c
red
its) a
re d
isb
urse
d e
ither
dire
ctly
by
the
do
nor
or
thro
ugh
sep
arat
e b
ank
acco
unts
, whi
ch a
re n
ot
linke
d t
o
the
Co
nso
lidat
ed F
und
.
In p
ract
ice,
less
tha
n 35
% o
f ext
erna
l fina
nce
(i.e.
onl
y b
udg
et
sup
po
rt) u
ses
the
trea
sury
and
oth
er n
atio
nal p
roce
dur
es. D
ono
rs
pro
vid
ing
pro
ject
sup
po
rt u
se t
heir
ow
n p
roce
dur
es. T
he m
ain
dev
elo
pm
ent
par
tner
s g
ivin
g b
udg
et s
upp
ort
(tho
se g
ivin
g m
ore
th
an U
S$10
mill
ion
in 2
006)
incl
ude
the
Wo
rld B
ank,
the
UK
, the
A
fDB
, the
EC
, the
Net
herla
nds,
Can
ada
and
Ger
man
y.
The
Min
istr
y o
f Fin
ance
and
Eco
nom
ic P
lann
ing
is w
ork
ing
to
es
tab
lish
an S
TA a
nd fr
om
the
re t
o c
ons
olid
ate
go
vern
men
t re
sour
ces.
Thi
s sy
stem
co
uld
eve
ntua
lly in
clud
e d
ono
r fu
nds,
cu
rren
tly h
eld
in s
epar
ate
ban
k ac
coun
ts.
Som
e d
ono
rs a
re re
po
rted
ly re
luct
ant
to p
ut t
heir
fund
s th
roug
h g
ove
rnm
ent
syst
ems
with
out
str
ong
er fi
duc
iary
co
ntro
ls: t
he la
rge
dev
iatio
ns b
etw
een
bud
get
ed a
nd a
ctua
l exp
end
iture
s an
d w
eak
cap
aciti
es fo
r in
tern
al/e
xter
nal a
udit
are
likel
y to
be
imp
ort
ant
fact
ors
.
87
GH
An
A
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/w
hy n
ot?
)
On
acco
unt
All
fund
s p
aid
into
the
Co
nso
lidat
ed F
und
, sp
ecifi
cally
bud
get
su
pp
ort
(bo
th g
ener
al a
nd s
ecto
r) a
re re
cord
ed a
nd a
cco
unte
d fo
r b
y th
e C
ont
rolle
r an
d A
cco
unta
nt G
ener
al u
sing
a c
om
bin
atio
n o
f a p
aper
-bas
ed s
yste
m a
nd t
he d
evel
op
ing
IFM
IS, k
now
n as
the
B
udg
et a
nd P
ublic
Exp
end
iture
Man
agem
ent
Syst
em. T
hus
bud
get
su
pp
ort
, gen
eral
and
sec
tor,
is a
cco
unte
d fo
r in
thi
s w
ay.
In a
dd
itio
n, s
ecto
rs o
per
atin
g S
WA
ps
pre
par
e co
nso
lidat
ed fi
scal
ac
coun
ts a
nd t
hus
may
be
cons
ider
ed t
o a
cco
unt
for
thei
r re
sour
ces
in a
way
sim
ilar
to t
he C
ons
olid
ated
Fun
d.
An
estim
ated
50%
of t
ota
l aid
reso
urce
s w
ere
acco
unte
d fo
r us
ing
na
tiona
l pro
ced
ures
in 2
006.
Ho
wev
er, a
s th
is fi
gur
e as
sum
es
100%
co
vera
ge
of p
oo
led
fund
s, it
rep
rese
nts
the
max
imum
leve
l p
oss
ible
, and
the
act
ual p
erce
ntag
e is
like
ly t
o b
e co
nsid
erab
ly
low
er.
As
acco
unts
are
co
nso
lidat
ed in
the
STA
, the
sys
tem
co
uld
ev
entu
ally
incl
ude
do
nor
fund
s cu
rren
tly h
eld
in s
epar
ate
ban
k ac
coun
ts, w
hich
wo
uld
the
n p
ote
ntia
lly in
crea
se t
he a
id o
n ac
coun
t.
The
Co
ntro
ller
and
Acc
oun
tant
Gen
eral
co
uld
imp
rove
the
co
mp
rehe
nsiv
enes
s o
f the
ir m
ont
hly
bud
get
exe
cutio
n re
po
rts
by
incl
udin
g d
ono
r p
roje
ct a
id. T
he C
ont
rolle
r an
d A
cco
unta
nt
Gen
eral
ind
icat
es t
hat
such
dat
a ar
e av
aila
ble
by
MD
A, s
o it
wo
uld
b
e p
oss
ible
to
incl
ude
it in
ter
ms
of t
he s
ame
clas
sific
atio
n as
for
do
mes
tic fi
nanc
e.
On
aud
itTh
e G
hana
Aud
it Se
rvic
e au
dits
cen
tral
go
vern
men
t’s a
nnua
l fina
l ac
coun
ts (C
ons
olid
ated
Fun
d) a
nd a
sel
ectio
n o
f iss
ues
for
MD
As
(on
an a
dve
rse
op
inio
n b
asis
), co
ntai
ned
in t
wo
sep
arat
e au
dit
volu
mes
. As
ind
icat
ed u
nder
“O
n ac
coun
t”, t
erm
s o
f ext
erna
l fin
ance
, the
Co
nso
lidat
ed F
und
(and
thu
s th
e au
dit
of t
he a
nnua
l fin
anci
al s
tate
men
t) in
clud
es b
udg
et s
upp
ort
(bo
th m
ulti-
do
nor
bud
get
sup
po
rt a
nd s
ecto
r su
pp
ort
). Th
e vo
lum
e o
n M
DA
s lo
oks
at
sp
ecifi
c ex
pen
ditu
re c
om
plia
nce
issu
es fo
r re
sour
ces
cons
ider
ed
to b
e in
the
pub
lic a
cco
unts
(mai
nly
Co
nso
lidat
ed F
und
, but
als
o
inte
rnal
ly g
ener
ated
fund
s an
d e
xter
nally
fina
nced
exp
end
iture
s) fo
r in
div
idua
l MD
As.
As
par
t o
f the
mul
ti-d
ono
r b
udg
et s
upp
ort
co
mm
on
fram
ewo
rk
agre
emen
t, e
xter
nal a
udits
, ter
med
“sp
ecia
l flo
ws
aud
its”,
may
be
carr
ied
out
; the
se a
re c
arrie
d o
ut b
y p
rivat
e se
cto
r au
dito
rs. T
he fi
rst
of t
hese
aud
it re
po
rts
cam
e o
ut in
200
4, a
nd t
he s
eco
nd is
cur
rent
ly
in d
raft
form
. Po
ole
d fu
nds
are
req
uire
d t
o b
e au
dite
d “
by
the
Gha
na A
udit
Serv
ice”
(i.e
. eith
er b
y th
e G
hana
Aud
it Se
rvic
e its
elf o
r b
y an
aud
itor
app
oin
ted
by
the
Gha
na A
udit
Serv
ice)
and
are
pai
d
for
with
ext
erna
l fun
ds.
Whi
le b
oth
aud
it vo
lum
es a
re s
ent
to p
arlia
men
t an
d a
re
sub
seq
uent
ly a
vaila
ble
to
inte
rest
ed p
artie
s, a
lack
of r
eso
urce
s m
eans
tha
t th
ey o
nly
cove
r a
(rela
tivel
y sm
all)
po
rtio
n o
f the
Gha
na
Aud
it Se
rvic
e’s
man
dat
e, w
hich
incl
udes
ann
ual a
udits
of a
ll M
DA
s,
met
rop
olit
an, m
unic
ipal
, and
dis
tric
t as
sem
blie
s, s
tatu
tory
fund
s an
d
pub
lic c
orp
ora
tions
.
Taki
ng t
he d
efini
tion
of “
on
aud
it” t
o in
clud
e th
ose
fund
s w
hich
the
G
hana
Aud
it Se
rvic
e is
man
dat
ed t
o a
udit,
dire
ctly
or
thro
ugh
an
app
oin
ted
/ap
pro
ved
ext
erna
l aud
itor,
this
wo
uld
incl
ude:
(i) t
hose
ex
tern
al re
sour
ces
whi
ch a
re p
art
of t
he C
ons
olid
ated
Fun
d (i
.e.
bud
get
sup
po
rt);
and
(ii)
po
ole
d fu
nds
who
se a
udito
r is
ap
pro
ved
b
y th
e A
udito
r G
ener
al. U
sing
dat
a fo
r 20
06, i
t is
est
imat
ed t
hat
up t
o 5
0% o
f to
tal a
id re
sour
ces
wer
e ac
coun
ted
for
usin
g n
atio
nal
pro
ced
ures
. Ho
wev
er, t
his
figur
e re
pre
sent
s th
e m
axim
um le
vel
po
ssib
le, a
s it
assu
mes
100
% c
ove
rag
e o
f po
ole
d fu
nds,
and
the
ac
tual
per
cent
age
is li
kely
to
be
cons
ider
ably
low
er.
Wea
knes
ses
in a
udit,
bo
th in
tern
al a
nd e
xter
nal,
can
und
erm
ine
the
effe
ctiv
enes
s o
f ove
rsig
ht o
f the
use
of p
ublic
fund
s. S
ince
the
in
tern
al a
udit
func
tion
is re
lativ
ely
new
, it
cann
ot
be
exp
ecte
d t
o b
e o
per
atin
g a
t m
axim
um e
ffect
iven
ess.
Whi
le t
he q
ualit
y o
f ext
erna
l au
dits
sub
mitt
ed t
o p
arlia
men
t is
reas
ona
ble
, the
tim
elin
ess
and
co
vera
ge
of t
hese
aud
its is
wea
ker.
Lim
ited
reso
urce
s an
d c
apac
ities
(e
ven
with
the
pra
ctic
e o
f sup
ple
men
ting
Gha
na A
udit
Serv
ice
staf
f with
priv
ate
sect
or
aud
itors
to
co
ver
spec
ific
aud
its in
the
se
rvic
e’s
man
dat
e) p
reve
nt t
he G
hana
Aud
it Se
rvic
e fr
om
co
verin
g
its s
tatu
tory
man
dat
e in
full.
In p
artic
ular
, mo
st o
f the
acc
oun
ts a
nd
finan
cial
sta
tem
ents
for
pub
lic b
oar
ds,
co
rpo
ratio
ns a
nd s
tatu
tory
in
stitu
tions
lag
beh
ind
by
seve
ral y
ears
. The
aud
its o
f MD
A a
cco
unts
te
nd t
o b
e ca
rrie
d o
ut s
elec
tivel
y, b
ased
on
an a
dve
rse
op
inio
n.
Follo
w-u
p o
n th
e fin
din
gs
of t
he A
udito
r G
ener
al, p
artic
ular
ly w
here
th
e ac
coun
ts a
re q
ualifi
ed, n
eed
s to
str
eng
then
ed.
Cap
acity
co
nstr
aint
s w
ill b
e d
ifficu
lt to
ove
rco
me
in t
he s
hort
te
rm, p
artic
ular
ly g
iven
the
lim
ited
reso
urce
s av
aila
ble
for
exte
rnal
au
dit
and
the
wid
e m
and
ate.
The
ava
ilab
le re
sour
ces
need
to
be
prio
ritis
ed t
o w
here
the
gre
ates
t p
oss
ible
gai
ns a
nd m
axim
um
imp
act
can
be
achi
eved
.
88
GH
An
A
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/w
hy n
ot?
)
On
rep
ort
Rep
ort
ing
on
the
bud
get
co
nsis
ts o
f mo
nthl
y ex
pen
ditu
re re
po
rts
fro
m t
he C
ont
rolle
r an
d A
cco
unta
nt G
ener
al, a
nd t
he a
nnua
l fin
anci
al s
tate
men
t. A
s in
dic
ated
und
er “
On
acco
unt”
, the
se c
ove
r th
e C
ons
olid
ated
Fun
d (w
hich
co
mp
rises
go
vern
men
t d
om
estic
ex
pen
ditu
res
for
MD
As,
and
HIP
C fu
nds
by
MD
A).
Thus
, the
y in
clud
e b
udg
et s
upp
ort
gra
nts
and
cre
dits
.
In a
dd
itio
n, t
he 2
007
Bud
get
Sta
tem
ent
incl
udes
info
rmat
ion
on
the
dis
bur
sem
ent
of e
xter
nal r
eso
urce
s (p
roje
ct/p
rog
ram
me
assi
stan
ce)
for
the
pre
vio
us y
ear
in a
num
ber
of a
pp
end
ices
, inc
lud
ing
: (i)
dis
bur
sem
ents
of p
roje
ct lo
ans
and
gra
nts
by
cred
itor/
do
nor
and
M
DA
; (ii)
det
ails
of p
roje
ct/p
rog
ram
me
loan
s b
y d
ono
r, tit
le o
f p
roje
ct, d
ate
of l
oan
, orig
inal
cur
renc
y, a
mo
unt
in o
rigin
al c
urre
ncy
and
US$
eq
uiva
lent
, mat
urity
, int
eres
t ra
te, a
nd p
erce
ntag
e g
rant
el
emen
t; (i
ii) d
etai
ls o
f pro
ject
/pro
gra
mm
e g
rant
s b
y d
ono
r, tit
le,
dat
e, o
rigin
al c
urre
ncy,
and
am
oun
t in
orig
inal
cur
renc
y an
d U
S$
equi
vale
nt.
Taki
ng “
on
rep
ort
” to
incl
ude
dat
a o
n ac
tual
dis
bur
sem
ents
p
rovi
ded
in: (
i) C
ont
rolle
r an
d A
cco
unta
nt G
ener
al b
udg
et
exec
utio
n re
po
rts
(Co
nso
lidat
ed F
und
, sp
ecifi
cally
bud
get
sup
po
rt);
(ii) t
he fo
llow
ing
yea
r’s B
udg
et S
tate
men
t (t
he m
ajo
rity
of p
roje
ct
fund
s); a
nd (i
ii) M
DA
co
nso
lidat
ed re
po
rts
fro
m p
oo
led
fund
s. (I
n 20
06 a
roun
d 9
3% o
f ext
erna
l aid
reso
urce
s w
ere
cap
ture
d in
fisc
al
rep
ort
s.)
In re
cent
yea
rs, c
ons
ider
able
pro
gre
ss h
as b
een
mad
e in
cap
turin
g
mo
re in
form
atio
n o
n d
ono
r re
sour
ces
in fi
scal
rep
ort
s. In
ord
er t
o
bro
aden
the
bud
get
’s re
po
rtin
g, t
he g
ove
rnm
ent
is a
lso
tak
ing
st
eps
to in
clud
e in
form
atio
n o
n al
l lo
ans
and
gra
nt a
gre
emen
ts in
th
e B
udg
et S
tate
men
t. A
co
mp
aris
on
of t
he B
udg
et S
tate
men
ts
for
2004
, 200
5 an
d 2
006
ind
icat
es s
igni
fican
tly g
reat
er c
ove
rag
e o
f d
ono
r flo
ws
in re
cent
yea
rs, p
artic
ular
ly g
rant
s. P
rog
ress
is b
eing
m
ade
with
exe
cutio
n re
po
rts.
In t
he 2
005
Ann
ual R
epo
rt a
nd
Fina
ncia
l Sta
tem
ent,
dis
bur
sem
ents
of g
rant
s b
y in
div
idua
l do
nor
wer
e in
clud
ed fo
r th
e fir
st t
ime.
Wo
rk is
als
o c
ont
inui
ng w
ith t
he C
ont
rolle
r an
d A
cco
unta
nt G
ener
al
to in
clud
e flo
ws
of e
xter
nal a
ssis
tanc
e in
reg
ular
(qua
rter
ly/m
ont
hly)
C
ont
rolle
r an
d A
cco
unta
nt G
ener
al re
po
rts.
Sin
ce d
ono
r fu
nds
are
onl
y p
artia
lly re
po
rted
(gra
nts
incl
uded
in t
he C
ont
rolle
r an
d
Acc
oun
tant
Gen
eral
rep
ort
rep
rese
nt a
bo
ut h
alf o
f the
am
oun
t es
timat
ed b
y th
e A
id a
nd D
ebt
Man
agem
ent
Div
isio
n), t
og
ethe
r w
ith C
ons
olid
ated
Fun
d fu
nds,
reta
ined
inte
rnal
ly g
ener
ated
re
sour
ces,
and
HIP
C re
sour
ces,
the
re is
no
t a
sing
le s
our
ce fo
r re
po
rtin
g o
n to
tal g
ove
rnm
ent
fisca
l op
erat
ions
. Thi
s ha
mp
ers
effe
ctiv
e o
vers
ight
of t
he u
se o
f pub
lic re
sour
ces.
Mo
nthl
y an
d
qua
rter
ly d
isb
urse
men
t in
form
atio
n, p
artic
ular
ly fo
r p
roje
ct/
pro
gra
mm
e lo
ans,
is a
vaila
ble
to
Min
istr
y o
f Fin
ance
and
Eco
nom
ic
Plan
ning
and
thu
s co
uld
be
rep
ort
ed o
n in
the
Co
ntro
ller
and
A
cco
unta
nt G
ener
al re
po
rts.
The
dat
a in
fras
truc
ture
is in
pla
ce t
o
pro
duc
e co
mp
rehe
nsiv
e re
po
rts;
ho
wev
er, t
here
nee
ds
to b
e a
clea
r co
mm
itmen
t fr
om
bo
th d
evel
op
men
t p
artn
ers
and
the
go
vern
men
t to
pro
vid
e an
d in
clud
e th
is in
form
atio
n. T
his
is a
n ar
ea t
hat
Co
ntro
ller
and
Acc
oun
tant
Gen
eral
is c
urre
ntly
wo
rkin
g t
o im
pro
ve.
With
thi
s d
evel
op
men
t, t
he C
ont
rolle
r an
d A
cco
unta
nt G
ener
al
rep
ort
s ar
e ex
pec
ted
, ove
r tim
e, t
o e
nco
mp
ass
reve
nue
and
ex
pen
ditu
re in
form
atio
n o
n ex
tern
ally
fund
ed p
roje
cts
(incl
udin
g
do
nor
pro
ject
aid
) and
reta
ined
inte
rnal
ly g
ener
ated
fund
s.
Ther
e ar
e w
eakn
esse
s in
info
rmat
ion
flow
s b
oth
bet
wee
n th
e g
ove
rnm
ent
and
do
nors
and
acr
oss
go
vern
men
t ai
d m
anag
emen
t in
stitu
tions
the
mse
lves
, lea
din
g t
o d
ifficu
lties
in re
conc
iliat
ions
b
etw
een
aid
flo
ws
cap
ture
d a
nd/o
r re
po
rted
by
IPs,
the
Aid
and
D
ebt
Man
agem
ent
Div
isio
n, E
xter
nal R
eso
urce
s M
anag
emen
t an
d
MD
As.
Thi
s is
due
par
tly t
o w
eakn
esse
s in
co
-ord
inat
ion
acro
ss
div
isio
ns a
nd a
cro
ss M
DA
s, a
nd p
artly
to
diff
eren
ces
in t
he s
tatu
s o
f d
efini
tions
of a
ssis
tanc
e, p
artic
ular
ly w
hen
a p
roje
ctio
n is
co
nsid
ered
to
be
offi
cial
.
89
MA
Li
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/W
hy n
ot?
)
On
pla
nG
BS
– In
the
ory
, GB
S is
pla
nned
in t
he s
ame
way
as
do
mes
tic
reso
urce
s. B
ut a
s th
ey a
re n
ot
refle
cted
in M
ali’s
bud
get
d
ocu
men
tatio
n, it
is d
ifficu
lt to
det
erm
ine
how
the
se fu
nds
are
trea
ted
in p
ract
ice.
Sect
or b
udg
et s
upp
ort –
In t
heo
ry, s
ecto
r b
udg
et s
upp
ort
sho
uld
al
so b
e p
lann
ed in
the
sam
e w
ay a
s d
om
estic
reso
urce
s, w
ith t
he
diff
eren
ce t
hat
the
rele
vant
sec
tor
min
istr
ies
are
sup
po
sed
to
ac
hiev
e p
artic
ular
resu
lts in
acc
ord
ance
with
the
acc
om
pan
ying
se
cto
r p
olic
y m
atrix
. In
pra
ctic
e in
Mal
i, th
ese
reso
urce
s ar
e tr
eate
d
diff
eren
tly t
o d
om
estic
reso
urce
s: in
effe
ct, t
hey
are
trea
ted
as
if th
ey w
ere
pro
ject
reso
urce
s. T
hey
are
pla
nned
sep
arat
ely,
and
the
ir us
e is
sep
arat
ely
iden
tified
(as
line
item
s) a
nd t
rack
ed in
the
bud
get
an
d in
bud
get
exe
cutio
n re
po
rts.
Fur
ther
mo
re, t
hey
are
dire
cted
at
par
ticul
ar p
urp
ose
s (e
.g. t
rain
ing
wo
rksh
op
s), w
hich
may
or
may
no
t b
e ho
w t
hey
wo
uld
hav
e b
een
used
had
the
y b
een
pla
nned
to
get
her
with
the
rest
of t
he m
inis
try’
s re
sour
ces.
Proj
ect a
ssis
tanc
e –
With
the
sep
arat
ion
bet
wee
n th
e re
curr
ent
and
in
vest
men
t p
lann
ing
pro
cess
es, p
roje
ct a
ssis
tanc
e re
sour
ces,
as
par
t o
f the
Sp
ecia
l Inv
estm
ent B
udg
et (B
udg
et S
péc
ial d
’Inve
stis
sem
ent),
ar
e p
lann
ed s
epar
atel
y fr
om
recu
rren
t re
sour
ces.
To
the
ext
ent
that
in
vest
men
t ex
pen
ditu
res
are
finan
ced
by
exte
rnal
reso
urce
s (t
he
vast
maj
orit
y ar
e), t
hese
ap
pea
r in
the
Sp
ecia
l Inv
estm
ent
Bud
get
, an
d t
he p
lann
ing
/man
agem
ent
arra
ngem
ents
ten
d t
o b
e p
lann
ed
in c
onj
unct
ion
with
the
rele
vant
do
nor(
s) (r
athe
r th
an a
s p
art
of a
p
rog
ram
mat
ic a
pp
roac
h).
Ass
umin
g t
hat
in p
ract
ice
GB
S is
ind
istin
gui
shab
le fr
om
do
mes
tic
reso
urce
s, a
roun
d 2
2% o
f to
tal a
id d
isb
urse
men
ts11
are
est
imat
ed t
o
be
incl
uded
in n
atio
nal p
lann
ing
pro
ced
ures
.
The
Spec
ial I
nves
tmen
t B
udg
et c
ont
ains
det
ails
of e
xter
nally
fin
ance
d p
roje
cts
by
min
istr
y an
d b
y p
rog
ram
me.
Whi
le t
he
info
rmat
ion
is re
aso
nab
ly c
om
pre
hens
ive
for
pro
ject
s fin
ance
d b
y ex
tern
al lo
ans,
it is
so
mew
hat
less
so
for
pro
ject
s fin
ance
d b
y g
rant
s.
In p
ract
ice,
the
bud
get
inco
rpo
rate
s m
ainl
y kn
ow
n co
mm
itmen
ts
for
pro
gra
mm
es c
ove
red
by
und
erly
ing
go
vern
men
t/d
ono
r ag
reem
ents
. Op
erat
ions
fund
ed b
y g
rant
s an
d c
arrie
d o
ut d
irect
ly
bet
wee
n d
ono
rs, N
GO
s (re
latin
g t
o g
ove
rnm
ent
op
erat
ions
) and
b
enefi
ciar
ies,
with
out
pas
sing
thr
oug
h th
e M
inis
try
of E
cono
my
and
Fin
ance
(Min
istè
re d
e l’E
cono
mie
et d
es F
inan
ces)
, ten
d n
ot
to
be
cap
ture
d. T
hese
are
est
imat
ed b
y th
e M
inis
try
of E
cono
my
and
Fi
nanc
e to
be
rela
tivel
y sm
all,
but
may
be
mo
re t
han
10%
of t
he
tota
l val
ue o
f gra
nts.
In t
erm
s o
f mo
ves
tow
ard
s p
lann
ing
bud
get
ary
reso
urce
s as
a
who
le, t
he p
olic
y b
asis
for
the
bud
get
has
imp
rove
d in
rece
nt
year
s th
roug
h: (i
) the
intr
od
uctio
n o
f the
Med
ium
-Ter
m B
udg
etar
y Fr
amew
ork
(Cad
re B
udg
étai
re à
Moy
en T
erm
e), (
ii) t
he b
egin
ning
s o
f the
pre
par
atio
n o
f sec
tor
exp
end
iture
str
ateg
ies,
and
(iii)
the
p
rep
arat
ion
of b
udg
et p
rog
ram
mes
and
incl
usio
n in
an
anne
x to
the
st
ate
bud
get
(Bud
get
d’E
tat).
Stra
teg
ic b
udg
etin
g is
in t
he e
arly
sta
ges
. Onl
y he
alth
and
ed
ucat
ion
have
a fu
ll m
ediu
m-t
erm
sec
tor
stra
teg
y. T
here
are
lim
ited
cap
aciti
es a
t lin
e m
inis
try
leve
l fo
r co
stin
g a
ctiv
ities
and
for
linki
ng p
olic
y o
bje
ctiv
es a
nd re
sour
ce a
lloca
tions
.
At
the
sam
e tim
e, t
he la
ck o
f ear
ly C
abin
et in
volv
emen
t in
the
se
ttin
g o
f str
ateg
ic b
udg
et p
aram
eter
s an
d li
mite
d a
naly
tical
ca
pac
ities
to
co
st s
trat
egie
s co
nstr
ain
the
go
vern
men
t’s a
bili
ty t
o
use
the
bud
get
as
a st
rate
gic
to
ol t
o im
ple
men
t its
med
ium
-ter
m
po
licie
s, in
clud
ing
tho
se in
the
PR
SP (C
adre
Str
atég
ique
pou
r la
Cro
issa
nce
et la
Réd
uctio
n d
e la
Pau
vret
é).
90
MA
Li
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/W
hy n
ot?
)
On
bud
get
The
annu
al B
udg
et L
aw (L
oi d
e Fi
nanc
es) i
nclu
des
tw
o t
ypes
of
exte
rnal
aid
reso
urce
s: (i
) ext
erna
lly fu
nded
pro
ject
reso
urce
s (b
oth
lo
an a
nd g
rant
-fina
nced
); an
d (i
i) se
cto
r b
udg
et s
upp
ort
(cur
rent
ly,
all g
rant
-fina
nced
). D
etai
ls o
f pro
ject
reso
urce
s (e
xpen
ditu
re
typ
e, a
ctiv
ities
fina
nced
, and
sec
tor)
are
incl
uded
in t
he S
pec
ial
Inve
stm
ent
Bud
get
, whi
le ju
st t
he t
ota
ls fo
r ex
tern
al g
rant
and
loan
-fin
ance
d p
roje
ct a
ssis
tanc
e ar
e sh
ow
n in
the
mai
n an
nual
Bud
get
La
w d
ocu
men
t. In
the
wid
er s
tate
bud
get
, do
mes
tic a
nd e
xter
nally
fin
ance
d S
pec
ial I
nves
tmen
t B
udg
et re
sour
ces
are
sho
wn
by
sect
or
min
istr
y.
The
bud
get
do
cum
ents
– t
he s
tate
bud
get
(inc
lud
ing
the
ann
ual
Bud
get
Law
), an
d t
he a
cco
mp
anyi
ng S
pec
ial I
nves
tmen
t B
udg
et
– d
o n
ot
incl
ude:
(i) e
xter
nally
fina
nced
pro
ject
reso
urce
s in
whi
ch
do
nors
pro
vid
e fin
ance
dire
ctly
to
no
n-p
ublic
sec
tor
inst
itutio
ns,
such
as
NG
Os
(eith
er d
om
estic
or
exte
rnal
NG
Os)
; (ii)
gra
nt-
finan
ced
GB
S; a
nd (i
ii) c
red
it /l
oan
-fina
nced
bud
get
sup
po
rt (a
t p
rese
nt, o
nly
GB
S –
not
sect
or
bud
get
sup
po
rt –
is fi
nanc
ed b
y cr
edits
/lo
ans)
.
That
is, t
he b
udg
et d
oes
no
t ac
coun
t fo
r re
sid
ual a
id (a
ny a
id
out
sid
e b
udg
et s
upp
ort
and
pro
ject
aid
). N
or
do
es it
incl
ude
spen
din
g o
utsi
de
Mal
i, o
r ai
d t
hat
is n
ot
del
iver
ed b
y th
e g
ove
rnm
ent
sect
or
(e.g
. by
NG
Os)
.
In te
rms
of c
omp
lete
ness
of c
over
age
of b
udg
et s
upp
ort:
• In
the
annu
al B
udg
et L
aw, g
rant
-fina
nced
sec
tor b
udg
et s
upp
ort
(in p
ract
ice,
all
of s
ecto
r bud
get
sup
por
t) is
sho
wn
as a
sou
rce
of
fund
ing
(und
er re
ceip
ts).
The
pla
nned
use
of w
hat i
s te
rmed
sec
tor
bud
get
sup
por
t is
show
n (a
s ea
rmar
ked
fund
ing
) und
er in
div
idua
l se
ctor
min
istr
ies.
Gra
nt-fi
nanc
ed G
BS
is n
ot in
clud
ed a
s a
sour
ce o
f fu
ndin
g (u
nder
rece
ipts
). Th
e us
e of
thes
e re
sour
ces
is n
ot in
clud
ed
und
er e
xpen
ditu
res.
But
they
are
incl
uded
in e
x p
ost r
epor
ting
, and
th
ere
is th
us a
diff
eren
ce b
etw
een
the
app
rop
riate
d b
udg
et a
nd
the
exec
uted
bud
get
. Thi
s si
tuat
ion
is c
ater
ed fo
r in
the
Org
anic
B
udg
et L
aw: i
f the
rece
ipts
turn
out
to b
e m
ore
than
exp
ecte
d (e
.g.
thro
ugh
add
ition
al g
rant
-fina
nced
bud
get
sup
por
t) th
e M
inis
ter
of E
cono
my
and
Fin
ance
can
incr
ease
the
open
cre
dits
by
dec
ree
up to
the
amou
nt o
f the
ad
diti
onal
rece
ipts
. But
the
fact
that
th
ese
add
ition
al re
sour
ces
are
not a
pp
rop
riate
d b
y th
e N
atio
nal
Ass
emb
ly (e
.g. t
hrou
gh
a su
pp
lem
enta
ry b
udg
et) u
nder
min
es th
e tr
ansp
aren
cy a
nd a
ccou
ntab
ility
of t
heir
pla
nned
use
– th
ere
is n
o
exp
licit
(ext
erna
l) m
echa
nism
to e
nsur
e th
at th
e p
riorit
isat
ion
of
thes
e re
sour
ces
is in
line
with
ove
rall
gov
ernm
ent o
bje
ctiv
es.
• C
red
it-fin
ance
d G
BS
is in
clud
ed im
plic
itly
as a
sou
rce
of fi
nanc
ing
th
e d
efici
t, al
ong
with
rese
rves
– a
lthou
gh
the
exac
t sp
lit b
etw
een
rese
rves
and
bud
get
sup
por
t is
not g
iven
, nor
are
det
ails
of t
he
bud
get
sup
por
t. Th
e us
e of
thes
e re
sour
ces
is im
plic
itly
incl
uded
as
par
t of g
ener
al e
xpen
ditu
res
and
, in
line
with
the
norm
for b
udg
et
sup
por
t, th
e fu
nds
are
not e
arm
arke
d.
A c
omp
aris
on o
f the
info
rmat
ion
in th
e an
nual
Bud
get
Law
with
that
fr
om d
onor
s sh
ows
that
in 2
005
exte
rnal
dis
bur
sem
ents
incl
uded
in
the
bud
get
wer
e 40
% lo
wer
than
the
info
rmat
ion
on d
isb
urse
men
ts
furn
ishe
d b
y d
onor
s, a
nd m
ore
than
25%
low
er in
200
6. F
or
2007
, the
gap
bet
wee
n th
e g
over
nmen
t and
don
ors
on p
lann
ed
dis
bur
sem
ents
is 3
3%. T
he a
ctua
l gap
for i
ndiv
idua
l pro
ject
s co
uld
b
e la
rger
.
Sect
or b
udg
et s
upp
ort s
how
n b
y m
inis
try
in th
e st
ate
bud
get
is
clas
sifie
d in
acc
ord
ance
with
the
gov
ernm
ent’s
cla
ssifi
catio
n sy
stem
. Pr
ojec
ted
exp
end
iture
s in
the
Spec
ial I
nves
tmen
t Bud
get
are
sh
own
in a
ccor
dan
ce w
ith it
s ow
n cl
assi
ficat
ion
(pro
ject
num
ber
s).
Info
rmat
ion
on d
isb
urse
men
ts o
f the
pre
viou
s ye
ar’s
sect
or b
udg
et
sup
por
t and
pro
ject
ass
ista
nce
is in
clud
ed in
the
ann
ual B
udg
et
Law
/sta
te b
udg
et a
nd t
he S
pec
ial I
nves
tmen
t B
udg
et, r
esp
ectiv
ely.
Tim
elin
ess
of in
form
atio
n on
pla
nned
dis
bur
sem
ents
is a
n is
sue.
In
line
with
the
bud
get
cyc
le, r
elia
ble
info
rmat
ion
on p
roje
cted
ext
erna
l re
sour
ces
need
s to
be
rece
ived
by
May
, but
oft
en is
not
rece
ived
un
til S
epte
mb
er o
r lat
er. T
his
app
lies
bot
h to
bud
get
sup
por
t and
to
pro
ject
aid
, and
can
ad
vers
ely
affe
ct b
oth
the
com
ple
tene
ss o
f the
in
form
atio
n an
d th
e ef
ficie
ncy
of th
e b
udg
et p
roce
ss.
The
mai
n d
irect
reas
on
for
the
less
co
mp
rehe
nsiv
e ca
ptu
re o
f ex
tern
al d
isb
urse
men
ts is
the
exc
lusi
on
of G
BS
(bo
th g
rant
- an
d
loan
-fina
nced
).
It is
po
ssib
le t
hat
the
Min
istr
y o
f Eco
nom
y an
d F
inan
ce re
ceiv
es
info
rmat
ion
on
gra
nt-fi
nanc
ed s
upp
ort
to
o la
te t
o b
e in
clud
ed in
the
b
udg
et d
ocu
men
ts.
At
the
sam
e tim
e, p
roje
ct a
id is
no
t p
red
icta
ble
, and
its
timin
g
is n
ot
alig
ned
with
the
bud
get
pro
cess
, no
r is
the
tim
ing
for
dis
bur
sem
ents
alig
ned
am
ong
do
nors
As
for
loan
-fina
nced
GB
S, t
he a
nnua
l Bud
get
Law
do
es n
ot
incl
ude
a b
reak
do
wn
of t
he c
om
po
nent
s o
f defi
cit
finan
cing
as
a ru
le.
91
MA
Li
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/W
hy n
ot?
)
On
par
liam
ent
The
Nat
iona
l Ass
emb
ly v
ote
s o
n th
e ap
pro
pria
tions
in t
he a
nnua
l B
udg
et L
aw, w
hich
incl
udes
det
aile
d in
form
atio
n o
n se
cto
r b
udg
et
sup
po
rt (b
oth
rece
ipts
and
pro
gra
mm
ed e
xpen
ditu
re u
se),
and
ag
gre
gat
e fig
ures
for
exte
rnal
ly fi
nanc
ed p
roje
ct a
ssis
tanc
e. T
here
is
no
info
rmat
ion
on
GB
S in
the
ann
ual B
udg
et L
aw. A
cco
mp
anyi
ng
info
rmat
ion
for
the
Nat
iona
l Ass
emb
ly in
clud
es d
etai
ls o
n p
roje
ct
exp
end
iture
s.
The
annu
al B
udg
et L
aw in
clud
es t
he n
atio
nal b
udg
et (B
udg
et
Nat
iona
l) fo
r th
e lin
e m
inis
trie
s, t
he re
gio
nal b
udg
ets
(Bud
get
s Ré
gio
naux
), th
e su
bsi
die
s to
pub
lic e
nter
pris
es a
nd e
arm
arke
d
fund
s (B
udg
ets
Ann
exes
, Com
pte
s, e
t Fon
ds
Spéc
iaux
), an
d t
he
inve
stm
ent
bud
get
(Sp
ecia
l Inv
estm
ent
Bud
get
). A
cco
mp
anyi
ng t
he
bud
get
as
info
rmat
ion
is t
he M
ediu
m-T
erm
Bud
get
ary
Fram
ewo
rk
and
the
bud
get
pro
gra
mm
es, f
oun
d in
the
sta
te b
udg
et.
The
Nat
iona
l Ass
emb
ly a
lso
ap
pro
ves
the
annu
al B
udg
et E
xecu
tion
Law
(Loi
de
Règ
lem
ent).
The
allo
catio
ns o
f pro
ject
ass
ista
nce
and
sec
tor
bud
get
sup
po
rt
in t
he a
nnua
l Bud
get
Law
whi
ch t
he N
atio
nal A
ssem
bly
scr
utin
ises
re
pre
sent
aro
und
80%
of t
ota
l aid
dis
bur
sem
ents
.12
The
Nat
iona
l Ass
emb
ly (t
hro
ugh
the
Fina
nce
Co
mm
issi
on)
has
at
leas
t tw
o m
ont
hs t
o s
crut
inis
e th
e an
nual
Bud
get
Law
. The
re a
re
conc
erns
ab
out
the
qua
lity
of t
he N
atio
nal A
ssem
bly
’s d
ebat
e, w
hich
is
focu
sed
rela
tivel
y na
rro
wly
on
the
annu
al li
ne it
em a
pp
rop
riatio
ns.
The
Fina
nce
Co
mm
issi
on
do
es n
ot
app
ear
to e
xam
ine
in d
etai
l the
b
road
er s
ecto
ral c
ont
ext,
incl
udin
g t
he a
pp
rop
riate
ness
of,
and
co
nsis
tenc
y w
ith, t
he o
bje
ctiv
es a
nd a
ctiv
ities
in t
he M
ediu
m-T
erm
B
udg
etar
y Fr
amew
ork
and
the
bud
get
pro
gra
mm
es. T
he e
xten
t to
whi
ch t
he c
om
mis
sio
n ex
amin
es t
he p
lann
ed u
se o
f ext
erna
l re
sour
ces
is n
ot
clea
r.
Whi
le t
he N
atio
nal A
ssem
bly
is g
iven
suf
ficie
nt t
ime
to d
ebat
e th
e d
raft
ann
ual B
udg
et L
aw, i
ts t
echn
ical
fina
ncia
l man
agem
ent
cap
aciti
es a
re in
suffi
cien
t to
exp
lore
in d
etai
l the
use
of a
ll re
sour
ces
for
sect
or
pro
gra
mm
es n
or
the
imp
licat
ions
for
mee
ting
sec
tora
l o
bje
ctiv
es.
Acc
ord
ing
to
the
PE
FA a
sses
smen
t, b
esid
es t
he d
elay
s in
rece
ivin
g
the
annu
al B
udg
et E
xecu
tion
Law
, lim
ited
cap
aciti
es fo
r re
view
ing
au
dite
d a
cco
unts
co
ntrib
ute
to w
eakn
esse
s in
the
Nat
iona
l A
ssem
bly
’s ex
po
st re
view
.
The
Nat
iona
l Ass
emb
ly’s
scru
tiny
of t
he a
nnua
l Bud
get
Exe
cutio
n La
w h
as b
een
hind
ered
by
del
ays
in t
he s
ubm
issi
on
of c
ertifi
ed
acco
unts
by
the
Sect
ion
des
Co
mp
tes,
an
aud
it in
stitu
tion
whi
ch is
a
bra
nch
of t
he S
upre
me
Co
urt.
Ho
wev
er, t
he re
cent
cle
arin
g o
f the
in
stitu
tion’
s b
ackl
og
will
po
tent
ially
allo
w t
he N
atio
nal A
ssem
bly
to
ex
erci
se it
s ex
tern
al s
crut
iny
role
mo
re a
ctiv
ely
Mo
re b
road
ly, i
t is
po
ssib
le t
hat,
with
the
rela
tive
yout
h o
f the
co
untr
y’s
dem
ocr
acy,
the
re is
stil
l lim
ited
dem
and
fro
m t
he p
ublic
fo
r g
reat
er t
rans
par
ency
and
acc
oun
tab
ility
on
pub
lic e
xpen
ditu
res.
On
trea
sury
Pub
lic e
xpen
ditu
res
are
chan
nelle
d t
hro
ugh
two
typ
es o
f acc
oun
t: (i
) th
ose
acc
oun
ts li
nked
to
(and
co
ntro
lled
by)
the
STA
at
the
Cen
tral
B
ank
of W
est
Afr
ican
Sta
tes
(Ban
que
Cen
tral
e d
es É
tats
de
l’Afr
ique
d
e l’O
uest
) – t
hese
incl
ude
do
mes
tic fu
nds,
GB
S an
d s
ecto
r b
udg
et
sup
po
rt; a
nd (i
i) th
ose
acc
oun
ts in
co
mm
erci
al b
anks
op
ened
and
m
anag
ed b
y PI
Us
or
oth
er g
ove
rnm
ent
agen
cies
out
sid
e o
f the
tr
easu
ry’s
cont
rol –
the
se in
clud
e p
roje
ct re
sour
ces.
In g
ener
al, o
nly
bud
get
sup
po
rt o
per
atio
ns m
ay b
e co
nsid
ered
to
p
ass
thro
ugh
the
trea
sury
sys
tem
. Out
sid
e o
f bud
get
sup
po
rt (e
.g.
Pove
rty
Red
uctio
n Su
pp
ort
Cre
dits
and
IMF
sup
po
rt) t
here
is li
mite
d
use
of t
he t
reas
ury
for
exte
rnal
fina
nce.
The
po
ole
d fu
nds
use
sep
arat
e g
ove
rnm
ent-
des
igna
ted
acc
oun
ts, b
ut t
hese
are
no
t lin
ked
to
tre
asur
y ac
coun
ts. P
roje
ct fu
nds,
man
aged
thr
oug
h PI
Us,
may
use
g
ove
rnm
ent
or
non-
go
vern
men
t ac
coun
ts, b
ut t
hese
are
no
t lin
ked
to
tre
asur
y ac
coun
ts.
In 2
006,
aro
und
33%
of e
xter
nal fi
nanc
e (i.
e. o
nly
bud
get
sup
po
rt)
used
the
tre
asur
y an
d o
ther
nat
iona
l pro
ced
ures
– a
n in
crea
se
fro
m 2
6% in
200
5. D
ono
rs p
rovi
din
g p
roje
ct s
upp
ort
use
the
ir o
wn
pro
ced
ures
. The
mai
n d
evel
op
men
t p
artn
ers
giv
ing
bud
get
sup
po
rt
(bas
ed o
n th
ose
giv
ing
mo
re t
han
US$
10 m
illio
n in
200
6) in
clud
e th
e A
fDB
, the
EC
, Fra
nce,
the
Net
herla
nds,
and
Sw
eden
(with
Can
ada
fro
m 2
007)
.
The
lack
of p
red
icta
bili
ty o
f the
dis
bur
sem
ents
of s
ecto
r b
udg
et
sup
po
rt b
y so
me
do
nors
(par
ticul
arly
with
the
var
iab
le t
ranc
he) i
s a
pro
ble
m: i
t le
ads
to c
ash
man
agem
ent
issu
es fo
r th
e tr
easu
ry a
nd
chal
leng
es in
ad
dre
ssin
g s
ecto
ral p
rog
ram
me
ob
ject
ives
.
Hes
itatio
n b
y d
ono
rs t
o u
se t
he t
reas
ury
acco
unt
for
bud
get
ex
ecut
ion
is p
artly
a q
uest
ion
of c
onfi
den
ce in
the
PFM
sys
tem
. Pe
rcei
ved
wea
knes
ses
in e
xter
nal a
udit
are
likel
y to
be
an im
po
rtan
t fa
cto
r.
Fro
m t
he g
ove
rnm
ent
sid
e, t
he la
ck o
f pre
dic
tab
ility
of
dis
bur
sem
ents
of b
udg
et s
upp
ort
, par
ticul
arly
sec
tor
bud
get
su
pp
ort
, is
of s
igni
fican
t co
ncer
n an
d h
as b
eco
me
a ca
sh
man
agem
ent
issu
e fo
r th
e tr
easu
ry. T
his
unp
red
icta
bili
ty, w
hich
o
ften
rela
tes
to t
he s
eco
nd, p
erfo
rman
ce-r
elat
ed t
ranc
he o
f sec
tor
bud
get
sup
po
rt, c
oul
d b
e lin
ked
to
wea
knes
ses
in t
he s
ettin
g o
f b
udg
et s
upp
ort
resu
lt in
dic
ato
rs, s
om
e o
f whi
ch m
ay b
e o
utsi
de
of
the
agen
cies
’ co
ntro
l.
Perc
eive
d w
eakn
esse
s in
ext
erna
l aud
it ar
e al
so a
n im
po
rtan
t fa
cto
r. Th
e re
cent
intr
od
uctio
n o
f the
Offi
ce o
f the
Aud
itor
Gen
eral
(Bur
eau
du
Vérifi
cate
ur G
énér
al) h
as p
rovi
ded
an
add
itio
nal p
ote
ntia
l so
urce
o
f ext
erna
l ove
rsig
ht, a
ltho
ugh
clar
ity o
n its
role
and
co
ord
inat
ion
with
the
Sup
rem
e C
our
t’s a
udit
inst
itutio
n re
mai
n im
po
rtan
t is
sues
.
92
MA
Li
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/W
hy n
ot?
)
On
acco
unt
Acc
oun
ting
sys
tem
s in
fran
cop
hone
co
untr
ies
are
typ
ical
ly s
pec
ified
fo
rmal
ly b
y a
dec
ree;
in M
ali,
the
reg
ulat
ions
are
co
ntai
ned
in a
la
w. I
n th
e fr
anco
pho
ne s
yste
m, t
wo
sep
arat
e ty
pes
of a
cco
untin
g
are
carr
ied
out
: (i)
The
Fina
ncia
l Co
ntro
ller
(Con
trôl
eur F
inan
cier
) in
the
Dep
artm
ent
of F
inan
cial
Co
ntro
llers
(Dire
ctio
n N
atio
nale
de
Con
trôl
e Fi
nanc
ière
) acc
oun
ts fo
r an
nual
Bud
get
Law
reso
urce
s ex
ecut
ed t
hro
ugh
the
trea
sury
(inc
lud
ing
do
mes
tic re
sour
ces,
GB
S in
the
ann
ual B
udg
et L
aw,13
and
sec
tor
bud
get
sup
po
rt) t
hro
ugh
the
exec
utio
n st
age
of i
ssua
nce
of a
pay
men
t o
rder
(ord
onna
ncem
ent).
(ii
) Pub
lic a
cco
unta
nts
in t
he t
reas
ury
(in t
he D
epar
tmen
t o
f Nat
iona
l Tr
easu
ry a
nd P
ublic
Acc
oun
tant
s [D
irect
ion
Nat
iona
le d
u Tr
ésor
et
de
la C
omp
tab
ilité
Pub
lique
]) ac
coun
t fo
r th
e fin
al s
tag
e o
f bud
get
ex
ecut
ion,
the
pay
men
t st
age.
Acc
ord
ing
to
Mal
i’s O
rgan
ic B
udg
et L
aw, i
n co
mm
on
with
oth
er
fran
cop
hone
sys
tem
s, t
he re
mit
of t
he F
inan
cial
Co
ntro
ller
is li
mite
d
to e
xpen
ditu
res
in t
he a
nnua
l Bud
get
Law
; thu
s, o
nly
bud
get
su
pp
ort
and
do
mes
tic re
sour
ces
are
sub
ject
to
the
Fin
anci
al
Co
ntro
ller’s
co
ntro
l and
acc
oun
ting
.
All
bud
get
sup
po
rt fo
llow
s na
tiona
l bud
get
exe
cutio
n p
roce
dur
es
in li
ne w
ith d
om
estic
reso
urce
s. It
is e
stim
ated
tha
t ar
oun
d 4
0% o
f to
tal a
id re
sour
ces
wer
e ac
coun
ted
for
usin
g n
atio
nal p
roce
dur
es in
20
06.14
Acc
oun
ting
for
reso
urce
s o
utsi
de
of t
he a
nnua
l Bud
get
Law
(e.g
. Sp
ecia
l Inv
estm
ent
Bud
get
reso
urce
s) is
car
ried
out
sep
arat
ely
by
PIU
acc
oun
tant
s an
d re
cord
ed in
sep
arat
e re
po
rts.
Prep
arin
g c
ons
olid
ated
acc
oun
ts (e
.g. f
or
po
ole
d fu
nds)
is n
ot
und
erta
ken;
thi
s p
ract
ice
do
es n
ot
app
ear
to b
e a
prio
rity
for
the
go
vern
men
t o
r IP
s.
The
emp
hasi
s in
the
Mal
ian
(and
oth
er fr
anco
pho
ne) b
udg
et
syst
ems
on
com
plia
nce
with
the
rele
vant
law
s an
d p
roce
dur
es is
an
imp
ort
ant
fact
or,
sinc
e b
udg
et e
xecu
tion
pro
ced
ures
as
set
out
in
the
law
are
lim
ited
to
tre
asur
y fu
nds
sho
wn
in t
he a
nnua
l Bud
get
La
w.
On
aud
itE
xter
nal s
crut
iny
of p
ublic
reso
urce
s is
the
prim
ary
resp
ons
ibili
ty o
f th
e Su
pre
me
Co
urt’s
aud
it in
stitu
tion.
Eac
h ye
ar, t
he in
stitu
tion
is
req
uire
d t
o e
xam
ine
the
annu
al B
udg
et E
xecu
tion
Law
and
cer
tify
that
the
tre
asur
y’s
acco
unts
(com
pte
s d
e g
estio
n) a
nd t
he B
udg
et
Dep
artm
ent’s
acc
oun
ts (c
omp
tes
adm
inis
trat
ifs) a
re fu
lly c
om
pat
ible
. Fo
llow
ing
its
revi
ew, t
he in
stitu
tion
issu
es a
rep
ort
on
the
exec
utio
n o
f the
ann
ual B
udg
et L
aw, t
og
ethe
r w
ith a
gen
eral
dec
lara
tion
of c
onf
orm
ity o
f the
acc
oun
ts. I
n p
rinci
ple
, the
ann
ual B
udg
et
Exe
cutio
n La
w c
an o
nly
be
pre
sent
ed t
o t
he N
atio
nal A
ssem
bly
fo
r ad
op
tion
onc
e th
e in
stitu
tion
has
issu
ed t
his
dec
lara
tion
of
conf
orm
ity. O
ther
wis
e, a
ny a
udits
car
ried
out
on
pro
ject
s (i.
e. t
hose
in
the
Sp
ecia
l Inv
estm
ent
Bud
get
) do
no
t in
volv
e th
e Su
pre
me
Co
urt’s
aud
it in
stitu
tion,
and
the
inst
itutio
n is
no
t a
reg
ular
reci
pie
nt
of t
hese
rep
ort
s.
The
pro
po
rtio
n o
f aid
reso
urce
s su
bje
ct t
o e
xter
nal s
crut
iny
thro
ugh
the
Sup
rem
e C
our
t’s a
udit
inst
itutio
n w
as ju
st u
nder
70%
in 2
004
(the
mo
st re
cent
rep
ort
).15 T
he in
stitu
tion
face
s a
sig
nific
ant
bac
klo
g
of a
nnua
l Bud
get
Exe
cutio
n La
ws
to a
udit,
und
erm
inin
g t
he
effe
ctiv
enes
s o
f its
revi
ews.
The
ann
ual B
udg
et E
xecu
tion
Law
s fr
om
19
99–2
004
wer
e p
artic
ular
ly d
elay
ed.
Ext
erna
l scr
utin
y in
ter
ms
of e
xter
nal a
udit
is w
eak
in M
ali a
nd
is a
co
ncer
n fo
r b
oth
the
ext
erna
l aud
itor
and
for
dev
elo
pm
ent
par
tner
s. O
ne o
f the
prim
ary
conc
erns
is t
he la
ck o
f hum
an a
nd
mat
eria
l res
our
ces
for
the
Sup
rem
e C
our
t’s a
udit
inst
itutio
n: o
nly
15
mag
istr
ates
ann
ually
aud
it al
l ins
titut
ions
rece
ivin
g p
ublic
mo
ney
(incl
udin
g e
ntiti
es a
t b
oth
cen
tral
and
loca
l lev
els)
, am
oun
ting
to
ap
pro
xim
atel
y 89
inst
itutio
ns. T
he li
mite
d c
apac
ity h
as re
sulte
d
in d
elay
s in
sub
mitt
ing
the
ann
ual B
udg
et E
xecu
tion
Law
to
the
N
atio
nal A
ssem
bly
, with
the
200
4 an
nual
Bud
get
Exe
cutio
n La
w
rep
ort
the
mo
st re
cent
to
hav
e b
een
sub
mitt
ed. T
he re
view
of t
he
2005
ann
ual B
udg
et E
xecu
tion
Law
is o
nly
exp
ecte
d b
y th
e en
d o
f th
e ye
ar.
Part
of t
he re
aso
n fo
r lim
ited
cap
aciti
es is
tha
t th
e Su
pre
me
Co
urt’s
au
dit
inst
itutio
n is
no
t fin
anci
ally
ind
epen
den
t –
it is
one
am
ong
co
mp
etin
g c
ham
ber
s o
f the
Sup
rem
e C
our
t, a
nd m
ust
rely
on
the
Sup
rem
e C
our
t fo
r its
fund
ing
. In
add
itio
n, t
here
is a
co
nstit
utio
nal
limit
on
the
num
ber
of S
upre
me
Co
urt
mag
istr
ates
.
93
MA
Li
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/W
hy n
ot?
)
On
rep
ort
The
Min
istr
y o
f Eco
nom
y an
d F
inan
ce p
rod
uces
four
typ
es o
f b
udg
et e
xecu
tion
rep
ort
s: (i
) qua
rter
ly b
udg
et e
xecu
tion
rep
ort
s b
y th
e Fi
nanc
ial C
ont
rolle
r co
verin
g t
he e
xecu
tion
stag
es u
p t
o t
he
sub
mis
sio
n o
f the
pay
men
t o
rder
(ord
onna
ncem
ent);
(ii)
mo
nthl
y an
d q
uart
erly
tre
asur
y b
alan
ces,
set
ting
out
the
ban
k b
alan
ces
and
thu
s co
verin
g t
he p
aym
ent
stag
e; (i
ii) t
he c
ons
olid
ated
an
nual
fina
ncia
l sta
tem
ent,
in t
he fo
rm o
f the
ann
ual B
udg
et
Exe
cutio
n La
w, w
hich
is in
tend
ed t
o re
conc
ile t
he t
wo
form
s o
f ac
coun
ting
info
rmat
ion;
and
(iv)
qua
rter
ly a
nd a
nnua
l Sp
ecia
l In
vest
men
t B
udg
et re
po
rts
on
pro
ject
reso
urce
s as
par
t o
f the
Sp
ecia
l Inv
estm
ent
Bud
get
. In
add
itio
n, t
he t
reas
ury
pro
duc
es t
he
cons
olid
ated
fisc
al t
able
(Tab
leau
des
Op
érat
ions
Fin
anci
ères
de
l’Eta
t), w
hich
pro
vid
es a
fisc
al o
verv
iew
, inc
lud
ing
det
ails
of t
he
defi
cit
and
its
finan
cing
.
Thus
, fo
r b
udg
et re
sour
ces
op
erat
ing
thr
oug
h na
tiona
l sys
tem
s,
rep
ort
ing
is s
epar
ated
bet
wee
n ex
ecut
ion
stag
es u
p t
o
ord
onna
ncem
ent u
nder
the
resp
ons
ibili
ty o
f the
Fin
anci
al C
ont
rolle
r o
n th
e o
ne h
and
and
pai
emen
t, re
po
rted
by
the
trea
sury
as
par
t o
f ban
k re
conc
iliat
ions
, on
the
oth
er. T
his
sep
arat
ion
is re
flect
ed in
th
e an
nual
acc
oun
ts, w
hich
co
mp
rise:
(i) t
he B
udg
et D
epar
tmen
t’s
acco
unts
, pro
vid
ing
det
ails
of r
even
ues
and
exp
end
iture
s b
y se
cto
r m
inis
trie
s up
to
, and
incl
udin
g, t
he s
ubm
issi
on
of t
he p
aym
ent
ord
er (o
rdon
nanc
emen
t), p
rep
ared
by
the
Bud
get
Dep
artm
ent;
and
(ii
) the
tre
asur
y’s
acco
unts
, whi
ch s
how
the
acc
oun
t b
alan
ces
and
tr
ansa
ctio
ns –
at
the
enca
shm
ent
stag
e fo
r re
venu
es a
nd t
he c
ash
pay
men
t st
age
for
exp
end
iture
s.
The
2004
ann
ual B
udg
et E
xecu
tion
Law
incl
uded
gra
nt-fi
nanc
ed
bud
get
sup
po
rt, s
how
n as
“ex
cep
tiona
l rec
eip
ts”
(whi
le s
how
ing
a
0 in
the
co
rres
po
ndin
g p
lann
ed/b
udg
eted
[i.e
. ann
ual B
udg
et L
aw]
colu
mn)
, ind
icat
ing
tha
t th
ese
reso
urce
s w
ent
thro
ugh
the
trea
sury
sy
stem
but
wer
e no
t p
lann
ed e
xplic
itly
bef
ore
hand
: thi
s is
the
bas
is
for
rep
ort
ing
on
bud
get
sup
po
rt in
the
co
nso
lidat
ed fi
scal
tab
le.
Ove
rall,
the
pro
po
rtio
n o
f to
tal a
id re
sour
ces
incl
uded
in t
he
norm
al b
udg
et re
po
rtin
g p
roce
ss is
est
imat
ed t
o b
e ar
oun
d 9
0%.16
H
ow
ever
, acc
ord
ing
to
the
PE
FA a
sses
smen
t, t
here
are
sig
nific
ant
wea
knes
ses
with
in-y
ear
and
end
-yea
r re
po
rtin
g, i
nclu
din
g t
he la
ck
of c
om
pre
hens
iven
ess
of t
he d
ata
for
bud
get
exe
cutio
n (t
he d
ata
do
no
t in
clud
e th
e p
aym
ent
stag
e), a
nd s
igni
fican
t d
elay
s in
the
p
rod
uctio
n o
f the
se re
po
rts.
Prep
arin
g c
ons
olid
ated
acc
oun
ts (e
.g. f
or
po
ole
d fu
nds)
is n
ot
und
erta
ken;
thi
s p
ract
ice
do
es n
ot
app
ear
to b
e a
prio
rity
for
the
go
vern
men
t o
r IP
s.
In p
rinci
ple
, the
shi
ft in
aid
mo
dal
ity fr
om
po
ole
d fu
nds
to s
ecto
r b
udg
et s
upp
ort
sho
uld
lead
to
incr
ease
d t
rans
par
ency
of r
epo
rtin
g
in t
he a
nnua
l Bud
get
Exe
cutio
n La
w, p
artic
ular
ly s
ince
sec
tor
bud
get
sup
po
rt fu
nds
are
exp
licitl
y sh
ow
n in
the
bud
get
do
cum
ents
b
y in
div
idua
l min
istr
ies.
The
200
5 an
nual
Bud
get
Exe
cutio
n La
w
sho
uld
co
nfirm
thi
s. T
he 2
004
annu
al B
udg
et E
xecu
tion
Law
refe
rs
to t
he p
erio
d b
efo
re t
he m
ove
to
sec
tor
bud
get
sup
po
rt.
The
stro
ng e
mp
hasi
s o
n co
ntro
l co
mm
on
to fr
anco
pho
ne
coun
trie
s, w
ith t
he s
epar
atio
n o
f res
po
nsib
ilitie
s fo
r th
e fir
st t
hree
st
ages
of b
udg
et e
xecu
tion
fro
m t
he p
aym
ent
stag
e, le
ads
to a
n o
vere
mp
hasi
s o
n th
e fin
anci
al a
cco
untin
g s
ide
at t
he e
xpen
se o
f b
udg
et m
anag
emen
t, fo
cusi
ng o
n b
udg
et o
utco
mes
.
94
Mo
ZA
MB
iQu
e
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(Ho
w w
ell?
Ho
w u
sefu
l?)
exp
lana
tio
ns(W
hy/W
hy n
ot?
)
On
pla
nTh
e M
oza
mb
ican
PR
SP, t
he P
AR
PA-I
I, w
hich
is t
he g
ove
rnm
ent’s
m
ain
inst
rum
ent
for
stra
teg
ic p
lann
ing
, inc
lud
es in
its
mac
roec
ono
mic
sce
nario
thr
ee p
roje
ctio
ns o
f the
exp
ecte
d re
sour
ce
enve
lop
e, b
ased
on
varia
tions
of e
xter
nal O
DA
flo
ws
mad
e av
aila
ble
to
Mo
zam
biq
ue.
The
go
vern
men
t’s m
acro
eco
nom
ic fr
amew
ork
mo
del
incl
udes
thr
ee
spre
adsh
eets
with
reas
ona
bly
det
aile
d p
roje
ctio
ns o
f exp
ecte
d
aid
flo
ws
to M
oza
mb
ique
, inc
lud
ing
tho
se d
isb
urse
d t
o t
he
go
vern
men
t, o
ver
a th
ree-
year
per
iod
.
The
pub
lic v
ersi
on
of M
oza
mb
ique
’s M
ediu
m-T
erm
Fis
cal
Fram
ewo
rk a
lso
incl
udes
a b
reak
do
wn
of m
ediu
m-t
erm
p
rog
ram
med
inve
stm
ent
exp
end
iture
, by
sour
ce o
f fina
nce,
ove
r a
thre
e-ye
ar p
erio
d. I
t d
istin
gui
shes
bet
wee
n in
tern
al a
nd e
xter
nal
sour
ces,
the
latt
er c
orr
esp
ond
ing
to
aid
fund
ing
to
the
go
vern
men
t.
The
go
vern
men
t al
so in
corp
ora
tes
info
rmat
ion
on
exp
ecte
d a
id
dis
bur
sem
ents
at
low
er le
vels
of p
lann
ing
, inc
lud
ing
at
sect
or
and
p
rovi
ncia
l lev
els.
The
info
rmat
ion
on
exte
rnal
fund
ing
cur
rent
ly p
rese
nted
in t
he
Med
ium
-Ter
m F
isca
l Fra
mew
ork
and
mac
roec
ono
mic
fram
ewo
rk is
st
ill li
mite
d, a
s a
resu
lt o
f: (i)
do
nors
’ diffi
culti
es in
sup
ply
ing
suc
h in
form
atio
n an
d t
heir
inab
ility
to
defi
ne m
ulti-
year
aid
co
mm
itmen
ts
for
all t
heir
pro
gra
mm
es in
Mo
zam
biq
ue; a
nd (i
i) co
ord
inat
ion
and
co
mm
unic
atio
n p
rob
lem
s b
etw
een
the
Min
istr
y o
f Fin
ance
an
d t
he M
inis
try
of P
lann
ing
and
Dev
elo
pm
ent
on
one
sid
e, a
nd
line
min
istr
ies
and
pro
vinc
ial g
ove
rnm
ents
on
the
oth
er, a
s w
ell a
s th
e w
eak
inte
gra
tion
of p
lann
ing
inst
rum
ents
at
vario
us le
vels
of
go
vern
men
t.
A k
ey fa
cto
r b
ehin
d t
he c
ons
ider
able
suc
cess
in t
his
sphe
re is
th
e re
aso
nab
ly s
tro
ng d
ialo
gue
, at
vario
us le
vels
, bet
wee
n th
e g
ove
rnm
ent
and
do
nors
. Thi
s is
evi
den
ced
in t
he n
umer
ous
form
al
and
info
rmal
inst
itutio
nal a
rran
gem
ents
tha
t ha
ve e
mer
ged
ove
r th
e p
ast
dec
ade.
The
OD
AM
OZ
dat
abas
e o
f aid
-fun
ded
pro
ject
s ha
s d
etai
led
in
form
atio
n o
n al
l pro
ject
s fu
nded
by
the
do
nors
par
ticip
atin
g
in t
his
aid
dis
sem
inat
ion
initi
ativ
e an
d is
an
imp
ort
ant
sour
ce o
f in
form
atio
n fo
r th
e g
ove
rnm
ent.
95
Mo
ZA
MB
iQu
e
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(Ho
w w
ell?
Ho
w u
sefu
l?)
exp
lana
tio
ns(W
hy/W
hy n
ot?
)
On
bud
get
66%
of O
DA
by
Pro
gra
mm
e A
id P
artn
er g
roup
of d
ono
rs (w
hich
in
clud
es m
ost
of t
he m
ain
inte
rnat
iona
l dev
elo
pm
ent
agen
cies
o
per
atin
g in
Mo
zam
biq
ue) w
as re
po
rted
on
bud
get
in 2
006.
By
spec
ific
do
nor,
this
per
cent
age
was
par
ticul
arly
low
for
coun
trie
s su
ch a
s G
erm
any
(44%
), It
aly
(42%
), Sp
ain
(29%
), Po
rtug
al (1
3%)
and
Den
mar
k (6
%).
Oth
er d
ono
rs w
ere
bet
ter
rep
ort
ed: t
he W
orld
B
ank
(100
%),
Bel
giu
m (9
7%),
Irela
nd (9
4%),
Swed
en (9
0%),
the
Net
herla
nds
(89%
), th
e U
K (8
7%),
Fran
ce (8
4%) a
nd F
inla
nd (8
3%).
Whi
le n
o re
cent
dat
a o
n ai
d o
n b
udg
et e
xist
s fo
r d
ono
rs o
utsi
de
the
Pro
gra
mm
e A
id P
artn
ers
gro
up (s
uch
as t
he U
N a
gen
cies
, USA
ID o
r th
e ne
w v
ertic
al g
lob
al fu
nds)
, ava
ilab
le e
vid
ence
sug
ges
ts t
hat
thei
r ai
d is
sig
nific
antly
less
wel
l cap
ture
d o
n b
udg
et.
The
bud
get
pre
sent
ed t
o p
arlia
men
t g
ives
a b
reak
do
wn
of
inve
stm
ent
exp
end
iture
by
sour
ce o
f fun
din
g, d
istin
gui
shin
g
bet
wee
n in
tern
al a
nd e
xter
nal s
our
ces,
with
the
latt
er c
orr
esp
ond
ing
to
OD
A-fi
nanc
ed g
ove
rnm
ent
exp
end
iture
. Ho
wev
er, t
here
is n
o
det
aile
d b
reak
do
wn
of t
he e
xact
so
urce
(i.e
. do
nor)
of t
he fu
ndin
g.
Sim
ilarly
, the
sec
tion
in t
he b
udg
et t
hat
pre
sent
s re
sour
ce
avai
lab
ility
als
o in
dic
ates
go
vern
men
t re
sour
ces
orig
inat
ing
fro
m
gra
nts
and
loan
s, b
ut w
ith n
o d
etai
led
bre
akd
ow
n o
f the
ir ex
act
sour
ce o
r na
ture
(e.g
. GB
S, c
om
mo
n fu
nds
finan
cing
, pro
ject
aid
, et
c.).
The
go
vern
men
t ha
s co
mm
itted
to
imp
rovi
ng t
he q
ualit
y an
d
det
ail o
f the
info
rmat
ion
it re
po
rts
on
reso
urce
ava
ilab
ility
.
The
Min
istr
y o
f Fin
ance
co
llect
s fa
r m
ore
info
rmat
ion
on
OD
A
fund
ing
tha
n th
at re
po
rted
in t
he b
udg
et d
ocu
men
t, s
ugg
estin
g
ther
e is
sig
nific
ant
sco
pe
to im
pro
ve t
he q
ualit
y an
d d
etai
l of
info
rmat
ion
on
OD
A fu
ndin
g o
f go
vern
men
t p
rog
ram
mes
in t
he
bud
get
and
its
acco
mp
anyi
ng d
ocu
men
t.
Ther
e ar
e al
so c
onc
erns
ab
out
the
qua
lity
of t
he in
form
atio
n o
n ex
tern
al fu
ndin
g fo
r in
vest
men
t p
roje
cts
whi
ch is
incl
uded
in t
he
bud
get
s p
asse
d o
n b
y lin
e m
inis
trie
s to
the
Min
istr
y o
f Fin
ance
d
urin
g t
he b
udg
et p
rep
arat
ion
pro
cess
. Fo
r ex
amp
le, t
he re
cent
g
ove
rnm
ent
aud
it o
f the
wat
er s
ecto
r fin
ds
very
imp
ort
ant
dis
crep
anci
es b
etw
een
the
rep
ort
ed v
alue
s o
f ext
erna
lly fu
nded
in
vest
men
t p
roje
cts
incl
uded
in t
he 2
005
bud
get
and
the
act
ual o
r re
al fu
ndin
g a
vaila
ble
for
thes
e sa
me
pro
ject
s –
larg
ely
refle
ctin
g
mis
calc
ulat
ions
due
to
inco
rrec
t fo
reig
n ex
chan
ge
rate
s.
In t
he p
ast,
effo
rts
to b
ring
aid
on
bud
get
wer
e hi
nder
ed b
y th
e (u
nint
end
ed) d
isto
rtio
ns g
ener
ated
by
the
incl
usio
n o
f a d
om
estic
p
rimar
y d
efici
t ta
rget
in t
he IM
F’s
mac
roec
ono
mic
and
fisc
al s
upp
ort
p
rog
ram
me
to t
he g
ove
rnm
ent.
Oth
er fa
cto
rs:
Ind
ivid
ual d
onor
s’ d
ecis
ions
– e
.g. D
enm
ark’
s d
ecis
ion
not
to
dis
bur
se t
hree
-qua
rter
s o
f the
ir co
mm
itted
GB
S fu
ndin
g in
resp
ons
e to
aud
its c
arrie
d o
ut in
var
ious
Dan
ida-
fund
ed p
roje
cts,
whi
ch w
ere
felt
to re
veal
cas
es o
f ser
ious
mis
man
agem
ent,
if n
ot
dire
ctly
co
rrup
t p
ract
ices
.
Diffi
culti
es th
at d
onor
s so
met
imes
hav
e in
pro
vid
ing
relia
ble
and
tim
ely
info
rmat
ion
on
thei
r ai
d p
rog
ram
mes
in M
oza
mb
ique
and
ac
cura
te s
ched
ules
of t
heir
exp
ecte
d a
id d
isb
urse
men
ts.
Don
ors’
cho
ice
of a
id m
odal
ity, w
hich
larg
ely
pre
det
erm
ines
the
d
egre
e to
whi
ch t
hey
can
rep
ort
aid
on
bud
get
at
this
and
oth
er
leve
ls o
f the
bud
get
cyc
le:
(i) S
om
e co
untr
ies
dis
bur
se a
larg
e p
rop
ort
ion
of t
heir
aid
to
g
ove
rnm
ent
thro
ugh
NG
Os,
whi
ch a
re t
hen
resp
ons
ible
for
imp
lem
entin
g t
hese
fund
s an
d a
sso
ciat
ed p
roje
cts
with
in t
he
go
vern
men
t se
cto
rs in
whi
ch t
hey
op
erat
e, a
dd
ing
ano
ther
act
or
to
the
aid
imp
lem
enta
tion
pro
cess
, and
furt
her
hurd
les
for
brin
gin
g a
id
on
bud
get
.
(ii) O
ther
ag
enci
es, s
uch
as m
ost
ag
enci
es o
per
atin
g w
ithin
the
UN
sy
stem
, lar
gel
y b
ase
thei
r ai
d t
o t
he g
ove
rnm
ent
on
the
pro
visi
on
of
TA, o
ften
fina
nced
by
third
-par
ty a
gen
cies
, whi
ch c
an s
om
etim
es b
e d
ifficu
lt to
inco
rpo
rate
in t
he b
udg
et fo
rmul
atio
n p
roce
ss.
(iii)
A k
ey d
river
beh
ind
the
gro
win
g v
olu
mes
of a
id o
n b
udg
et in
M
oza
mb
ique
has
bee
n d
ono
rs’ i
ncre
asin
g u
se o
f pro
gra
mm
atic
aid
m
od
aliti
es.
96
Mo
ZA
MB
iQu
e
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(Ho
w w
ell?
Ho
w u
sefu
l?)
exp
lana
tio
ns(W
hy/W
hy n
ot?
)
On
par
liam
ent
The
Ass
emb
ly o
f the
Rep
ublic
has
no
leg
al a
utho
rity
to a
pp
rove
aid
to
the
go
vern
men
t. W
hat
it d
oes
ap
pro
ve is
the
nat
iona
l bud
get
. In
asm
uch
as t
he g
ove
rnm
ent’s
bud
get
pro
po
sal i
nco
rpo
rate
s, a
s p
art
of i
ts re
sour
ce b
ase,
exp
ecte
d a
id d
isb
urse
men
ts, t
he A
ssem
bly
o
f the
Rep
ublic
pla
ys a
n in
dire
ct ro
le in
deb
atin
g a
nd s
anct
ioni
ng
the
use
that
the
go
vern
men
t m
akes
of t
hat
aid
.
Sim
ilarly
, the
fact
tha
t b
udg
et p
erfo
rman
ce, i
nclu
din
g t
he u
se
of e
xter
nal r
eso
urce
s, is
als
o e
xam
ined
by
the
Ass
emb
ly o
f the
R
epub
lic, i
mp
lies
that
par
liam
ent
ind
irect
ly p
lays
a ro
le in
ass
essi
ng
the
use
that
go
vern
men
t m
akes
of d
evel
op
men
t ai
d.
This
bud
get
revi
ew p
roce
ss t
akes
pla
ce t
wic
e a
year
dur
ing
the
d
ebat
es o
rgan
ised
to
exa
min
e b
oth
the
go
vern
men
t’s b
udg
et
exec
utio
n re
po
rt –
Rel
ator
io d
e Ex
ecuç
ão d
o O
rçam
ento
, and
the
im
ple
men
tatio
n o
f the
go
vern
men
t’s a
nnua
l so
cial
and
eco
nom
ic
pla
n –
Bal
anço
do
PES.
The
qua
lity
of b
oth
the
revi
ew a
nd a
pp
rova
l of t
he b
udg
et a
nd it
s ac
com
pan
ying
do
cum
ents
is g
ener
ally
co
nsid
ered
wea
k.17
Parli
amen
t’s la
ck o
f tec
hnic
al, h
uman
and
fina
ncia
l res
our
ces
und
erm
ines
its
abili
ty t
o t
horo
ughl
y an
alys
e an
d c
ont
rol b
udg
et
pro
po
sals
and
exe
cutio
n, in
clud
ing
its
OD
A c
om
po
nent
.
In a
dd
itio
n, t
he u
nder
lyin
g t
ensi
ons
bet
wee
n th
e (F
RE
LIM
O)
go
vern
men
t an
d t
he m
ain
op
po
sitio
n p
arty
(RE
NA
MO
) affe
ct t
he
qua
lity
of p
arlia
men
tary
bud
get
deb
ates
, im
ped
ing
in-d
epth
and
o
bje
ctiv
e ex
amin
atio
ns o
f the
go
vern
men
t’s b
udg
et p
rop
osa
ls a
nd
sub
seq
uent
bud
get
ary
per
form
ance
.
97
Mo
ZA
MB
iQu
e
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(Ho
w w
ell?
Ho
w u
sefu
l?)
exp
lana
tio
ns(W
hy/W
hy n
ot?
)
On
trea
sury
Dur
ing
the
last
dec
ade,
brin
gin
g a
id o
n tr
easu
ry h
as b
een
at t
he
cent
re o
f muc
h o
f do
nors
’ and
the
go
vern
men
t’s e
ffort
s to
brin
g
aid
on
bud
get
. Im
pro
ving
aid
on
trea
sury
has
bee
n o
ne o
f the
m
ain
ob
ject
ives
of t
he d
ono
r–g
ove
rnm
ent
on/
off
bud
get
tas
kfo
rce
crea
ted
in 2
005
to in
crea
se t
he u
se o
f lo
cal p
lann
ing
, bud
get
ing
and
PF
M s
yste
ms
to m
anag
e ai
d-r
elat
ed fu
ndin
g.
The
ove
rall
cap
ture
of a
id o
n tr
easu
ry fo
r th
e Pr
og
ram
me
Aid
Pa
rtne
rs g
roup
of d
ono
rs in
200
6 is
est
imat
ed a
t 43
% o
f the
ir ac
tual
ai
d t
o g
ove
rnm
ent.
As
with
OD
A o
n b
udg
et, t
here
is c
ons
ider
able
var
iab
ility
in t
he
Pro
gra
mm
e A
id P
artn
ers
gro
up re
gar
din
g t
he e
xten
t to
whi
ch
they
mak
e us
e o
f the
go
vern
men
t tr
easu
ry a
nd p
aym
ents
sys
tem
. D
ono
r co
untr
ies
or
agen
cies
suc
h as
Bel
giu
m, D
FID
, Fin
land
, the
N
ethe
rland
s an
d Ir
elan
d c
urre
ntly
dis
bur
se m
ost
of t
heir
aid
into
the
tr
easu
ry –
thi
s sh
are
is a
bo
ve 7
5% fo
r al
l five
. Oth
ers
in t
he g
roup
, ho
wev
er, h
ave
bee
n p
artic
ular
ly s
low
in m
eetin
g t
he P
rog
ram
me
Aid
Par
tner
s ta
rget
of 4
5%, e
spec
ially
Sp
ain
(19%
), Po
rtug
al (1
3%)
and
Den
mar
k (6
%),
alth
oug
h fo
r D
enm
ark
this
ap
pea
rs t
o b
e re
late
d
to it
s co
ncer
ns o
ver
the
use
the
go
vern
men
t w
as m
akin
g o
f its
aid
fu
nds.
Out
sid
e th
e Pr
og
ram
me
Aid
Par
tner
s g
roup
, pro
gre
ss in
usi
ng
trea
sury
sys
tem
s ha
s b
een
sig
nific
antly
slo
wer
, with
the
sha
re o
f aid
p
roce
ssed
thr
oug
h th
ese
syst
ems
ove
r to
tal a
id t
o t
he g
ove
rnm
ent
estim
ated
at
14.3
% fo
r Ja
pan
, jus
t o
ver
1% fo
r U
N a
gen
cies
and
0%
for
USA
ID a
nd fu
nds
fro
m G
AV
I. Th
e G
FATM
is a
s an
exc
eptio
n am
ong
the
no
n-Pr
og
ram
me
Aid
Par
tner
s d
ono
r g
roup
– a
ll o
f its
fu
nds
to t
he g
ove
rnm
ent
are
dis
bur
sed
thr
oug
h th
e tr
easu
ry s
yste
m.
Dire
ct e
xpen
ditu
re e
xecu
tion
thro
ugh
the
STA
(kno
wn
as C
UT)
sy
stem
is s
till l
imite
d, a
nd a
lmo
st n
on-
exis
tent
for
pro
ject
s.
Est
imat
es fo
r 20
06 s
ugg
est
that
onl
y 4.
3% o
f bud
get
tra
nsac
tions
(c
orr
esp
ond
ing
to
2.7
% o
f to
tal b
udg
et e
xpen
ditu
re) w
ere
dire
ctly
ex
ecut
ed t
hro
ugh
the
Inte
gra
ted
Sys
tem
for
Stat
e Fi
nanc
ial
Man
agem
ent
(e-S
ISTA
FE).
The
syst
em s
till c
anno
t m
anag
e m
ulti-
curr
ency
op
erat
ions
, whi
ch
has
long
bee
n id
entifi
ed a
s an
imp
ort
ant
imp
edim
ent
in b
ring
ing
ai
d o
n tr
easu
ry.
In a
dd
itio
n, t
here
are
stil
l man
y ag
enci
es, n
amel
y at
dis
tric
t le
vel,
that
can
not
op
erat
e th
e In
teg
rate
d S
yste
m fo
r St
ate
Fina
ncia
l M
anag
emen
t to
exe
cute
the
ir b
udg
eted
fund
s, s
ince
the
y d
o n
ot
have
the
req
uire
d fa
cilit
ies
– re
gul
ar e
lect
ricity
sup
ply
, IT
syst
ems
or
loca
l ban
king
faci
litie
s. S
imila
rly, t
he s
yste
m c
anno
t m
anag
e al
l co
mp
one
nts
of t
he b
udg
et, u
nder
min
ing
effo
rts
to b
ring
all
pay
men
ts, i
nclu
din
g O
DA
fund
ing
, thr
oug
h th
e tr
easu
ry s
yste
m.
Fina
lly, n
ot
all c
lient
and
co
ntra
cto
rs’/
sup
plie
rs’ b
ank
acco
unts
ha
ve b
een
incl
uded
in t
he s
yste
m. A
s re
sult,
a la
rge
pro
po
rtio
n o
f th
e g
ove
rnm
ent’s
fina
ncia
l tra
nsac
tions
co
ntin
ue t
o b
e un
der
take
n un
der
the
old
sys
tem
of p
aym
ents
by
cheq
ue.
Ther
e is
stil
l sig
nific
ant
wo
rk t
o b
e d
one
by
bo
th d
ono
rs a
nd t
he
go
vern
men
t to
ach
ieve
the
Per
form
ance
Ass
ista
nce
Fram
ewo
rk
ind
icat
or
of h
avin
g a
t le
ast
60%
of O
DA
cha
nnel
led
thr
oug
h th
e ST
A
by
2009
– a
pp
roxi
mat
ely
US$
900
mill
ion.
Past
con
stra
ints
(i) In
the
pas
t, a
key
fact
or
det
errin
g d
ono
rs a
nd re
cip
ient
g
ove
rnm
ent
agen
cies
alik
e fr
om
usi
ng t
he c
entr
al t
reas
ury
and
p
aym
ents
sys
tem
was
the
lack
of r
elia
bili
ty, t
imel
ines
s, fl
exib
ility
and
tr
ansp
aren
cy o
f the
Mo
zam
bic
an P
FM s
yste
m a
nd t
he a
sso
ciat
ed
liqui
dity
pro
ble
ms
that
freq
uent
ly a
ffect
ed t
he c
entr
al t
reas
ury.
(ii) I
n ad
diti
on,
the
way
the
tre
asur
y w
as o
rigin
ally
co
ncei
ved
mad
e it
bur
den
som
e an
d c
ost
ly t
o p
roce
ss fo
reig
n cu
rren
cy p
aym
ents
fo
r im
po
rts
of g
oo
ds
and
ser
vice
s, w
hich
wer
e o
ften
an
imp
ort
ant
com
po
nent
of t
hese
aid
/dev
elo
pm
ent
pro
ject
s. D
isb
ursi
ng a
nd
man
agin
g a
id t
hro
ugh
the
cent
ral t
reas
ury
and
pay
men
ts s
yste
m
cam
e at
the
co
st o
f red
ucin
g t
he e
ffect
iven
ess
of d
ono
r-fu
nded
p
roje
cts.
Ong
oing
con
stra
int
Som
e d
ono
r ag
enci
es a
re b
arre
d b
y th
eir
go
vern
men
ts fr
om
usi
ng
the
loca
l PFM
sys
tem
s, o
r, si
mila
rly, t
heir
ow
n fin
anci
al m
anag
emen
t ru
les
are
not
fully
co
mp
atib
le w
ith t
hose
of t
he M
oza
mb
ican
g
ove
rnm
ent.
Driv
ers
of c
aptu
ring
aid
on
trea
sury
(i) A
s w
ith e
ffort
s to
cap
ture
aid
on
bud
get
, cap
turin
g a
id o
n tr
easu
ry a
lso
ap
pea
rs t
o h
ave
bee
n d
riven
, to
a c
ons
ider
able
ext
ent,
b
y th
e g
row
ing
use
of p
rog
ram
mat
ic m
od
aliti
es o
f aid
dis
bur
sem
ent
and
man
agem
ent
by
do
nor
agen
cies
in M
oza
mb
ique
.
(ii) T
he In
teg
rate
d S
yste
m fo
r St
ate
Fina
ncia
l Man
agem
ent
PFM
re
form
pac
kag
e ha
s al
low
ed t
he g
ove
rnm
ent
to a
dd
ress
so
me
of
the
pro
ble
ms
cons
trai
ning
PFM
op
erat
ions
and
has
thu
s fa
cilit
ated
th
e p
roce
ss o
f brin
gin
g a
id o
n b
udg
et a
t tr
easu
ry, p
ublic
acc
oun
ting
an
d p
rocu
rem
ent
syst
em le
vels
.
(iii)
The
go
vern
men
t ha
s b
een
wo
rkin
g t
o m
erg
e al
l fina
ncia
l ac
coun
ts in
to o
ne S
TA, a
nd t
o b
ring
all
sour
ces
of f
und
ing
– in
tern
al
and
ext
erna
l – u
nder
the
um
bre
lla o
f the
go
vern
men
t in
teg
rate
d
bud
get
and
PFM
sys
tem
. Thi
s ha
s he
lped
to
brin
g o
rder
and
cla
rity
to t
he t
reas
ury
and
thu
s in
crea
se it
s tr
ansp
aren
cy, e
nco
urag
ing
d
ono
rs t
o d
isb
urse
the
ir fu
nds
thro
ugh
the
trea
sury
, and
g
ove
rnm
ent
agen
cies
to
rely
on
thes
e sy
stem
s m
ore
will
ing
ly.
98
Mo
ZA
MB
iQu
e
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(Ho
w w
ell?
Ho
w u
sefu
l?)
exp
lana
tio
ns(W
hy/W
hy n
ot?
)
On
acco
unt
No
pre
cise
and
co
mp
rehe
nsiv
e es
timat
es e
xist
on
the
volu
me
of
aid
cur
rent
ly b
eing
pro
cess
ed t
hro
ugh
the
Mo
zam
bic
an a
cco
untin
g
syst
em.
By
defi
nitio
n, a
ll d
ono
r fu
nds
dis
bur
sed
as
GB
S ar
e p
roce
ssed
usi
ng
acco
untin
g s
yste
ms
and
are
the
refo
re c
aptu
red
on
acco
unt.
Als
o m
ost
aid
fund
s ch
anne
lled
thr
oug
h co
mm
on
fund
ing
sch
emes
in
ag
ricul
ture
, ed
ucat
ion
and
hea
lth h
ave
bee
n d
isb
urse
d t
hro
ugh
the
STA
sys
tem
sin
ce 2
007.
The
strin
gen
t d
ocu
men
tary
req
uire
men
ts a
sso
ciat
ed w
ith t
he
duo
déc
imo
syst
em o
f ad
vanc
e p
aym
ents
, co
mb
ined
with
d
ifficu
lties
tha
t g
ove
rnm
ent
agen
cies
oft
en h
ad w
ith m
eetin
g
thes
e re
qui
rem
ents
on
time
and
the
num
ero
us p
rob
lem
s o
f m
ism
anag
emen
t an
d p
oo
r ac
coun
ting
, hav
e le
d m
any
do
nors
to
o
pt,
alto
get
her,
for
par
alle
l (co
nseq
uent
ly, o
ff b
udg
et) a
cco
untin
g
arra
ngem
ents
in t
heir
dea
ling
s w
ith g
ove
rnm
ent
agen
cies
.
Ref
orm
s b
eing
und
erta
ken
und
er t
he In
teg
rate
d S
yste
m fo
r St
ate
Fina
ncia
l Man
agem
ent
refo
rm p
acka
ge
are
seen
as
inst
rum
enta
l in
enco
urag
ing
do
nors
to
brin
g a
id o
n b
udg
et, a
lso
at
the
acco
untin
g
leve
l. Th
e re
form
s ar
e le
adin
g t
o t
he s
imp
lifica
tion
of d
ocu
men
tary
an
d a
cco
untin
g p
roce
dur
es.
A k
ey p
reco
nditi
on
for
succ
ess
in t
his
sphe
re is
str
ong
inte
rnal
co
ntro
l and
aud
it sy
stem
s, in
par
ticul
ar fo
r w
arra
ntin
g t
hat
the
roll-
out
of t
he In
teg
rate
d S
yste
m fo
r St
ate
Fina
ncia
l Man
agem
ent
and
d
irect
bud
get
exe
cutio
n tr
ansl
ate
into
co
ncre
te in
crea
ses
in t
he
deg
ree
of a
id c
aptu
red
on
acco
unt.
On
aud
itIn
tern
al a
udit
func
tions
– A
t p
rese
nt t
he In
spec
tor
Gen
eral
of
Fina
nce
(Insp
ecçã
o G
eral
de
Fina
nças
) car
ries
out
per
iod
ic in
tern
al
aud
its o
f go
vern
men
t p
olic
y an
d a
dm
inis
trat
ive
syst
ems.
The
se h
ave
incl
uded
exe
rcis
es t
hat
have
exa
min
ed e
xist
ing
PFM
pra
ctic
es a
nd
syst
ems,
and
in s
om
e ca
ses
have
incl
uded
det
aile
d a
sses
smen
ts
of s
yste
ms
and
mec
hani
sms
in p
lace
with
in t
he P
FM fr
amew
ork
to
m
anag
e ai
d fu
ndin
g.
Exte
rnal
aud
it fu
nctio
ns –
The
go
vern
men
t’s G
ener
al A
cco
unt
Rep
ort
(Con
ta G
eral
do
Esta
do)
has
bee
n su
bm
itted
to
the
M
oza
mb
ican
Ad
min
istr
ativ
e Tr
ibun
al fo
r its
ext
erna
l aud
it si
nce
2004
. Thi
s ex
tern
al a
udit
exer
cise
incl
udes
an
exha
ustiv
e an
alys
is
of t
he u
se m
ade
by
the
go
vern
men
t o
f ext
erna
l so
urce
s o
f fun
din
g
for
each
of t
he m
ain
prio
rity
sect
ors
, with
a fu
ll se
ctio
n o
f the
rep
ort
d
evo
ted
to
thi
s p
urp
ose
.
Sinc
e 20
05, t
his
anal
ysis
is n
ot
limite
d t
o O
DA
fund
ing
incl
uded
in
the
bud
get
(aid
cap
ture
d o
n b
udg
et):
the
Ad
min
istr
ativ
e Tr
ibun
al
now
als
o a
udits
fina
ncia
l inf
orm
atio
n re
ceiv
ed fr
om
the
go
vern
men
t o
n ex
tern
ally
fund
ed in
vest
men
t p
roje
cts
dire
ctly
exe
cute
d b
y d
ono
r ag
enci
es o
r lin
e m
inis
trie
s an
d w
hich
do
no
t ap
pea
r in
the
g
ove
rnm
ent
bud
get
. Thi
s ex
tern
al a
udit
is a
leg
al d
ocu
men
t an
d
fully
ava
ilab
le t
o t
he p
ublic
.
Acc
ord
ing
to
the
OE
CD
,18 in
200
5 32
.9%
of a
id w
as c
aptu
red
by
the
Ad
min
istr
ativ
e Tr
ibun
al’s
annu
al e
xter
nal a
udit
of t
he 2
005
Gen
eral
A
cco
unt
Rep
ort
. With
a fe
w e
xcep
tions
(e.g
. Can
ada,
Ger
man
y,
GFA
TM),
the
deg
ree
of a
id c
aptu
red
on
aud
it fo
llow
s si
mila
r p
atte
rns
to t
he o
ther
dim
ensi
ons
.
The
2007
Jo
int
Rev
iew
co
nsid
ers
that
the
re h
ave
bee
n co
nsid
erab
le
imp
rove
men
ts in
the
ab
ility
of t
he In
spec
tor
Gen
eral
of F
inan
ce t
o
fulfi
l its
role
of i
nter
nal a
udito
r. Fu
rthe
rmo
re, t
he n
umb
er o
f int
erna
l au
dits
thi
s ag
ency
und
erto
ok
in 2
006
is m
ore
tha
n d
oub
le t
hose
un
der
take
n in
200
5.
Ho
wev
er, b
eyo
nd t
his
spec
ific
inst
itutio
n, t
he n
atio
nal i
nter
nal
cont
rol s
yste
m –
in t
he fo
rm o
f the
var
ious
Inte
rnal
Co
ntro
l Uni
ts
esta
blis
hed
thr
oug
hout
the
Mo
zam
bic
an p
ublic
ad
min
istr
atio
n –
rem
ains
wea
k, m
ainl
y as
a re
sult
of l
ack
of fi
nanc
ial a
nd h
uman
re
sour
ces.
In it
s re
view
of t
he w
ork
und
erta
ken
by
the
Ad
min
istr
ativ
e Tr
ibun
al,
the
2007
Jo
int
Rev
iew
rep
ort
s th
at im
po
rtan
t st
eps
have
bee
n ta
ken
to im
pro
ve t
he e
xter
nal a
udit
sub
-sys
tem
: (i)
The
Ad
min
istr
ativ
e Tr
ibun
al h
as re
cent
ly c
onc
lud
ed it
s C
orp
ora
te P
lan
for
2007
–10.
(ii
) The
re h
as a
lso
bee
n a
sub
stan
tial i
ncre
ase
in it
s ex
tern
al a
udit
activ
ities
– t
he n
umb
er o
f fina
ncia
l aud
its in
crea
sed
by
148%
b
etw
een
2005
and
200
6; 5
5 au
dits
wer
e un
der
take
n, fo
r th
e fir
st
time,
at
the
dis
tric
t le
vel;
and
the
num
ber
of a
udit
pro
noun
cem
ents
un
der
take
n b
y th
e tr
ibun
al h
as g
row
n fr
om
11
in 2
005
to 4
5 in
20
06. (
iii) F
urth
erm
ore
, in
2006
the
trib
unal
was
ab
le t
o s
ubm
it its
au
dit
rep
ort
of t
he 2
005
Gen
eral
Acc
oun
t R
epo
rt t
o t
he N
atio
nal
Ass
emb
ly w
ithin
the
tim
e lim
it es
tab
lishe
d b
y la
w.
The
Join
t R
evie
w m
akes
a c
all t
o im
pro
ve t
he d
ialo
gue
bet
wee
n th
e g
ove
rnm
ent
and
the
Ad
min
istr
ativ
e Tr
ibun
al in
ord
er t
o s
olv
e th
e cr
itica
l po
ints
rep
eate
dly
bro
ught
up
in t
he t
ribun
al re
po
rts.
This
revi
ew a
lso
find
s th
at t
he w
ork
und
erta
ken
by
the
Ad
min
istr
ativ
e Tr
ibun
al c
ont
inue
s to
pre
sent
imp
ort
ant
del
ays
and
th
at it
s re
po
rts
on
the
Gen
eral
Acc
oun
t R
epo
rt a
re o
nly
mad
e p
ublic
af
ter
a le
ngth
y le
gal
pro
cess
, red
ucin
g t
he t
rans
par
ency
of t
his
aud
it ex
erci
se.
99
Mo
ZA
MB
iQu
e
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(Ho
w w
ell?
Ho
w u
sefu
l?)
exp
lana
tio
ns(W
hy/W
hy n
ot?
)
On
rep
ort
The
go
vern
men
t ha
s b
een
wo
rkin
g t
o im
pro
ve it
s re
po
rtin
g o
n b
udg
et e
xecu
tion
of e
xter
nal fi
nanc
ing
. The
Nat
iona
l Dire
cto
rate
fo
r Pu
blic
Acc
oun
ts is
no
w p
rod
ucin
g u
p t
o fo
ur q
uart
erly
Bud
get
E
xecu
tion
Rep
ort
s an
d t
he a
nnua
l co
nso
lidat
ed fi
nanc
ial s
tate
men
t –
the
Gen
eral
Acc
oun
t R
epo
rt.
Bo
th d
ocu
men
ts re
po
rt o
n b
udg
et e
xecu
tion
leve
ls fo
r ex
tern
ally
fu
nded
inve
stm
ent
pro
ject
s, a
s ac
coun
ted
for
by
the
Mo
zam
bic
an
Nat
iona
l Bud
get
Dire
cto
rate
. The
Bud
get
Exe
cutio
n R
epo
rts
follo
w
a si
mila
r st
ruct
ure
to t
hat
used
in t
he b
udg
et d
ocu
men
t, in
ter
ms
of h
ow
info
rmat
ion
is o
rgan
ised
and
pre
sent
ed –
for
inst
ance
, in
vest
men
t p
roje
cts
are
pre
sent
ed b
y so
urce
of f
und
ing
, acc
ord
ing
to
the
org
anic
cla
ssifi
er, i
nclu
din
g w
heth
er in
vest
men
t is
inte
rnal
ly o
r ex
tern
ally
fund
ed.
Acc
ord
ing
to
the
OE
CD
, in
2005
aid
cap
ture
on
offi
cial
fina
ncia
l re
po
rtin
g in
stru
men
ts re
ache
d a
lmo
st 3
5% o
f to
tal a
id. T
he s
hare
s fo
r m
ost
co
untr
ies
wer
e si
mila
r. O
nly
thre
e ag
enci
es h
ad m
ore
tha
n 70
% o
f the
ir ai
d in
clud
ed in
the
se fi
nanc
ial r
epo
rtin
g e
xerc
ises
: Ire
land
(73.
7%),
the
UK
(78.
6%) a
nd t
he G
FATM
(100
%).
Exe
cutio
n le
vels
rep
ort
ed in
the
se d
ocu
men
ts fo
r th
e ex
tern
al
com
po
nent
of i
nves
tmen
t re
fer
onl
y to
ext
erna
lly fu
nded
pro
ject
s re
gis
tere
d in
the
bud
get
and
pro
cess
ed t
hro
ugh
the
pub
lic
acco
unts
sys
tem
. A v
ery
sig
nific
ant
num
ber
of e
xter
nally
fund
ed
inve
stm
ent
pro
ject
s ar
e th
us le
ft o
ut.
The
2007
Jo
int
Rev
iew
rep
ort
s co
nsid
erab
le im
pro
vem
ents
in t
he
qua
lity
of t
he b
udg
et e
xecu
tion
rep
ort
s, p
artic
ular
ly in
rela
tion
to: t
he n
ew fo
rmat
; the
tra
nsp
aren
cy o
f inf
orm
atio
n; a
nd t
imel
y su
bm
issi
on.
Fur
ther
imp
rove
men
ts a
re e
xpec
ted
as
the
volu
me
of b
udg
et e
xpen
ditu
re d
irect
ly e
xecu
ted
thr
oug
h th
e In
teg
rate
d
Syst
em fo
r St
ate
Fina
ncia
l Man
agem
ent’s
dire
ct b
udg
et e
xecu
tion
mo
dul
e in
crea
ses.
In t
his
resp
ect,
it is
no
tew
ort
hy t
hat
in 2
006
the
Bud
get
Exe
cutio
n R
epo
rt w
as p
rod
uced
on
the
bas
is o
f inf
orm
atio
n g
ener
ated
by
this
sys
tem
.
Nev
erth
eles
s, g
reat
er e
ffort
nee
ds
to b
e p
ut in
to in
crea
sing
th
e am
oun
t o
f inf
orm
atio
n p
rese
nted
usi
ng t
he fu
nctio
nal a
nd
pro
gra
mm
atic
cla
ssifi
er in
ord
er t
o in
crea
se t
he u
sefu
lnes
s o
f the
se
rep
ort
s fo
r g
ove
rnm
ent
agen
cies
and
oth
er s
take
hold
ers.
The
2006
PE
FA re
po
rt o
n M
oza
mb
ique
find
s th
at t
he G
ener
al
Acc
oun
t R
epo
rt “
is p
rep
ared
on
a re
gul
ar b
asis
. The
det
ail o
f in
form
atio
n in
rela
tion
to re
venu
e, e
xpen
ditu
re, b
ank
bal
ance
s an
d
finan
cial
ass
ets
may
no
t al
way
s b
e co
mp
lete
but
the
om
issi
ons
are
no
t so
sig
nific
ant
as t
o u
nder
min
e its
util
ity. T
he le
vel o
f det
ail o
n in
com
e an
d e
xpen
ditu
re c
oul
d b
e m
uch
hig
her
than
it a
ctua
lly is
an
d is
exp
ecte
d t
o b
e ex
pan
ded
with
the
new
cla
ssifi
catio
n sy
stem
fo
r th
e In
teg
rate
d S
yste
m fo
r St
ate
Fina
ncia
l Man
agem
ent.
Sin
ce
2004
, the
Gen
eral
Acc
oun
t R
epo
rt h
as b
een
sub
mitt
ed fo
r ex
tern
al
aud
it w
ithin
five
mo
nths
of t
he e
nd o
f the
fisc
al y
ear.
The
stat
emen
ts
have
bee
n p
rese
nted
in a
co
nsis
tent
man
ner
ove
r th
e la
st fe
w y
ears
w
ith s
om
e d
iscl
osu
re o
f acc
oun
ting
sta
ndar
ds
used
, but
the
re a
re
po
ints
of i
nco
nsis
tenc
y w
ith in
tern
atio
nal s
tand
ard
s”.19
100
rW
An
dA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/W
hy n
ot?
)
On
pla
nTh
e cu
rren
t d
raft
Eco
nom
ic D
evel
op
men
t an
d P
ove
rty
Red
uctio
n St
rate
gy
incl
udes
one
fina
ncin
g s
cena
rio, w
hich
has
pro
ject
ions
of
bud
get
sup
po
rt a
nd p
roje
ct s
upp
ort
for
2008
–12.
Onl
y tw
o s
ecto
rs h
ave
dev
elo
ped
co
sted
med
ium
-ter
m s
trat
egie
s:
heal
th a
nd e
duc
atio
n.
Bud
get
sup
por
t – C
om
mitm
ents
are
cap
ture
d re
lativ
ely
wel
l in
the
mac
roec
ono
mic
fram
ewo
rk, w
hich
itse
lf d
efine
s th
e re
sour
ce
enve
lop
e o
n w
hich
the
bud
get
cei
ling
s ar
e b
ased
for
the
thre
e fo
llow
ing
yea
rs a
nd s
ets
the
fram
ewo
rk fo
r th
e 20
08–1
2 E
cono
mic
D
evel
op
men
t an
d P
ove
rty
Red
uctio
n St
rate
gy.
Proj
ect s
upp
ort –
The
PIP
is s
upp
ose
d t
o p
rovi
de
info
rmat
ion
to
pre
par
e th
e m
acro
eco
nom
ic fr
amew
ork
. Exp
ecte
d d
isb
urse
men
ts
fro
m p
roje
cts
for
the
next
five
yea
rs a
re c
olle
cted
in t
he P
IP. I
n th
eory
, the
PIP
sho
uld
be
the
inve
stm
ent
leg
of t
he M
TEF,
but
in
pra
ctic
e th
e tw
o p
roce
sses
stil
l lac
k co
ord
inat
ion.
Dis
tric
t lev
el –
Man
y p
roje
cts
are
imp
lem
ente
d (n
ego
tiate
d a
t ce
ntra
l go
vern
men
t le
vel o
r ca
rrie
d o
ut b
y N
GO
s) fo
r w
hich
dis
tric
ts
have
diffi
culty
co
llect
ing
info
rmat
ion.
Gui
dan
ce h
as b
een
pro
vid
ed t
o d
istr
icts
tha
t al
l pro
ject
s w
ithin
the
d
istr
ict
mus
t b
e fo
rced
to
rep
ort
to
the
dis
tric
t ad
min
istr
atio
n o
n th
eir
activ
ities
and
bud
get
for
the
com
ing
yea
r. Pr
oje
cts
that
fail
to d
o s
o s
houl
d n
ot
be
auth
oris
ed t
o w
ork
in t
he d
istr
ict.
In o
rder
to
imp
rove
co
llab
ora
tion
bet
wee
n d
istr
ict
auth
orit
ies
and
do
nors
, N
GO
s an
d o
ther
org
anis
atio
ns a
ctiv
e in
the
dis
tric
t, a
co
nsul
tativ
e fo
rum
for
all p
artn
ers
has
bee
n p
ut in
pla
ce, u
nder
the
lead
ersh
ip o
f th
e M
inis
try
of L
oca
l Go
vern
men
t, G
oo
d G
ove
rnan
ce, C
om
mun
ity
Dev
elo
pm
ent
and
So
cial
Affa
irs –
the
Jo
int
Act
ion
Foru
m.
Mo
st d
ono
rs a
re u
nab
le t
o g
ive
ind
icat
ive
com
mitm
ents
ove
r a
five-
year
per
iod
: the
fina
ncin
g fr
amew
ork
of t
he E
cono
mic
Dev
elo
pm
ent
and
Po
vert
y R
educ
tion
Stra
teg
y an
d t
he e
stim
ated
fina
ncin
g g
ap
are
ther
efo
re b
ased
on
estim
ates
mad
e b
y th
e M
acro
Dep
artm
ent
in
the
Min
istr
y o
f Fin
ance
and
Eco
nom
ic P
lann
ing
.
Bud
get
sup
por
t – M
ainl
y d
ue t
o t
he c
oo
rdin
atio
n p
roce
ss s
et u
p
aro
und
bud
get
sup
po
rt a
nd t
he n
atur
e o
f the
inst
rum
ent,
dat
a o
n ex
pec
ted
bud
get
sup
po
rt d
isb
urse
men
t ar
e o
f bet
ter
qua
lity
than
d
ata
on
exp
ecte
d p
roje
ct a
id.
Proj
ect s
upp
ort –
Pro
ject
do
cum
ents
are
no
t al
way
s m
ade
avai
lab
le
in t
ime
to t
he t
eam
in c
harg
e o
f up
dat
ing
the
PIP
(in
bud
get
and
p
lann
ing
dep
artm
ents
in t
he M
acro
Dep
artm
ent
in t
he M
inis
try
of
Fina
nce
and
Eco
nom
ic P
lann
ing
), an
d “
PIP
fiche
s p
roje
t” a
re o
ften
ei
ther
no
t co
mp
lete
d o
r no
t co
mp
lete
d a
pp
rop
riate
ly b
y p
roje
ct
coo
rdin
ato
rs.
The
clus
ter s
yste
m (S
WG
s) –
In t
heo
ry, b
y p
rovi
din
g fo
r jo
int
do
nor–
go
vern
men
t d
iscu
ssio
n at
sec
tor
or
sub
-sec
tor
leve
l, th
e cl
uste
r sy
stem
co
uld
be
the
chan
nel t
hro
ugh
whi
ch b
oth
par
ties
ensu
re
that
do
nor
pro
ject
s ar
e ad
equa
tely
tak
en in
to a
cco
unt
in s
ecto
ral
pla
nnin
g a
nd b
udg
etin
g. B
ut t
he s
yste
m h
as o
ften
no
t co
mp
lete
ly
met
thi
s o
bje
ctiv
e. T
he m
ain
reas
ons
are
the
func
tioni
ng o
f the
cl
uste
r its
elf (
som
e d
ono
rs d
o n
ot
par
ticip
ate
activ
ely,
dis
cuss
ions
fo
cus
excl
usiv
ely
on
spec
ific
po
licy
issu
es),
and
lack
of i
nfo
rmat
ion
on
the
do
nor
sid
e. F
or
exam
ple
, the
Min
istr
y o
f Gen
der
and
Pr
om
otio
n o
f Wo
men
out
lined
the
diffi
culty
in in
clud
ing
in t
heir
annu
al p
lan
and
bud
get
cer
tain
UN
pro
ject
s, s
imp
ly b
ecau
se t
hey
are
info
rmed
of t
he U
N s
upp
ort
aft
er t
he b
egin
ning
of t
he y
ear.
101
rW
An
dA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/W
hy n
ot?
)
On
bud
get
Acc
ord
ing
to
OE
CD
DA
C d
ata,
49%
of t
ota
l OD
A p
rovi
ded
to
R
wan
da
in 2
005
was
reco
rded
in t
he n
atio
nal b
udg
et. T
akin
g in
to
acco
unt
that
bud
get
sup
po
rt re
pre
sent
s ap
pro
xim
atel
y 41
% o
f to
tal O
DA
dis
bur
sed
to
Rw
and
a in
200
5, a
nd t
hat
bud
get
sup
po
rt
is 1
00%
acc
oun
ted
for
in t
he b
udg
et, t
he im
plic
atio
n is
tha
t p
roje
ct
sup
po
rt is
ver
y b
adly
cap
ture
d.
The
dra
ft P
EFA
Rep
ort
(Sep
t 20
07) s
core
s in
dic
ato
r PI
-7 (i
i) (in
com
e/ex
pen
ditu
re in
form
atio
n o
n d
ono
r-fu
nded
pro
ject
s w
hich
is in
clud
ed
in fi
scal
rep
ort
s) w
ith a
B. I
t es
timat
es t
he p
rop
ort
ion
of d
ono
r p
roje
ct d
isb
urse
men
t off
bud
get
at 6
6%, b
y co
mp
arin
g d
ata
ente
red
in
the
Dev
elo
pm
ent
Ass
ista
nce
Dat
abas
e an
d in
the
Nat
iona
l B
udg
et L
aw fo
r 20
05. I
t es
timat
es t
hat
of t
he 6
6%, a
sub
stan
tial
amo
unt
is s
upp
ort
to
NG
Os,
whi
ch s
houl
d n
ot
be
incl
uded
in t
he
natio
nal b
udg
et b
ecau
se o
f the
ab
senc
e o
f sig
nific
ant
cont
rol b
y g
ove
rnm
ent
on
how
the
se fu
nds
are
used
.
Bud
get
cei
ling
s se
nt o
ut b
y th
e M
acro
Dep
artm
ent
in t
he M
inis
try
of
Fina
nce
and
Eco
nom
ic P
lann
ing
, in
the
Bud
get
Cal
l Circ
ular
, co
ver
bo
th d
evel
op
men
t an
d re
curr
ent
exp
end
iture
s.
Dat
a co
llect
ion
on
pro
ject
sup
po
rt is
muc
h m
ore
tim
e co
nsum
ing
fo
r th
e M
acro
Dep
artm
ent
in t
he M
inis
try
of F
inan
ce a
nd E
cono
mic
Pl
anni
ng d
epar
tmen
ts a
nd li
ne m
inis
trie
s. M
ore
ove
r, p
rep
arin
g t
he
exte
rnal
ly fi
nanc
ed p
art
of t
he d
evel
op
men
t b
udg
et is
mer
ely
a d
ata
colle
ctio
n ex
erci
se, a
imed
at
pro
vid
ing
info
rmat
ion
to u
pd
ate
the
mac
roec
ono
mic
fram
ewo
rk. N
ever
thel
ess,
thi
s d
oes
no
t p
rovi
de
any
flexi
bili
ty o
n se
cto
r al
loca
tion
of O
DA
dur
ing
the
bud
get
p
rep
arat
ion
pro
cess
. Pro
ject
s in
clud
ed in
the
dev
elo
pm
ent
bud
get
ha
ve a
lread
y b
een
sig
ned
: the
cho
ice
of s
ecto
r al
loca
tion
has
alre
ady
bee
n m
ade
in a
noth
er c
ont
ext.
Unt
il th
e 20
07 b
udg
et, a
ll d
evel
op
men
t p
roje
cts
wer
e cl
assi
fied
in
the
dev
elo
pm
ent
bud
get
usi
ng a
sp
ecifi
c ec
ono
mic
cla
ssifi
catio
n w
hich
did
no
t al
low
the
sep
arat
ion
of r
ecur
rent
and
cap
ital
exp
end
iture
. The
200
8 b
udg
et w
ill s
tart
usi
ng t
he n
ew c
hart
of
acco
unts
, whi
ch n
o lo
nger
pro
vid
es fo
r a
sep
arat
e d
evel
op
men
t b
udg
et, b
ut c
lear
ly s
epar
ates
recu
rren
t an
d c
apita
l exp
end
iture
. Pr
oje
cts’
exp
end
iture
will
nee
d t
o b
e cl
assi
fied
acc
ord
ing
to
thi
s sa
me
char
t o
f acc
oun
ts, i
.e. s
epar
atin
g re
curr
ent
and
cap
ital
exp
end
iture
.
Dis
tric
ts’ b
udg
ets
– D
istr
icts
rece
ive
dire
ct s
upp
ort
fro
m d
ono
rs, i
n th
e fo
rm o
f bud
get
sup
po
rt o
r p
roje
cts,
whi
ch is
(in
theo
ry) s
how
n in
d
istr
icts
’ bud
get
s b
ut d
oes
no
t ye
t ap
pea
r in
the
Nat
iona
l Bud
get
La
w. A
sum
mar
y o
f dis
tric
ts’ b
udg
ets
is e
xpec
ted
to
be
anne
xed
to
th
e 20
08 N
atio
nal B
udg
et L
aw.
The
PEFA
rep
ort
giv
es R
wan
da
a sc
ore
of D
to
ind
icat
or
D-2
fo
r fin
anci
al in
form
atio
n p
rovi
ded
by
do
nors
for
bud
get
ing
and
re
po
rtin
g o
n p
roje
ct a
nd p
rog
ram
me
aid
.
Tim
elin
ess
– D
ono
rs o
ften
do
no
t p
rovi
de
info
rmat
ion
on
pla
nned
d
isb
urse
men
ts o
n tim
e to
be
incl
uded
in t
he b
udg
et (i
.e. b
efo
re J
uly
of t
he p
revi
ous
yea
r fo
r in
clus
ion
in t
he c
urre
nt y
ear’s
bud
get
). Le
ss
than
hal
f of d
ono
rs p
rovi
de
bud
get
est
imat
es fo
r d
isb
urse
men
ts
thre
e m
ont
hs b
efo
re t
he n
ew fi
nanc
ial y
ear
(PE
FA re
po
rt, p
.90)
.
Com
pre
hens
iven
ess
– So
me
do
nors
pro
vid
e ve
ry li
ttle
info
rmat
ion
on
pro
ject
sup
po
rt t
o t
he b
udg
et d
epar
tmen
t. In
gen
eral
, fo
r m
ultil
ater
al d
ono
rs, t
he p
rob
lem
is le
ss a
cute
, sin
ce m
ost
pro
ject
s p
rep
are
a p
lan
and
bud
get
do
cum
ent
whi
ch p
rovi
de
eno
ugh
info
rmat
ion
to u
pd
ate
the
PIP.
Relia
bili
ty o
f pro
ject
ions
– B
ecau
se s
om
e p
roje
cts
have
a p
oo
r re
cord
of p
roje
ct e
xecu
tion
and
in t
he p
ast
the
dev
elo
pm
ent
bud
get
ten
ded
to
be
und
er-e
xecu
ted
, in
the
bud
get
pre
par
atio
n p
roce
ss t
he B
udg
et D
epar
tmen
t d
isco
unts
(at
varia
ble
rat
es)
exp
ecte
d p
roje
ct d
isb
urse
men
ts.
Qua
lity
of c
lass
ifica
tion
– D
ono
r-fin
ance
d p
roje
cts
have
bee
n fo
llow
ing
func
tiona
l and
pro
gra
mm
atic
cla
ssifi
catio
ns s
ince
the
y w
ere
intr
od
uced
in 2
000.
But
pro
ject
s ar
e no
t d
ivid
ed in
to s
ub-
com
po
nent
s: c
ross
-pro
gra
mm
e, c
ross
-sec
tor
or
cro
ss-m
inis
try
pro
ject
s ar
e no
t p
rese
nted
und
er t
he s
pec
ific
pro
gra
mm
e, s
ecto
r o
r m
inis
try
to w
hich
the
y co
ntrib
ute,
but
ap
pea
r in
the
bud
get
as
one
ent
ity u
nder
the
pro
gra
mm
e, s
ecto
r an
d m
inis
try
to w
hich
the
y co
ntrib
ute
mo
st. T
his
do
es n
ot
allo
w a
n ac
cura
te o
verv
iew
of w
hich
se
cto
rs a
nd p
rog
ram
mes
the
dev
elo
pm
ent
bud
get
co
ntrib
utes
to
.
Dis
tric
ts’ b
udg
ets
– O
ne o
f the
Co
mm
on
Dev
elo
pm
ent
Fund
’s o
bje
ctiv
es w
as t
o c
oo
rdin
ate
and
co
nso
lidat
e su
pp
ort
to
inve
stm
ent
in d
istr
icts
– fr
om
go
vern
men
t tr
ansf
ers
or
fro
m d
ono
rs. S
o fa
r –
cont
rary
to
the
go
vern
men
t’s a
id p
olic
y –
the
vast
maj
orit
y o
f do
nor
sup
po
rt t
o d
istr
icts
stil
l byp
asse
s th
e C
om
mo
n D
evel
op
men
t Fu
nd
and
go
es d
irect
ly t
o d
istr
icts
.
On
par
liam
ent
Parli
amen
t ap
pro
ves
the
Bud
get
Law
as
a w
hole
, whi
ch c
om
pris
es
bud
get
sup
po
rt a
nd t
he d
ono
r-fin
ance
d p
roje
cts
pre
sent
ed in
the
d
evel
op
men
t b
udg
et.
By
law
, eac
h lo
an m
ust
be
app
rove
d b
y p
arlia
men
t. W
hile
the
M
inis
ter
of F
inan
ce a
pp
rove
s al
l ext
erna
lly fi
nanc
ed s
upp
ort
, p
arlia
men
t ap
pro
ves
onl
y lo
ans.
The
dis
cuss
ion
in p
arlia
men
t fo
cuse
s o
n re
curr
ent
bud
get
and
on
inte
rnal
ly fi
nanc
ed d
evel
op
men
t b
udg
et. T
he o
nly
aid
mo
dal
ity o
n w
hich
par
liam
ent
has
a re
al s
ay is
bud
get
sup
po
rt.
Alth
oug
h th
e ex
tern
ally
fina
nced
par
t o
f the
bud
get
go
es t
o
par
liam
ent,
it o
nly
incl
udes
pro
ject
s th
at h
ave
bee
n si
gne
d a
lread
y –
pro
ject
s th
at p
arlia
men
t ca
nno
t m
od
ify.
102
rW
An
dA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/W
hy n
ot?
)
On
trea
sury
Sing
le tr
easu
ry a
ccou
nt –
The
Law
on
Stat
e Fi
nanc
e an
d P
rop
erty
st
ates
in it
s ar
ticle
7 t
hat
“Fo
r ef
fect
ive
man
agem
ent
of t
he B
udg
et
in t
he c
entr
al G
ove
rnm
ent,
a c
ons
olid
ated
fund
sha
ll b
e es
tab
lishe
d,
whi
ch c
ons
titut
es a
ll re
venu
es a
nd o
ther
pub
lic m
oni
es, i
nclu
din
g
earm
arke
d re
venu
es o
f ext
ra b
udg
etar
y fu
nds
and
ext
erna
l lo
ans
and
gra
nts
rece
ived
in g
ener
al”
and
tha
t “T
he M
inis
ter
[of F
inan
ce]
[…] s
hall
be
the
chie
f man
ager
of t
he re
venu
es a
nd e
xpen
ditu
res
in t
he C
entr
al G
ove
rnm
ent
cons
olid
ated
fund
”; a
nd in
its
artic
le
63 t
hat
“All
rais
ed o
r re
ceiv
ed c
entr
al G
ove
rnm
ent
mo
ney
shal
l b
e cr
edite
d in
to a
sin
gle
Tre
asur
y ac
coun
t in
the
Nat
iona
l Ban
k o
f R
wan
da”
(the
sam
e is
req
uire
d fr
om
loca
l go
vern
men
ts in
art
icle
s 8
and
64)
.
Bud
get
fina
ncia
l man
agem
ent s
yste
m –
At
pre
sent
, the
bul
k o
f d
evel
op
men
t b
udg
et e
xpen
ditu
res,
fina
nced
fro
m e
xter
nal s
our
ces,
ar
e no
t re
flect
ed in
the
go
vern
men
t’s b
udg
et m
anag
emen
t sy
stem
. In
man
y ca
ses,
the
se e
xpen
ditu
res
are
not
even
kno
wn
to t
he
finan
cial
offi
cer
of t
he c
onc
erne
d li
ne m
inis
try,
as
the
info
rmat
ion
is o
ften
no
t p
rovi
ded
by
the
pro
ject
man
ager
. Mo
st p
roje
ct-
rela
ted
tra
nsac
tions
are
mad
e d
irect
ly fr
om
ind
ivid
ual p
roje
ct b
ank
acco
unts
, whi
ch a
re o
utsi
de
the
cont
rol o
f the
tre
asur
y. P
roje
ct
man
ager
s, t
here
fore
, do
no
t ne
ed o
r us
e th
e SM
AR
TGO
V s
yste
m t
o
pro
cess
the
ir p
aym
ent
req
uest
s, le
t al
one
the
ir co
mm
itmen
ts.
So fa
r o
nly
bud
get
sup
po
rt a
ctua
lly fl
ow
s th
roug
h g
ove
rnm
ent’s
ST
A.
Mo
st d
ono
r-fu
nded
pro
ject
s ar
e m
anag
ed b
y PI
Us
and
op
erat
e se
par
ate
ban
k ac
coun
ts, w
hich
rece
ive
fund
s b
oth
fro
m t
he
trea
sury
and
dire
ctly
fro
m d
ono
rs. B
ut li
ttle
or
no a
cco
untin
g o
r b
ank
reco
ncili
atio
n in
form
atio
n is
pro
vid
ed b
ack
to t
he M
acro
D
epar
tmen
t in
the
Min
istr
y o
f Fin
ance
and
Eco
nom
ic P
lann
ing
. The
20
02 R
epo
rt o
n th
e Pu
blic
Sec
tor
Ban
k A
cco
unts
list
s 1,
468
ban
k ac
coun
ts o
pen
ed in
the
nam
e o
f var
ious
go
vern
men
t in
stitu
tions
.
Sinc
e ea
rly 2
001,
the
Mac
ro D
epar
tmen
t in
the
Min
istr
y o
f Fin
ance
an
d E
cono
mic
Pla
nnin
g h
as p
rep
ared
a c
ash
bud
get
for
all
min
istr
ies
afte
r th
e b
udg
et h
as b
een
app
rove
d fo
r us
e b
y m
inis
trie
s,
pro
vinc
es a
nd a
gen
cies
. Ho
wev
er, u
ncer
tain
ties
abo
ut t
he re
leas
e o
f ext
erna
l fina
ncin
g –
in p
artic
ular
bud
get
sup
po
rt, w
hich
acc
oun
ts
for
a su
bst
antia
l pro
po
rtio
n o
f bud
get
reso
urce
s –
have
imp
acte
d
neg
ativ
ely
on
the
accu
racy
of t
he c
ash
man
agem
ent
pla
ns.
Follo
win
g t
he s
igni
ng o
f the
Par
tner
ship
Fra
mew
ork
on
Bud
get
Su
pp
ort
, bud
get
sup
po
rt d
ono
rs h
ave
atte
mp
ted
fro
nt-lo
adin
g t
heir
dis
bur
sem
ents
in t
he fi
rst
qua
rter
of t
he y
ear
to e
ase
go
vern
men
t ca
sh-fl
ow
man
agem
ent.
The
intr
od
uctio
n o
f the
STA
in t
he O
rgan
ic B
udg
et L
aw a
nd
the
go
vern
men
t’s e
ffort
s to
effe
ct t
hese
pro
visi
ons
rep
rese
nt a
si
gni
fican
t st
ep in
the
no
rmal
isat
ion
of t
he t
reas
ury
man
agem
ent
syst
em. N
ever
thel
ess,
the
bul
k o
f dev
elo
pm
ent
exp
end
iture
is
exte
rnal
ly fi
nanc
ed a
nd d
oes
no
t p
ass
thro
ugh
the
trea
sury
sys
tem
.
The
IMF’
s vi
ew is
tha
t, g
iven
the
maj
or
imp
rove
men
ts in
ban
king
se
rvic
es in
Rw
and
a, t
here
see
ms
to b
e lit
tle ju
stifi
catio
n fo
r m
aint
aini
ng s
epar
ate
com
mer
cial
ban
k ac
coun
ts fo
r p
roje
cts.
With
the
imp
rove
men
ts in
SM
AR
TGO
V, p
artic
ular
ly it
s in
stal
latio
n in
mo
st li
ne m
inis
trie
s an
d p
rovi
nces
, the
Bud
get
Uni
t is
exa
min
ing
w
ays
to e
xten
d it
s fu
nctio
nalit
y to
ena
ble
all
pro
ject
tra
nsac
tions
to
be
reco
rded
. One
of t
he o
ptio
ns is
to
req
uire
pro
ject
man
ager
s to
sub
mit
mo
nthl
y ac
coun
ts t
o t
heir
line
min
istr
y’s
finan
cial
m
anag
emen
t un
it, w
hich
wo
uld
the
n re
cord
the
m in
SM
AR
TGO
V.
The
pro
ject
man
ager
s co
uld
be
pro
gre
ssiv
ely
req
uire
d t
o s
ubm
it co
mm
itmen
t an
d p
aym
ent
ord
er re
que
sts
to b
e p
roce
ssed
thr
oug
h SM
AR
TGO
V in
the
sam
e w
ay a
s re
curr
ent
exp
end
iture
s. T
his
wo
uld
al
low
pay
men
ts t
o c
ont
inue
to
be
mad
e o
ut o
f ind
ivid
ual p
roje
ct
acco
unts
whe
re n
eces
sary
, in
acco
rdan
ce w
ith p
artic
ular
pro
ject
fin
anci
ng a
gre
emen
ts.
On
acco
unt
The
first
co
mp
rehe
nsiv
e g
ove
rnm
ent
acco
unts
and
fina
ncia
l rep
ort
w
ill b
e p
rod
uced
in 2
007,
co
verin
g 2
006.
Th
e ne
w c
hart
of a
cco
unts
pro
vid
es fo
r re
cord
ing
exp
end
iture
s o
n al
l pro
ject
s cu
rren
tly in
clud
ed in
the
dev
elo
pm
ent
bud
get
, and
cl
assi
fyin
g e
xpen
ditu
res
per
typ
e an
d s
our
ce o
f fun
din
g (l
oan
, gra
nt,
exte
rnal
ly fi
nanc
ed).
The
Law
on
Stat
e Fi
nanc
e an
d P
rop
erty
and
the
Man
ual o
f Po
licie
s an
d P
roce
dur
es fo
r Fi
nanc
ial M
anag
emen
t an
d A
cco
untin
g p
rovi
de
that
the
fina
ncia
l rep
ort
s m
ust
incl
ude
a b
ank
reco
ncili
atio
n st
atem
ent.
Thi
s is
an
ince
ntiv
e fo
r th
e B
udg
et A
gen
cy t
o re
po
rt
on
reve
nue
that
was
no
t b
udg
eted
for,
in p
artic
ular
do
nor
pro
ject
s th
at p
rovi
de
fund
s to
go
vern
men
t ac
coun
ts b
ut w
ere
not
incl
uded
in
the
bud
get
. Ove
r th
e m
ediu
m t
erm
, thi
s sh
oul
d p
rovi
de
for
pro
gre
ssiv
ely
bet
ter
cove
rag
e o
f ext
erna
lly fi
nanc
ed p
roje
cts
in t
he
dev
elo
pm
ent
bud
get
.
103
rW
An
dA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/W
hy n
ot?
)
On
aud
itTh
e O
ffice
of t
he A
udito
r G
ener
al o
f Sta
te F
inan
ces
aud
its
go
vern
men
t p
roje
cts
and
pro
gra
mm
es, b
oth
inte
rnal
ly a
nd
exte
rnal
ly fi
nanc
ed. T
he O
ffice
of t
he A
udito
r G
ener
al’s
rep
ort
on
2005
ind
icat
es t
hat
19 p
roje
cts
and
pro
gra
mm
es w
ere
aud
ited
, suc
h as
the
Mul
ti-se
cto
ral H
IV/A
IDS
Pro
ject
(MA
P) a
nd t
he D
airy
Cat
tle
Dev
elo
pm
ent
Sup
po
rt P
roje
ct (P
AD
EB
L), b
oth
fina
nced
by
the
Wo
rld B
ank.
The
Offi
ce o
f the
Aud
itor
Gen
eral
was
cre
ated
by
law
in 1
998.
In
its c
apac
ity a
s th
e co
untr
y’s
sup
rem
e au
dit
inst
itutio
n, t
he O
ffice
o
f the
Aud
itor
Gen
eral
has
the
aut
horit
y to
co
nduc
t an
aud
it o
f all
go
vern
men
t fin
ance
s an
d t
o re
po
rt t
o p
arlia
men
t o
n th
e g
ove
rnm
ent’s
ste
war
dsh
ip a
nd m
anag
emen
t o
f pub
lic fu
nds.
In 2
003,
the
Mac
ro D
epar
tmen
t in
the
Min
istr
y o
f Fin
ance
and
E
cono
mic
Pla
nnin
g a
ttem
pte
d t
o s
ubm
it a
dra
ft s
et o
f co
nso
lidat
ed
finan
cial
sta
tem
ents
for
the
2002
fina
ncia
l yea
r fo
r an
nual
aud
it,
but
a n
umb
er o
f inc
ons
iste
ncie
s p
reve
nted
the
m fr
om
bei
ng
aud
ited
. The
Offi
ce o
f the
Aud
itor
Gen
eral
no
ted
tha
t am
ong
the
sh
ort
com
ing
s w
ere
the
abse
nce
of:
(i) a
n o
pen
ing
bal
ance
and
ac
coun
ting
rep
ort
s fr
om
bud
get
inst
itutio
ns; (
ii) a
ban
k re
conc
iliat
ion
stat
emen
t o
n th
e m
ain
trea
sury
acc
oun
t; a
nd (i
ii) in
form
atio
n o
n b
alan
ces
on
go
vern
men
t b
ank
acco
unts
out
sid
e th
e co
ntro
l of t
he
trea
sury
. The
Offi
ce o
f the
Aud
itor
Gen
eral
pre
par
ed a
revi
ew o
f th
ese
acco
unts
and
mad
e a
num
ber
of w
ritte
n re
com
men
dat
ions
fo
r im
pro
vem
ent.
It p
rep
ared
sep
arat
e au
dit
rep
ort
s fo
r se
lect
ed
line
min
istr
ies
and
ass
oci
ated
ag
enci
es, w
hich
wer
e su
bm
itted
to
p
arlia
men
t to
get
her
in 2
003.
In a
dd
itio
n to
Offi
ce o
f the
Aud
itor
Gen
eral
aud
its, p
roje
cts’
ann
ual
finan
cial
sta
tem
ents
are
aud
ited
by
priv
ate
aud
iting
firm
s. T
he
aud
it re
po
rts,
with
lett
ers
do
cum
entin
g t
he a
udito
rs’ fi
ndin
gs,
are
su
bm
itted
to
go
vern
men
t, a
nd c
op
ies
are
pro
vid
ed t
o t
he re
leva
nt
do
nor.
The
PEFA
rep
ort
(ind
icat
or
PI-2
6) in
dic
ates
tha
t th
e O
ffice
of t
he
Aud
itor
Gen
eral
lack
s hu
man
cap
acity
, in
num
ber
s an
d s
kills
.
104
rW
An
dA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(H
ow
wel
l? H
ow
use
ful?
)e
xpla
nati
ons
(W
hy/W
hy n
ot?
)
On
rep
ort
The
Law
on
Stat
e Fi
nanc
e an
d P
rop
erty
req
uire
s th
e M
inis
try
of
Fina
nce
to p
rep
are
mo
nthl
y, q
uart
erly
and
six
-mo
nthl
y re
po
rts
on
reve
nue
and
exp
end
iture
. Qua
rter
ly re
po
rts
are
sub
mitt
ed t
o
Cab
inet
, and
six
-mo
nthl
y re
po
rts
to t
he C
ham
ber
s o
f Dep
utie
s.
Sinc
e 20
06, t
he C
entr
al P
roje
cts
and
Ext
erna
l Fin
ance
Bur
eau
has
pre
par
ed q
uart
erly
dev
elo
pm
ent
bud
get
exe
cutio
n re
po
rts.
In
Janu
ary
2007
, the
bur
eau
pro
duc
ed it
s fir
st a
nnua
l dev
elo
pm
ent
bud
get
exe
cutio
n re
po
rt. T
hese
rep
ort
s an
alys
e p
roje
ct
dis
bur
sem
ents
per
pro
ject
and
per
do
nor.
Ove
r th
e p
ast
year
s, fi
nanc
ial r
epo
rts
on
the
exec
utio
n o
f do
nor
pro
ject
s (C
entr
al P
roje
cts
and
Ext
erna
l Fin
ance
Bur
eau
rep
ort
s) h
ave
bee
n la
te a
nd a
re r
arel
y co
mp
rehe
nsiv
e. In
the
ab
senc
e o
f rel
iab
le
rep
ort
s, t
he B
udg
et D
epar
tmen
t us
es in
form
atio
n fr
om
pro
ject
ban
k ac
coun
ts, a
s a
pro
xy fo
r ex
pen
ditu
re d
ata,
whe
n p
rep
arin
g b
udg
et
exec
utio
n re
po
rts.
To p
rep
are
the
mo
nthl
y “fl
ash
rep
ort
s” (fi
scal
rep
ort
s), t
he M
acro
D
epar
tmen
t in
Min
istr
y o
f Fin
ance
and
Eco
nom
ic P
lann
ing
co
llect
s in
form
atio
n fr
om
ban
k ac
coun
ts a
t th
e N
atio
nal B
ank
of R
wan
da
(Ban
que
Nat
iona
le d
u R
wan
da)
to
sup
ple
men
t/re
conc
ile g
aps
in
reve
nue
and
exp
end
iture
info
rmat
ion.
One
of t
he re
aso
ns g
iven
by
the
PEFA
rep
ort
for
a C
sco
re o
n in
dic
ato
r PI
-7 (i
i) (in
com
e/ex
pen
ditu
re in
form
atio
n o
n d
ono
r-fu
nded
p
roje
cts
whi
ch is
incl
uded
in fi
scal
rep
ort
s) is
tha
t “fl
ash
rep
ort
s” a
re
gen
erat
ed a
nd d
isse
min
ated
with
out
rep
ort
ing
on
dev
elo
pm
ent
spen
din
g fu
nded
by
do
nors
.
Non
-fina
ncia
l rep
ortin
g –
It is
no
t cl
ear
whe
ther
min
istr
ies
incl
ude
do
nor-
fund
ed p
roje
cts’
ach
ieve
men
ts in
the
ir an
nual
rep
ort
s o
r no
t (re
po
rt o
n im
ple
men
tatio
n o
f ann
ual a
ctio
n p
lan,
rep
ort
on
per
form
ance
co
ntra
cts
for
dis
tric
ts).
Ther
e is
no
sys
tem
atic
or
spec
ific
rep
ort
ing
by
min
istr
ies
or
dis
tric
ts o
n th
e p
erfo
rman
ce o
f d
ono
r-fu
nded
pro
ject
s.
Rep
ort
ing
on
and
ana
lysi
s o
f pro
ject
imp
lem
enta
tion,
ach
ieve
men
ts,
and
per
form
ance
is t
he re
spo
nsib
ility
of t
he C
entr
al P
roje
cts
and
E
xter
nal F
inan
ce B
urea
u in
the
Mac
ro D
epar
tmen
t in
the
Min
istr
y o
f Fi
nanc
e an
d E
cono
mic
Pla
nnin
g. M
inis
trie
s o
ften
do
no
t re
ceiv
e th
e re
leva
nt in
form
atio
n fr
om
pro
ject
co
ord
inat
ors
.
The
PEFA
rep
ort
no
tes
(ind
icat
or
D-2
) tha
t q
uart
erly
rep
ort
s ar
e no
t p
rod
uced
by
do
nors
with
in t
wo
mo
nths
of e
nd o
f qua
rter
for
at le
ast
50%
of e
xter
nally
fina
nced
pro
ject
s in
the
bud
get
.
Som
e d
ono
rs m
entio
n co
mm
end
able
effo
rts
by
the
Cen
tral
Pro
ject
s an
d E
xter
nal F
inan
ce B
urea
u to
dev
elo
p a
nd d
isse
min
ate
a st
and
ard
re
po
rtin
g fo
rmat
for
do
nors
, to
be
com
ple
ted
eve
ry q
uart
er.
Thes
e ef
fort
s o
verla
p w
ith D
evel
op
men
t A
ssis
tanc
e D
atab
ase
rep
ort
ing
req
uire
men
ts a
nd e
ffort
s to
str
eng
then
the
co
vera
ge
of d
evel
op
men
t b
udg
et e
xecu
tion
in S
MA
RTG
OV,
but
the
y ha
ve
faci
litat
ed s
igni
fican
t im
pro
vem
ents
in t
he t
imel
ines
s an
d q
ualit
y o
f th
e b
urea
u’s
rep
ort
s o
ver
the
pas
t ye
ar.
Ho
wev
er, d
ono
rs a
lso
men
tion
the
need
to
re-t
hink
the
form
at
of r
epo
rts
on
dev
elo
pm
ent
bud
get
exe
cutio
n. T
here
is a
ne
ed fo
r m
ore
syn
thet
ic s
umm
ary
tab
les
and
for
a cr
oss
che
ck
bet
wee
n C
entr
al P
roje
cts
and
Ext
erna
l Fin
ance
Bur
eau
dat
a an
d
Dev
elo
pm
ent
Ass
ista
nce
Dat
abas
e d
ata
to s
epar
ate
wha
t ha
s b
een
put
on
the
bud
get
and
wha
t ha
s b
een
kep
t o
ut. T
his
is n
eces
sary
in
ord
er t
o a
sses
s if
pro
ject
s ar
e no
t o
n th
e b
udg
et b
ecau
se o
f (i)
lack
of i
nfo
rmat
ion;
(ii)
lack
of a
link
age
bet
wee
n th
e D
evel
op
men
t A
ssis
tanc
e D
atab
ase
and
the
dev
elo
pm
ent
bud
get
; or
(iii)
the
go
vern
men
t’s d
elib
erat
e d
ecis
ion.
105
uG
An
dA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(Ho
w w
ell?
Ho
w u
sefu
l?)
exp
lana
tio
ns(W
hy/W
hy n
ot?
)
On
pla
nA
id is
men
tione
d in
go
vern
men
t st
rate
gic
pla
ns o
nly
in v
ery
gen
eral
an
d a
gg
reg
ated
ter
ms.
The
Po
vert
y E
rad
icat
ion
Act
ion
Plan
in
clud
es a
long
-ter
m e
xpen
ditu
re fr
amew
ork
whi
ch in
corp
ora
tes
aid
p
roje
ctio
ns. S
ecto
r st
rate
gic
pla
ns a
nd t
heir
asso
ciat
ed fi
nanc
ing
st
rate
gie
s va
ry in
the
deg
ree
to w
hich
aid
is e
xplic
itly
inco
rpo
rate
d.
Pro
ject
ions
for
pro
ject
and
pro
gra
mm
e ai
d a
re in
teg
rate
d in
to t
he
thre
e-ye
ar m
acro
eco
nom
ic fr
amew
ork
, and
pro
ject
ions
for
sect
or
pro
ject
aid
are
pro
vid
ed t
o a
gen
cies
in t
he M
TEF
and
form
par
t o
f bud
get
cei
ling
s. A
id p
roje
ctio
ns a
re a
lso
inco
rpo
rate
d in
sec
tor
Bud
get
Fra
mew
ork
Pap
ers,
whi
ch a
re p
rep
ared
by
SWG
s. A
thr
ee-
year
PIP
set
s o
ut p
roje
ct a
lloca
tions
in t
he d
evel
op
men
t b
udg
et
ove
r th
e m
ediu
m t
erm
, and
ass
oci
ated
do
nor
finan
cing
.
In t
he P
ove
rty
Era
dic
atio
n A
ctio
n Pl
an a
nd s
trat
egic
pla
ns, t
he
cap
ture
of a
id in
ver
y g
ener
al t
erm
s is
larg
ely
app
rop
riate
.
Pro
ject
aid
in t
he o
vera
ll M
TEF,
whi
ch is
pre
sent
ed b
y se
cto
r, is
he
lpfu
l in
faci
litat
ing
hig
h-le
vel i
nter
- an
d in
tra-
sect
ora
l dec
isio
n m
akin
g, w
hile
bud
get
sup
po
rt is
ap
pro
pria
tely
cap
ture
d a
s a
reve
nue
onl
y. H
ow
ever
, aid
info
rmat
ion
in s
ecto
r B
udg
et F
ram
ewo
rk
Pap
ers
is m
ore
hap
haza
rd. A
id c
aptu
red
in t
he P
IP, d
esp
ite
bei
ng c
om
pre
hens
ive
and
co
nsis
tent
with
the
MTE
F (c
ove
ring
th
ree
year
s), i
s no
t p
rese
nted
in a
way
tha
t fa
cilit
ates
str
ateg
ic
reso
urce
allo
catio
n, a
s it
is ju
st a
list
of p
roje
cts
imp
lem
ente
d b
y ad
min
istr
ativ
e un
its.
Aid
dat
a p
rovi
ded
by
sect
ors
and
do
nor
agen
cies
for
the
bud
get
ye
ar is
ove
r-o
ptim
istic
; and
pro
ject
ions
for
the
out
er y
ears
usu
ally
un
der
estim
ate
futu
re le
vels
of a
id a
s ex
istin
g c
om
mitm
ents
tai
l o
ff o
ver
the
med
ium
ter
m. T
he M
inis
try
of F
inan
cial
Pla
nnin
g a
nd
Eco
nom
ic D
evel
op
men
t co
unte
rs t
his
by
(i) d
isco
untin
g a
id fo
r th
e b
udg
et y
ear
(diff
eren
t d
isco
unts
are
use
d fo
r b
udg
et s
upp
ort
and
p
roje
cts)
, and
(ii)
mak
ing
real
istic
pro
ject
ions
of b
udg
et s
upp
ort
(i.e
. p
roje
ctin
g it
ab
ove
co
mm
itted
leve
ls) o
ver
the
med
ium
ter
m.
The
pro
cess
is t
rans
par
ent,
and
sec
tor
and
nat
iona
l bud
get
fr
amew
ork
pap
ers
are
easi
ly a
vaila
ble
to
sta
keho
lder
s in
SW
Gs
and
ac
ross
go
vern
men
t.
The
finan
cial
aid
dat
a p
rovi
ded
by
do
nors
and
for
incl
usio
n in
the
PI
P an
d M
TEF
is la
rgel
y b
ased
on
finan
cial
co
mm
itmen
ts. T
he
num
ber
s fo
r th
e b
udg
et y
ear
are
ove
r-o
ptim
istic
bec
ause
do
nors
ra
rely
mee
t th
eir
com
mitm
ents
in fu
ll. A
id t
ails
off
ove
r th
e m
ediu
m
term
bec
ause
do
nors
are
unw
illin
g t
o m
ake
pro
ject
ions
ab
ove
wha
t th
ey h
ave
form
ally
co
mm
itted
to
, eve
n th
oug
h th
ey k
now
tha
t ne
w
pro
ject
s an
d p
rog
ram
mes
are
like
ly t
o a
pp
ear.
Dat
a is
wel
l pre
sent
ed in
the
nat
iona
l Bud
get
Fra
mew
ork
Pap
er,
larg
ely
bec
ause
the
Min
istr
y o
f Fin
anci
al P
lann
ing
and
Eco
nom
ic
Dev
elo
pm
ent
has
esta
blis
hed
cap
acity
and
a c
lear
and
sim
ple
sy
stem
for
sum
mar
isin
g t
he in
form
atio
n fr
om
sec
tor
Bud
get
Fr
amew
ork
Pap
ers.
Sec
tor
Bud
get
Fra
mew
ork
Pap
ers
are
less
wel
l p
rese
nted
, as
a re
sult
of t
he v
aria
ble
cap
acity
of S
WG
s an
d t
he
vag
ue g
uid
elin
es p
rovi
ded
by
the
Min
istr
y o
f Fin
anci
al P
lann
ing
an
d E
cono
mic
Dev
elo
pm
ent
in t
he B
udg
et C
all C
ircul
ar, w
hich
can
b
e in
terp
rete
d in
diff
eren
t w
ays.
In a
dd
itio
n, m
any
line
min
istr
y p
lann
ing
dep
artm
ents
hav
e lit
tle in
cent
ive
to k
eep
effe
ctiv
e tr
ack
of
the
per
form
ance
of d
iffer
ent
pro
ject
s.
106
uG
An
dA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(Ho
w w
ell?
Ho
w u
sefu
l?)
exp
lana
tio
ns(W
hy/W
hy n
ot?
)
On
bud
get
Aid
is c
aptu
red
in t
he a
nnua
l bud
get
do
cum
enta
tion
at t
hree
leve
ls:
The
bud
get
sp
eech
pro
vid
es a
sum
mar
y o
f the
MTE
F an
d
mac
roec
ono
mic
fram
ewo
rk, i
nco
rpo
rate
s o
n-b
udg
et p
rog
ram
me,
an
d in
corp
ora
tes
pro
ject
aid
to
go
vern
men
t.
The
Estim
ates
of R
even
ue a
nd E
xpen
ditu
re a
re t
he d
etai
led
ann
ual
bud
get
est
imat
es, a
nd in
clud
e th
e su
mm
ary
tab
les
fro
m t
he b
udg
et
spee
ch p
lus
det
aile
d in
form
atio
n o
f pro
ject
aid
in t
he d
evel
op
men
t b
udg
et e
stim
ates
. The
dev
elo
pm
ent
bud
get
is p
rese
nted
by
pro
ject
an
d li
ne it
em in
ter
ms
of l
oca
l allo
catio
ns, d
ono
r-fu
nded
allo
catio
ns
and
in t
ota
l. Th
e fu
ndin
g s
our
ces
are
also
pre
sent
ed –
eith
er t
he
go
vern
men
t b
udg
et o
r th
e sp
ecifi
c d
ono
r.
Min
iste
rial p
olic
y st
atem
ents
are
pre
par
ed b
y lin
e m
inis
trie
s an
d a
lso
ca
ptu
re p
roje
ct a
id in
a s
imila
r w
ay t
o B
udg
et F
ram
ewo
rk P
aper
s.
They
als
o o
ften
ap
pen
d t
he re
leva
nt e
xtra
cts
fro
m t
he d
raft
ann
ual
bud
get
est
imat
es fo
r th
e re
curr
ent
and
dev
elo
pm
ent
bud
get
, and
th
eref
ore
incl
ude
aid
.
All
pro
ject
aid
to
go
vern
men
t sp
end
ing
vo
tes
is s
upp
ose
d t
o b
e ca
ptu
red
in t
he a
nnua
l bud
get
by
law
, but
in p
ract
ice
ther
e ar
e g
aps
in m
ost
sec
tors
, and
est
imat
es s
how
tha
t th
is m
ay b
e up
to
a t
hird
. O
ther
pro
ject
aid
is le
giti
mat
ely
off
bud
get
– fo
r ex
amp
le, a
id t
o
NG
Os
and
par
asta
tals
.
The
allo
catio
ns w
hich
ap
pea
r in
the
ann
ual b
udg
et a
re b
ased
on
annu
al d
ono
r co
mm
itmen
ts r
athe
r th
an t
he d
isco
unte
d n
umb
ers
whi
ch a
pp
ear
in t
he m
acro
eco
nom
ic fr
amew
ork
and
MTE
F, b
ecau
se
it is
diffi
cult
to p
red
ict
whi
ch in
div
idua
l pro
ject
or
pro
gra
mm
e is
lik
ely
to u
nder
-per
form
.
The
dat
a in
the
bud
get
est
imat
es a
re la
rgel
y ad
min
istr
ativ
e, w
ith
sep
arat
e re
curr
ent
and
dev
elo
pm
ent
bud
get
s, m
akin
g it
diffi
cult
to g
et a
pic
ture
of o
vera
ll ai
d a
nd d
om
estic
go
vern
men
t sp
end
ing
. Pr
esen
tatio
n o
f the
dat
a in
min
iste
rial p
olic
y st
atem
ents
like
the
B
udg
et F
ram
ewo
rk P
aper
s is
no
t st
and
ard
ised
. The
se is
sues
rela
te
to o
vera
ll b
udg
et p
rese
ntat
ion
– ai
d is
wel
l int
egra
ted
in t
erm
s o
f st
ruct
ure
and
pro
cess
.
The
timin
g o
f fina
l aid
dat
a is
pro
ble
mat
ic: d
ono
rs o
ften
pro
vid
e fin
al in
form
atio
n la
te in
the
bud
get
pro
cess
, mak
ing
it d
ifficu
lt to
co
mp
are
and
reco
ncile
it w
ith t
he s
ecto
r-le
vel d
ata
pro
vid
ed in
b
udg
et s
ubm
issi
ons
.
Aid
has
bee
n in
teg
rate
d in
to b
udg
et/M
TEF
ceili
ngs.
Thi
s m
eans
th
at a
ny in
crea
ses
in p
roje
ct a
id re
sult
in a
red
uctio
n in
do
mes
tic
bud
get
allo
catio
ns t
o k
eep
a s
ecto
r w
ithin
the
cei
ling
. The
rat
iona
le
is t
o im
pro
ve a
lloca
tive
effic
ienc
y b
y en
cour
agin
g (i
) bud
get
ary
dec
isio
ns o
n th
e b
asis
of o
vera
ll re
sour
ces
to t
he s
ecto
r; a
nd (i
i) an
ex
amin
atio
n o
f ho
w s
ecto
rs s
houl
d b
e fin
ance
d. H
ow
ever
, to
avo
id
the
risk
of r
educ
ed g
ove
rnm
ent
do
mes
tic b
udg
et fu
ndin
g, s
ecto
rs
have
per
vers
e in
cent
ives
(i) n
ot
to d
iscl
ose
all
the
pro
ject
s fo
r w
hich
th
ey a
re re
ceiv
ing
do
nor
fund
s; a
nd (i
i) to
und
eres
timat
e th
e le
vel o
f th
is d
ono
r fu
ndin
g. T
his
is e
xace
rbat
ed b
y th
e fa
ct t
hat
the
bud
get
p
roce
sses
and
the
Dev
elo
pm
ent
Man
agem
ent
Syst
em re
ly o
n ai
d
dat
a g
ener
ated
by
sect
or
min
istr
ies
them
selv
es.
The
Mac
roec
ono
mic
Dep
artm
ent
of t
he M
inis
try
of F
inan
cial
Pl
anni
ng a
nd E
cono
mic
Dev
elo
pm
ent
atte
mp
ted
to
imp
rove
ai
d d
ata
by
colle
ctin
g it
dire
ctly
fro
m d
ono
rs, b
ut t
he t
imin
g o
f th
is d
ata
mak
es it
diffi
cult
to re
conc
ile w
ith t
hat
gen
erat
ed fr
om
lin
e m
inis
trie
s. N
o m
od
ality
for
do
ing
reco
ncili
atio
n ha
s b
een
esta
blis
hed
: the
dat
a fo
r m
acro
pro
ject
ions
and
for
the
annu
al
bud
get
and
PIP
rem
ain
unre
conc
iled
.
In re
cent
yea
rs, l
ittle
att
entio
n ha
s b
een
giv
en t
o h
ow
the
ann
ual
bud
get
and
min
iste
rial p
olic
y st
atem
ents
are
pre
sent
ed: t
he fo
cus
has
bee
n o
n B
udg
et F
ram
ewo
rk P
aper
s an
d t
he M
TEF
– th
e st
rate
gic
pha
se o
f the
bud
get
pro
cess
.
On
par
liam
ent
In t
he b
udg
et p
roce
ss a
nd P
FM c
ycle
, par
liam
ent
has
thre
e m
ain
stag
es o
f eng
agem
ent,
and
aid
is in
corp
ora
ted
in d
ocu
men
tatio
n in
al
l the
se s
tag
es:
Firs
t, it
revi
ews
and
mak
es re
com
men
dat
ions
to
the
exe
cutiv
e o
n th
e co
nten
ts o
f the
Nat
iona
l Bud
get
Fra
mew
ork
.
Seco
nd, i
t d
elib
erat
es o
n an
d a
pp
rove
s th
e na
tiona
l bud
get
. Th
e fo
cus
of p
arlia
men
tary
dis
cuss
ion
on
the
annu
al b
udg
et
estim
ates
are
Min
iste
rial P
olic
y St
atem
ents
. The
rele
vant
Sec
tora
l Se
ssio
nal C
om
mitt
ees
revi
ew t
he b
udg
et a
nd re
po
rt t
o t
he B
udg
et
Co
mm
ittee
.
Third
, the
Cen
tral
and
Lo
cal G
ove
rnm
ent
Pub
lic A
cco
unts
C
om
mitt
ees
revi
ew t
he A
udito
r G
ener
al’s
rep
ort
s.
Sinc
e 20
03 a
nd t
he a
do
ptio
n o
f the
Pub
lic F
inan
cial
and
A
cco
unta
bili
ty A
ct, p
roje
ct a
id h
as fo
rmal
ly b
een
vote
d o
n b
y p
arlia
men
t, fo
rmin
g p
art
of t
he a
nnua
l ap
pro
pria
tions
. Pre
vio
usly
, p
roje
ct a
id h
ad o
nly
bee
n p
rovi
ded
to
info
rm t
he b
udg
et p
roce
ss.
In a
dd
itio
n, p
arlia
men
t is
resp
ons
ible
for
app
rovi
ng lo
ans
and
g
uara
ntee
s, c
ase
by
case
. Par
liam
ent’s
Eco
nom
ic C
om
mitt
ee t
akes
th
e le
ad in
thi
s.
The
bud
get
pro
cess
pro
vid
es a
so
und
bas
is fo
r p
arlia
men
t to
sc
rutin
ise
aid
and
the
bud
get
, and
the
rele
vant
par
liam
enta
ry
com
mitt
ees
are
rela
tivel
y ac
tive.
Ho
wev
er, t
here
are
issu
es w
ith t
he
info
rmat
ion
pre
sent
ed t
o p
arlia
men
t an
d p
arlia
men
t’s c
apac
ity t
o
scru
tinis
e th
at in
form
atio
n.
Whi
le t
he n
atio
nal B
udg
et F
ram
ewo
rk P
aper
and
the
pre
sent
atio
n o
f aid
with
in it
pro
vid
es a
fairl
y us
eful
bas
is fo
r p
arlia
men
t to
co
mm
ent
on
the
exec
utiv
e’s
pro
po
sals
for
inte
r-se
cto
ral a
lloca
tions
, th
is u
sefu
lnes
s is
lim
ited
by
the
qua
lity
of t
he a
id d
ata.
The
way
in
form
atio
n is
pre
sent
ed in
Min
iste
rial P
olic
y St
atem
ents
and
th
e A
nnua
l Bud
get
Est
imat
es is
the
mo
re li
miti
ng fa
cto
r fo
r p
arlia
men
tary
dec
isio
n m
akin
g a
nd a
cco
unta
bili
ty. I
t is
diffi
cult
for
par
liam
enta
rians
to
see
wha
t p
roje
ct a
id w
ill b
e sp
ent
on
by
the
diff
eren
t sp
end
ing
ag
enci
es.
Parli
amen
t’s P
ublic
Acc
oun
ts C
om
mitt
ee o
nly
revi
ews
the
aud
its o
f p
roje
cts
in t
he A
udito
r G
ener
al’s
rep
ort
– a
min
orit
y o
f pro
ject
s –
des
pite
the
fact
tha
t th
ey v
ote
on
the
rela
ted
exp
end
iture
s.
The
pro
ble
ms
are
the
way
info
rmat
ion
is p
rese
nted
and
its
qua
lity.
Alth
oug
h fa
r fr
om
bei
ng s
tro
ng, t
he e
xist
ence
of a
par
liam
enta
ry
Bud
get
Offi
ce d
oes
hel
p in
the
qua
lity
of s
crut
iny
by
par
liam
enta
rians
, and
the
intr
od
uctio
n o
f mul
ti-p
arty
po
litic
s ha
s ad
ded
vig
our
to
the
dis
cuss
ions
.
The
Pub
lic A
cco
unts
Co
mm
ittee
do
es n
ot
dis
cuss
pro
ject
ex
pen
ditu
res
in fu
ll b
ecau
se g
ove
rnm
ent
acco
unts
and
the
Aud
itor
Gen
eral
’s re
po
rts
do
no
t ca
ptu
re a
ll p
roje
ct a
id.
107
uG
An
dA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(Ho
w w
ell?
Ho
w u
sefu
l?)
exp
lana
tio
ns(W
hy/W
hy n
ot?
)
On
trea
sury
All
pro
gra
mm
e ai
d, i
nclu
din
g G
BS
and
sec
tor
bud
get
sup
po
rt,
is d
isb
urse
d t
hro
ugh
the
Co
nso
lidat
ed F
und
usi
ng t
he t
reas
ury
syst
em in
its
entir
ety.
Bey
ond
mo
nito
ring
dis
bur
sem
ents
to
the
C
ons
olid
ated
Fun
d, t
here
is n
o t
rack
ing
of p
rog
ram
me
aid
thr
oug
h th
e ac
coun
ting
sys
tem
, as
any
earm
arki
ng is
no
tiona
l.
Alth
oug
h p
roje
ct a
id fo
rms
par
t o
f the
Co
nso
lidat
ed F
und
, it
do
es
not
use
trea
sury
sys
tem
s d
urin
g b
udg
et e
xecu
tion.
Rec
ently
, all
acco
unts
for
do
nor
pro
ject
s to
go
vern
men
t w
ere
tran
sfer
red
to
the
B
ank
of U
gan
da:
thi
s ha
s im
pro
ved
info
rmat
ion
on
aid
flo
ws
(as
wel
l as
liq
uid
ity m
anag
emen
t in
the
pub
lic s
ecto
r). T
he a
cco
unts
fall
und
er t
he s
uper
visi
on
of t
he A
cco
unta
nt G
ener
al, a
nd a
re m
anag
ed
by
thei
r re
spec
tive
go
vern
men
t sp
end
ing
ag
enci
es. S
pen
din
g
agen
cies
are
req
uire
d t
o fo
rwar
d re
que
sts
for
the
dis
bur
sem
ent
of f
und
s to
the
Acc
oun
tant
Gen
eral
, who
forw
ard
s th
e re
que
st t
o
do
nors
. Mo
st m
ajo
r d
ono
rs, b
ut n
ot
all,
resp
ect
this
.
Ann
ual d
isb
urse
men
ts fo
r g
ener
al a
nd s
ecto
r b
udg
et s
upp
ort
ha
ve b
eco
me
incr
easi
ngly
pre
dic
tab
le. S
om
e b
udg
et s
upp
ort
is
no
tiona
lly e
arm
arke
d t
o t
he P
ove
rty
Act
ion
Fund
. The
mai
n co
mm
itmen
t th
e g
ove
rnm
ent
mak
es w
ith re
spec
t to
the
Po
vert
y A
ctio
n Fu
nd is
to
ens
ure
that
bud
get
dis
bur
sem
ents
are
at
leas
t 95
% o
f bud
get
ed a
mo
unts
for
Pove
rty
Act
ion
Fund
pro
gra
mm
es.
Ho
wev
er, t
he fu
ndin
g s
our
ces
are
not
trac
ked
to
sp
ecifi
c ex
pen
ditu
re li
nes.
Dis
bur
sem
ents
of p
rog
ram
me
aid
to
the
C
ons
olid
ated
Fun
d a
re re
po
rted
on
in B
udg
et P
erfo
rman
ce re
po
rts.
The
pre
dic
tab
ility
of d
isb
urse
men
ts o
f pro
ject
aid
thr
oug
h p
aral
lel
syst
ems
is d
ifficu
lt to
ass
ess.
But
the
unp
red
icta
bili
ty o
f pro
ject
ai
d is
evi
den
t –
the
Min
istr
y o
f Fin
anci
al P
lann
ing
and
Eco
nom
ic
Dev
elo
pm
ent
sees
fit
to a
pp
ly a
dis
coun
t fa
cto
r to
pro
gra
mm
e ai
d
at t
he t
ime
of b
udg
etin
g. T
he B
ank
of U
gan
da
and
the
min
istr
y ar
e no
w a
ble
to
tra
ck d
isb
urse
men
ts t
o a
nd o
utflo
ws
fro
m p
roje
ct
acco
unts
, as
par
t o
f liq
uid
ity m
anag
emen
t. T
here
is n
o fu
lly
reco
ncile
d s
tate
men
t o
f bud
get
ed p
roje
ct a
id, a
id d
isb
urse
d a
nd
aid
use
d.
Pro
gra
mm
e ai
d is
by
defi
nitio
n o
n tr
easu
ry.
The
mo
vem
ent
of p
roje
ct a
cco
unts
into
the
Ban
k o
f Ug
and
a w
as
larg
ely
mo
tivat
ed b
y a
des
ire t
o im
pro
ve li
qui
dity
man
agem
ent,
as
the
go
vern
men
t ha
d s
ubst
antia
l id
le b
alan
ces
in c
om
mer
cial
ac
coun
ts.
On
acco
unt
Deb
t re
lief a
nd b
udg
et s
upp
ort
in a
ll its
form
s us
e g
ove
rnm
ent
acco
untin
g s
yste
ms
in fu
ll, a
nd e
xpen
ditu
res
fund
ed b
y b
udg
et
sup
po
rt a
re c
ove
red
in t
he a
cco
unts
of c
entr
al a
nd lo
cal
go
vern
men
ts. W
hile
pro
gra
mm
e ai
d a
pp
ears
in th
e re
venu
e b
udg
et,
activ
ities
fund
ed b
y p
rog
ram
me
aid
are
no
t se
par
atel
y id
entifi
able
o
n th
e ex
pen
ditu
re s
ide,
as
pro
gra
mm
e ai
d is
onl
y no
tiona
lly
earm
arke
d.
Pro
ject
aid
may
use
go
vern
men
t p
rocu
rem
ent
syst
ems,
but
p
roje
cts
at c
entr
al g
ove
rnm
ent
leve
l do
no
t an
d c
anno
t us
e th
e g
ove
rnm
ent’s
acc
oun
ting
sys
tem
s. M
any
pro
ject
s su
bm
it th
eir
acco
unts
for
aud
it la
te o
r us
e d
iffer
ent
(do
nor)
fina
ncia
l yea
rs, w
hich
co
mp
licat
es t
he a
cco
untin
g c
ycle
and
pre
vent
s sp
end
ing
ag
enci
es
fro
m p
rep
arin
g c
ons
olid
ated
pro
ject
acc
oun
ts fo
r a
giv
en fi
nanc
ial
year
. Thi
s m
eans
tha
t sp
end
ing
ag
enci
es’ a
nnua
l acc
oun
ts p
rep
ared
fo
r au
dit
do
no
t co
ver
pro
ject
aid
.
Som
e p
roje
ct a
id u
ses
loca
l go
vern
men
t ac
coun
ting
sys
tem
s (w
hich
ar
e m
anua
l) b
ut k
ept
in s
epar
ate
ban
k ac
coun
ts.
As
with
the
bud
get
and
tre
asur
y, e
xpen
ditu
res
finan
ced
by
pro
gra
mm
e ai
d a
re a
uto
mat
ical
ly c
ove
red
by
the
go
vern
men
t ac
coun
ting
sys
tem
. Pro
gra
mm
e ai
d is
no
t, h
ow
ever
, tra
cked
to
sp
ecifi
c ex
pen
ditu
res.
Ther
e is
litt
le o
r no
cap
ture
of p
roje
ct a
id in
the
go
vern
men
t ac
coun
ting
sys
tem
. It
do
es n
ot
app
ear
in t
he a
nnua
l fina
ncia
l st
atem
ents
of t
he s
pen
din
g a
gen
cy c
onc
erne
d, a
ltho
ugh
sep
arat
e p
roje
ct a
cco
unts
(usu
ally
usi
ng t
he d
ono
r sy
stem
) will
be
pre
par
ed.
Alth
oug
h th
e ac
coun
ting
sys
tem
allo
ws
the
trac
king
of e
xpen
ditu
res
by
sour
ce, t
his
is n
ot
do
ne, a
s an
y ea
rmar
king
of b
udg
et s
upp
ort
to
ex
pen
ditu
res
(via
the
Per
form
ance
Ass
ista
nce
Fram
ewo
rk o
r se
cto
rs)
is o
nly
notio
nal.
To d
ate
ther
e ha
s b
een
no m
echa
nism
to
cap
ture
do
nor
pro
ject
fu
ndin
g in
the
go
vern
men
t ac
coun
ting
sys
tem
, alth
oug
h a
pro
ject
m
anag
emen
t m
od
ule
is b
eing
dev
elo
ped
for
the
auto
mat
ed IF
MIS
, w
hich
will
be
able
to
man
age
do
nor
fund
s. T
his
is a
sym
pto
m o
f th
e p
rior
focu
s o
n th
e sh
ift t
ow
ard
s b
udg
et s
upp
ort
and
aw
ay
fro
m p
roje
cts
– al
igni
ng p
roje
cts
to g
ove
rnm
ent
syst
ems
was
no
t a
prio
rity.
Alth
oug
h p
roje
ct a
cco
unts
are
pre
par
ed, d
ono
r sy
stem
s an
d
finan
cial
yea
rs a
re o
ften
use
d. M
inis
trie
s an
d a
gen
cies
do
no
t (a
nd
are
pro
bab
ly u
nab
le t
o) p
rep
are
cons
olid
ated
ann
ual fi
nanc
ial
stat
emen
ts w
hich
incl
ude
do
nor
pro
ject
fund
ing
.
108
uG
An
dA
dim
ensi
on
evi
den
ce o
f ca
ptu
re
(Wha
t is
cap
ture
d?)
Qua
lity
of
cap
ture
(Ho
w w
ell?
Ho
w u
sefu
l?)
exp
lana
tio
ns(W
hy/W
hy n
ot?
)
On
aud
itD
ebt
relie
f and
bud
get
sup
po
rt in
all
its fo
rms
are
cove
red
in t
he
ove
rall
rep
ort
of t
he A
udito
r G
ener
al –
in t
erm
s o
f the
aud
its o
f bo
th
cent
ral g
ove
rnm
ent
vote
s an
d lo
cal g
ove
rnm
ents
.
The
Aud
itor
Gen
eral
aud
its d
ono
r p
roje
ct a
cco
unts
ind
ivid
ually
, and
si
nce
2004
/05
offe
rs o
pin
ions
on
thes
e p
roje
cts,
pre
sent
ed in
an
annu
al re
po
rt. W
here
rele
vant
, thi
s co
vers
pro
ject
aid
. In
2003
/04,
th
e A
udito
r G
ener
al w
as o
nly
able
to
aud
it tw
o-t
hird
s o
f do
nor-
fund
ed p
roje
ct a
cco
unts
: no
ne o
f the
se fo
rmed
par
t o
f the
ann
ual
rep
ort
.
Aud
it re
po
rts
sub
mitt
ed t
o p
arlia
men
t ar
e lim
ited
to
fina
ncia
l in
form
atio
n, a
ltho
ugh
som
e o
f the
pro
ject
rep
ort
s m
ay in
volv
e an
ele
men
t o
f val
ue fo
r m
one
y. A
s th
e g
ove
rnm
ent
bec
om
es
incr
easi
ngly
tim
ely
with
the
pre
par
atio
n o
f its
acc
oun
ts, s
o t
he
Aud
itor
Gen
eral
has
bee
n in
crea
sing
ly t
imel
y w
ith t
he a
uditi
ng.
Acc
ess
to a
udit
rep
ort
s is
go
od
– re
po
rts
to p
arlia
men
t ar
e in
the
p
ublic
do
mai
n an
d a
vaila
ble
on
the
web
site
, as
are
som
e ag
ency
-sp
ecifi
c ac
coun
ts.
Ther
e ar
e sp
ecifi
c p
rob
lem
s w
ith p
roje
ct a
id a
nd t
he c
ove
rag
e o
f p
roje
cts:
(i) t
he n
umb
er o
f pro
ject
s au
dite
d; a
nd (i
i) th
e fa
ct t
hat
not
all a
id is
incl
uded
in t
he b
udg
et.
The
cove
rag
e o
f do
nor
pro
ject
s is
lim
ited
by
the
num
ber
of p
roje
ct
acco
unts
ava
ilab
le fo
r au
dit
and
whe
ther
do
nors
hav
e su
pp
ort
ed a
n au
dit
of t
hose
fund
s en
do
rsed
by
the
Aud
itor
Gen
eral
. The
tim
ing
o
f the
sub
mis
sio
n o
f acc
oun
ts fo
r au
dite
d p
roje
cts
mea
ns t
hat
the
Aud
itor
Gen
eral
is u
sual
ly u
nab
le t
o in
clud
e th
em in
the
ann
ual
rep
ort
.
Alth
oug
h th
e ca
pac
ity o
f the
Offi
ce o
f the
Aud
itor
Gen
eral
is
limite
d fo
r ca
rryi
ng o
ut s
ubst
antiv
e au
dits
, the
Aud
itor
Gen
eral
is
incr
easi
ngly
inte
rest
ed in
cap
turin
g p
roje
ct s
upp
ort
, and
has
bee
n ra
isin
g is
sues
rela
ting
to
the
inco
mp
lete
ness
of t
he b
udg
et a
nd
acco
unts
in re
cent
ann
ual r
epo
rts.
On
rep
ort
GB
S an
d s
ecto
r b
udg
et s
upp
ort
are
fully
on
trea
sury
, on
acco
unt
and
on
rep
ort
, and
any
wea
knes
ses
are
sym
pto
mat
ic o
f gen
eral
g
ove
rnm
ent
finan
cial
man
agem
ent
and
acc
oun
ting
sys
tem
s.
In a
dd
itio
n to
the
ann
ual fi
nanc
ial s
tate
men
ts o
f sp
end
ing
ag
enci
es,
the
follo
win
g a
re t
he m
ain
vehi
cles
for
rep
ort
ing
on
aid
:
Proj
ect r
epor
ting
sys
tem
– A
s ea
rly a
s 20
00, t
he O
ffice
of t
he P
rime
Min
iste
r at
tem
pte
d t
o in
tro
duc
e a
pro
ject
rep
ort
ing
sys
tem
whi
ch
incl
uded
fina
ncia
l inf
orm
atio
n o
n ai
d, b
ut it
did
no
t ta
ke o
ff.
Bud
get
per
form
ance
rep
orts
– B
udg
et p
erfo
rman
ce re
po
rts,
p
rep
ared
eve
ry s
ix m
ont
hs, p
rovi
de
agg
reg
ate
info
rmat
ion
on
dis
bur
sem
ents
rela
ting
to
gra
nts
and
loan
s, p
roje
cts
and
bud
get
su
pp
ort
. In
add
itio
n, b
udg
et s
upp
ort
dis
bur
sem
ents
are
bro
ken
do
wn
by
do
nor.
Ann
ual d
evel
opm
ent c
oop
erat
ion
rep
orts
– E
ach
year
, the
A
id L
iais
on
Dep
artm
ent
of t
he M
inis
try
of F
inan
ce p
rep
ares
a
dev
elo
pm
ent
coo
per
atio
n re
po
rt, w
hich
att
emp
ts t
o p
rovi
de
com
pre
hens
ive
dat
a o
n ai
d d
isb
urse
men
ts fo
r th
e fin
anci
al y
ear.
Sect
or a
nnua
l per
form
ance
rep
orts
– S
om
e se
cto
rs, a
s p
art
of t
heir
SWA
ps,
tab
le a
nnua
l sec
tor
per
form
ance
rep
ort
s at
the
ir jo
int
revi
ews.
The
se s
et o
ut t
he o
vera
ll p
erfo
rman
ce o
f a s
ecto
r, in
clud
ing
ai
d.
The
pro
ject
rep
ort
ing
sys
tem
has
suf
fere
d fr
om
co
mp
lianc
e p
rob
lem
s, a
nd h
as n
ot
resu
lted
in c
ons
olid
ated
per
form
ance
in
form
atio
n.
Bud
get
per
form
ance
rep
ort
s p
rovi
de
agg
reg
ate
aid
info
rmat
ion
in a
us
eful
way
. But
pro
ject
info
rmat
ion
is n
ot
bro
ken
do
wn
into
sec
tors
, sp
end
ing
ag
enci
es o
r p
roje
cts,
and
onl
y re
flect
s d
isb
urse
men
ts a
nd
not
exp
end
iture
s.
Ann
ual d
evel
op
men
t co
op
erat
ion
rep
ort
s p
rovi
de
usef
ul
com
par
ativ
e in
form
atio
n o
ver
time
and
acr
oss
sec
tors
/do
nors
, b
ut t
hey
take
a lo
ng t
ime
to p
rod
uce
and
are
sub
ject
to
the
dat
a lim
itatio
ns o
f the
Dev
elo
pm
ent
Man
agem
ent
Syst
em.
Sect
or
per
form
ance
rep
ort
s co
uld
be
an im
po
rtan
t ve
hicl
e fo
r th
e d
iscu
ssio
n o
f aid
per
form
ance
, but
the
y fo
cus
mo
re o
n p
erfo
rman
ce
agai
nst
po
licy
ob
ject
ives
, rat
her
than
on
finan
cial
info
rmat
ion
and
ai
d.
The
onl
y fo
rmal
ly a
udite
d s
tate
men
ts o
f aid
are
the
ann
ual fi
nanc
ial
stat
emen
ts. A
ll o
ther
aid
rep
ort
ing
is n
ot
aud
ited
, alth
oug
h p
ast
figur
es m
ay h
ave
bee
n su
bje
ct t
o a
udit.
The
inef
fect
iven
ess
of t
he p
roje
ct re
po
rtin
g s
yste
m w
as la
rgel
y d
ue t
o t
he fa
ct t
hat
spen
din
g a
gen
cies
had
litt
le o
r no
ince
ntiv
e to
co
mp
ly w
ith t
he re
qui
rem
ents
, whi
ch w
ere
bur
den
som
e. T
here
is
usu
ally
a s
tro
nger
ince
ntiv
e to
rep
ort
to
do
nors
dire
ctly
, whi
ch
cont
ribut
es t
o fr
agm
ente
d re
po
rtin
g, w
ith d
iffer
ent
inst
itutio
ns
dem
and
ing
rep
ort
s.
The
lack
of e
xpen
ditu
re re
po
rtin
g o
vera
ll re
flect
s sh
ort
com
ing
s in
b
udg
et re
po
rtin
g a
nd n
ot
in a
id m
anag
emen
t. In
ter
ms
of p
roje
cts,
th
e ab
senc
e o
f a fu
nctio
ning
pro
ject
-bas
ed re
po
rtin
g s
yste
m
(bey
ond
the
co
mm
itmen
t co
ntro
l sys
tem
) mea
ns t
hat
cons
olid
atio
n is
diffi
cult,
alth
oug
h in
form
atio
n o
n d
isb
urse
men
ts a
nd e
xpen
ditu
res
sho
uld
be
avai
lab
le fr
om
Ban
k o
f Ug
and
a d
ata.
109
no
tes
to c
oun
try
case
stu
die
s1
IDD
& A
sso
ciat
es (2
006)
. Eva
luat
ion
of G
ener
al B
udg
et S
upp
ort:
Synt
hesi
s Re
por
t. M
ay 2
006.
Lo
ndo
n: D
FID
.
2 M
oni
torin
g S
urve
y K
enya
Rep
ort
(200
7), p
. 16–
4.
3 C
hijo
rica,
Mar
celli
na, a
nd S
wai
, To
bia
s (2
007)
. Int
egra
ting
Aid
into
the
Bud
get
ing
Mec
hani
sm o
f the
Go
vern
men
t of T
anza
nia.
Rep
ort
for t
he E
mb
assy
of
Jap
an, D
ar e
s Sa
laam
.
4 C
hijo
rica,
Mar
celli
na, a
nd S
wai
, To
bia
s (2
007)
. Int
egra
ting
Aid
into
the
Bud
get
ing
Mec
hani
sm o
f the
Go
vern
men
t of T
anza
nia.
Rep
ort
for t
he E
mb
assy
of
Jap
an, D
ar e
s Sa
laam
.
5 C
hijo
rica,
Mar
celli
na, a
nd S
wai
, To
bia
s (2
007)
. Int
egra
ting
Aid
into
the
Bud
get
ing
Mec
hani
sm o
f the
Go
vern
men
t of T
anza
nia.
Rep
ort
for t
he E
mb
assy
of
Jap
an, D
ar e
s Sa
laam
.
6 C
hijo
rica,
Mar
celli
na, a
nd S
wai
, To
bia
s (2
007)
. Int
egra
ting
Aid
into
the
Bud
get
ing
Mec
hani
sm o
f the
Go
vern
men
t of T
anza
nia.
Rep
ort
for t
he E
mb
assy
of
Jap
an, D
ar e
s Sa
laam
.
7 C
hijo
rica,
Mar
celli
na, a
nd S
wai
, To
bia
s (2
007)
. Int
egra
ting
Aid
into
the
Bud
get
ing
Mec
hani
sm o
f the
Go
vern
men
t of T
anza
nia.
Rep
ort
for t
he E
mb
assy
of
Jap
an, D
ar e
s Sa
laam
, p. 3
.
8 G
PRS
I, co
verin
g 2
003–
05,
and
GPR
S II,
co
verin
g 2
006–
09,
diff
er in
ter
ms
of
thei
r p
riorit
ies.
GPR
S II
focu
ses
mo
re o
n w
ealth
cre
atio
n an
d e
nhan
cing
g
row
th in
the
po
pul
atio
n’s
dis
po
sab
le in
com
e in
ord
er to
do
uble
the
size
of t
he e
cono
my
and
to ra
ise
Gha
na’s
per
cap
ita in
com
e to
mid
dle
inco
me
leve
l b
y 20
15. G
PRS
I fo
cuse
d m
ore
on
mac
ro s
tab
ilisa
tion
and
po
vert
y re
duc
tion
mea
sure
s.
9 C
om
par
ing
200
7 ap
pro
pria
tions
with
pla
nned
dis
bur
sem
ents
as
rep
ort
ed b
y IP
s. G
iven
tha
t th
ese
com
par
iso
ns a
re b
ased
on
agg
reg
ate
figur
es,
93%
re
pre
sent
s a
max
imum
co
nver
gen
ce b
etw
een
the
two
so
urce
s; w
hen
loo
king
at d
isag
gre
gat
ed d
ata,
the
actu
al c
orr
esp
ond
ence
bet
wee
n ap
pro
pria
tions
an
d d
isb
urse
men
ts is
like
ly t
o b
e le
ss.
10
The
bud
get
do
cum
ents
fo
r ea
ch M
DA
can
run
to
sev
eral
vo
lum
es.
Ther
e is
lim
ited
str
ateg
ic p
olic
y co
ntex
t in
clud
ed w
ith t
he b
udg
et e
stim
ates
; th
e vo
lum
es c
ons
ist
prim
arily
of l
arg
e q
uant
ities
of n
umb
ers
agai
nst,
unt
il re
cent
ly, a
ver
y la
rge
num
ber
of a
ctiv
ities
.
11
Bas
ed o
n 20
06 d
isb
urse
men
ts fr
om
info
rmat
ion
pro
vid
ed b
y d
ono
r so
urce
s.
12
Bas
ed o
n to
tal 2
006
dis
bur
sem
ents
fro
m in
form
atio
n p
rovi
ded
by
do
nor
sour
ces.
13
Imp
licitl
y G
BS,
bo
th lo
an fi
nanc
ed a
nd in
the
ory
gra
nt fi
nanc
ed.
14
Usi
ng d
ata
fro
m t
he C
ons
olid
ated
Fis
cal T
able
.
15
Taki
ng to
tal b
udg
et s
upp
ort
and
Sp
ecia
l Inv
estm
ent B
udg
et in
clud
ed in
the
2004
ann
ual B
udg
et E
xecu
tion
Law
as
a p
rop
ort
ion
of a
id re
sour
ces
rep
ort
ed
by
do
nors
.
16
Taki
ng in
to a
cco
unt
any
natio
nal b
udg
et re
po
rt (i
nclu
din
g t
he 2
006
Co
nso
lidat
ed F
isca
l Tab
le, o
n w
hich
thi
s es
timat
e is
bas
ed).
17
For
exam
ple
, de
Ren
zio
, Pao
lo a
nd S
ulem
ane,
Jo
sé (2
006)
. Int
egra
ting
Rep
ortin
g o
f PRS
and
Bud
get
Imp
lem
enta
tion:
The
Moz
amb
ican
Cas
e. L
ond
on:
O
vers
eas
Dev
elo
pm
ent
Inst
itute
(OD
I).
18
OE
CD
(20
07).
2006
Sur
vey
on M
onito
ring
the
Par
is D
ecla
ratio
n. C
ount
ry C
hap
ters
: M
ozam
biq
ue.
Dev
elo
pm
ent
Ass
ista
nce
Co
mm
ittee
(D
AC
). Pa
ris:
OE
CD
.
19
Law
son,
A. e
t al.
(200
6). A
sses
smen
t of P
ublic
Fin
ance
Man
agem
ent i
n M
ozam
biq
ue 2
004/
05. F
inal
Rep
ort
. Cur
rent
sta
tus
of P
FM s
yste
ms
and
pro
cess
es,
ove
rvie
w o
f ref
orm
s an
d p
ersp
ectiv
es fo
r 20
06. (
Mar
ch 2
006)
. SA
L an
d O
DI.
110
