1. “Aid on Budget” Study Progress Report and Preliminary Findings OECD DAC Joint Venture on PFM Paris, 20 December 2007 Stephen Lister

2. Paris, 20 December 2007"Aid on Budget" Study Progress and Preliminary Findings 2 Overview of Presentation  Study objectives  Components and outputs  Key concepts and definitions  Literature review findings  Incentives (for donors and governments)  What is the baseline?  Potential gains and what helps them  Outline recommendations  Next steps and follow up

3. Study Objectives  Commissioned by CABRI (Collaborative Africa Budget Reform Initiative) and SPA  Principal funding from DFID and JICA  Objectives: 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). Paris, 20 December 2007"Aid on Budget" Study Progress and Preliminary Findings 3

4. Study Components and Outputs Inception Report Country Reports:  5 broad brush (Group A): Burkina Faso, Ethiopia, Kenya, South Africa, Tanzania  5 in more depth (Group B): Ghana, Mali, Mozambique, Rwanda, Uganda Literature Review  to document good practice  review policies and guidelines of main aid agencies  light review of more countries (Bangladesh, Vietnam, Bolivia, Nicaragua, Botswana, Senegal) Synthesis Report Good Practice Note Paris, 20 December 2007"Aid on Budget" Study Progress and Preliminary Findings 4

5. Concept and Definitions (1) “on budget” is about budget process, not just budget documents. Hence: Paris, 20 December 2007"Aid on Budget" Study Progress and Preliminary Findings 5

6. Concept and Definitions (2) Be careful to note: Budget support is always on budget, but other aid modalities can and should also be on budget. (Different degrees of earmarking.) Identifying sources of public funds, vs. identifying uses of public funds. Different aid-disbursement channels:  Channel 1: funds disbursed via Ministry of Finance  Channel 2: funds disbursed to government, but via sector ministry or special project unit.  Channel 3: funds managed by donor, and government receives services or assets in kind. Multiple budgets (and budget-holders) in decentralised systems. Paris, 20 December 2007"Aid on Budget" Study Progress and Preliminary Findings 6

7. Paris, 20 December 2007"Aid on Budget" Study Progress and Preliminary Findings 7 Reasons for Capturing Aid  Reasons for capturing aid comprehensiveness transparency accountability strengthening government systems  Whether capturing aid at any of these points will have the desired effects depends on: quality of the information: completeness, credibility (predictability), disaggregation and detail, etc. 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.

8. Key Points from Literature Review  Strong overlap between agendas for better public finance management and for better aid effectiveness (cf. PEFA).  Paris Declaration is key pressure to bring more aid on budget. However: Different donors interpret objectives differently. Contrast between “reflecting aid in budget documents” and “integrating aid in budget processes”. Different interpretations of targets diminish the credibility of Paris Declaration monitoring survey.  Pattern of case study findings may be complex, because of: Many dimensions of “on budget”. Different interpretations by different donors. Variety of country contexts. Paris, 20 December 2007"Aid on Budget" Study Progress and Preliminary Findings 8

9. Donor perspectives and incentives  Likelihood of putting aid on budget depends on: Form of aid (grant, loan, TA, other aid in-kind) Modality (budget support, project aid) Aid partner (central/local government, NGO)  But also depends on donor characteristics: “reflectors” vs. “integrators” Importance of “visibility” Attitude to quality of public finance management Flexibility  Other factors: HQ vs. in-country perspective General vs. sector perspective Career incentives of staff Paris, 20 December 2007"Aid on Budget" Study Progress and Preliminary Findings 9

10. Government perspectives and incentives  Often conflicting interests and incentives on the government side, e.g.: Finance vs. sector ministries Centre vs. local government Vested interests of project management units.  Concerns may be about: discretion over resources reliability of disbursement through treasury  An example: perverse incentives (for government and donors) when aid is included in sector budget ceilings. Paris, 20 December 2007"Aid on Budget" Study Progress and Preliminary Findings 10

11. What is the baseline? How much aid is on/off budget?  Available data:  Diagnostics becoming more rigorous (e.g. PEFA vs. HIPC Assessment and Action Plans); Paris Declaration monitoring is the only quantified cross-country survey (2005 data).  But problems in consistency mean that comparisons across countries and over time are not reliable (Literature Review includes Paris Declaration summary across countries and donors).  PEFA Indicators D2-D3 also show weak scores.  Conclusions:  Large volumes of aid are (still) off-budget.  Budget support is (by definition) on budget; thus off-budget proportion of other modalities is often very high.  Large (but sometimes puzzling) differences in donor scores.  Lack of clear correlation between aid-on-budget and scores for quality of public finance management.  For aid effectiveness, very important to look behind the percentages and consider the quality of aid capture at different stages in the budget process. (This is the focus of our case studies.) Paris, 20 December 2007"Aid on Budget" Study Progress and Preliminary Findings 11

12. Where are the potential gains?  Budget support is automatically on budget in all dimensions. It makes sense for partner countries to create favourable conditions for (general and sector) budget support.  However, budget support has limits, and project aid has been growing. The bigger challenge, therefore, is to bring project aid on budget.  Need to consider quality of capture: timeliness, accuracy, and (above all) use of information about availability and disbursement of aid.  Also need to prioritise: Issues are more important in more aid-dependent countries. More attention to sectors where substantial aid flows support public services. Which dimension is most important for a particular instrument? Link to overall country strategy for reform of public finance management. Recognise that technical solutions won’t work unless incentives are right. Paris, 20 December 2007"Aid on Budget" Study Progress and Preliminary Findings 12

13. What helps to bring aid on budget?  Main factors that deter donors (and other stakeholders) from putting aid on budget: Weak planning process (poor costing of national strategies and government unwilling to prioritise). Budget that lacks credibility – gap between budgeted amounts and actual spending. Slow and unreliable accounting for public expenditure. Unsatisfactory procurement process (although aid can be “on treasury” but not “on procurement”).  Key factors for progress in bringing aid on budget: Strong government leadership (especially from Finance Ministry). Credible public finance management reform programme. Effective dialogue between government and donors (at general and sector levels). Donor commitment to strengthening government systems. Paris, 20 December 2007"Aid on Budget" Study Progress and Preliminary Findings 13

14. Recommendations for governments (1)  Explicit aid effectiveness strategies should: Link aid management strategy to strengthened public finance management reform (effectiveness of all resources). Go beyond indication of preference for budget support modality, to specify good donor practices and standard government procedures for other modalities. Include dialogue about the use and the additionality of aid funds (to reduce donor demand for earmarking and micro-management of donor funds). Paris, 20 December 2007"Aid on Budget" Study Progress and Preliminary Findings 14

15. Recommendations for governments (2)  Key PFM reforms include: Strengthen transparent medium term expenditure planning (to bring aid “on plan”). Clear planning and budgeting calendar. Better budget classification, including:  accurate identification of recurrent and capital expenditures (but moving away from separate “recurrent” and “development” budgets);  “source” codes to allow tracking of donor funds if required. 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 show aid sources of budget finance, and aggregate uses of aid;  to facilitate more meaningful parliamentary review. Paris, 20 December 2007"Aid on Budget" Study Progress and Preliminary Findings 15

16. Recommendations for donors  Improve medium-term aid predictability.  Support strengthened approach to public finance management.  Improve understanding of government public finance management systems and factor this into early stages of project/programme design.  Review agency-specific procedures and the extent to which they use regular government systems.  Provide aid flow information linked to government FY, government planning and budget calendar, and government budget classification. Paris, 20 December 2007"Aid on Budget" Study Progress and Preliminary Findings 16

17. Recommendations for joint action  Recognise importance of joint government-donor action (at general and sector level).  Explicitly consider different dimensions of “on budget” and monitor agency performance.  Incorporate “aid on budget” objectives in country-led public finance management strategies.  Strengthen Paris Declaration monitoring set standardised and less flexible definitions at international level; monitor proportion of project aid that is on budget; continue country-level focus on use of government systems.  Rationalise and cooperate on aid-related databases (but a database is not a free-standing solution) Paris, 20 December 2007"Aid on Budget" Study Progress and Preliminary Findings 17

18. Next Steps and Follow-up  Draft Synthesis Report (main volume) and Good Practice Note to be circulated shortly  Early 2008, feedback on drafts and finalisation of Good Practice Note  CABRI plans to make Aid on Budget a major part of its 2008 work programme. Paris, 20 December 2007"Aid on Budget" Study Progress and Preliminary Findings 18