Beyond Accra: What action should DFID take to meet our Paris and Accra commitments on aid effectiveness by 2010?

Beyond Accra: What action should DFID take to meet our Paris and Accra commitments on aid effectiveness by 2010? July 2009 Aid Effectiveness & Accountability Department (AEAD) 1

Contents Executive Summary 1. Introduction 2. The Accra High Level Forum & Accra Agenda for Action (AAA) 3. Results of 2008 Survey of Paris Declaration and progress since 2006 4. Recent evaluations of DFID s Performance in implementing Paris 5. Post-Accra priorities for DFID on increasing aid effectiveness Appendix Tables: Table 1: List of countries with DFID returns for 2006 and 2008 Paris Surveys Table 2: Results from DFID Country Offices explaining low/ unexpected scores against Paris targets 3 and 7 Table 3: Countries with 10 year Development Partnership Arrangements (DPAs) Table 4: DFID PSA countries with/without reviews of mutual accountability Annexes: A. Implementation Plan B. Accra Agenda for Action C. Bibliography Acronyms Used AAA... Accra Agenda for Action AiDA.Accessible Information on Development Activities AE... Aid effectiveness AEAD... Aid Effectiveness & Accountability Department, DFID CABRI... Collaborative Africa Budget Reform Initiative COs... Country Offices CSOs... Civil society organisations CSR... Comprehensive Spending Review DAC... Development Assistance Committee (of OECD) DPF... Divisional Performance Framework DSOs... Departmental Strategic Objectives EC... European Commission EU... European Union FA... Financial Aid HLF... High Level Forum HMT... Her Majesty s Treasury IATI... International Aid Transparency Initiative IDC... International Development Committee (of House of Commons) IMG... Independent monitoring group NAO... National Audit Office OECD... Organisation for Economic Co-operation & Development PBA... Programme Based Approach PD... Paris Declaration PFM... Public financial management PIUs... Project implementation units PPI... Publishing Project Information PRBS... Poverty Reduction Budget Support PRD... Policy & Research Division SID... Statistics on International Development SPA... Strategic Partnership with Africa TC/ TA... Technical cooperation/ technical assistance UNCAC... United Nations Convention Against Corruption UNDCF... United Nations Development Cooperation Forum WP-EFF... Working Party on Aid Effectiveness 2

Beyond Accra: Executive Summary The paper provides an overview of implications for DFID of the commitments on aid effectiveness made in the Accra Agenda for Action (AAA). It also analyses DFID s performance in the 2008 Paris Declaration Monitoring Survey. The paper recommends actions that DFID needs to take both at country level and at corporate level to meet our Paris and Accra commitments by 2010. DFID played a leading role in securing agreement to a strong, ambitious Accra Agenda for Action at the Third High Level Forum on Aid Effectiveness (HLF3) in Accra in September 2008. It is critical that DFID meet both our Paris and our Accra commitments for the following reasons: The Paris and Accra commitments are based on our experience of what makes aid effective in getting results for poor people. In the current economic climate, it becomes even more essential to ensure that every pound of aid we spend is effective and delivers value for money. DFID s Departmental Strategic Objectives for 2008/11 now entrench aid effectiveness as a corporate priority, with two of the seven objectives relating to aid effectiveness. DSO 22 reads Paris Declaration commitments implemented and targets met corporately and in country offices. To continue to be able to credibly influence other donors and partner countries, it is important that DFID demonstrates that we are fully committed to implementing not only the Paris Declaration, but also the Accra Agenda for Action, in our own policy and practice. We also need to act as a model of good practice in our own transparency and accountability, particularly given our ongoing leadership role in the new International Aid Transparency Initiative (IATI) launched at Accra. 2008 Paris Declaration Survey The 2008 DAC Paris Declaration Monitoring Survey showed that, in the subset of countries which participated in both the 2006 and the 2008 surveys, DFID is making good progress towards the Paris targets (see Table 1 of paper). We have already met seven of the 10 targets which are relevant to donors in these countries, three years ahead of schedule. Of the remaining three targets, one (avoiding parallel project implementation units PIUs) is very close to being met. On two of the indicators, aid on budget and predictability, we are on track, but concerted effort will be needed to meet these targets. The paper sets out our analysis of the reasons for weaker performance on these two indicators and makes recommendations on how to address these. Post-Accra Priorities for DFID Based on a set of criteria outlined in the paper, we have identified three priority areas in which DFID will need to take further action to ensure that our AAA commitments are met: Improving the predictability of our aid, to enable governments to plan better; Improving the transparency of our aid and ensuring all government-to-government aid is shown on partner country budgets; and Increased use of mutual accountability mechanisms at country level. 3

The paper makes 15 main commitments on work needed at corporate or regional/ country office level to address our commitments in these 3 areas. The paper also makes 13 further recommendations on other areas in the Accra Agenda for Action where work is ongoing (see Section 5 and Annex 1: Summary of Main Recommendations). Main commitments: Improving the predictability of our aid (see Section 5.5 5.13): R1. We will continue to provide rolling 3 year indicative resource allocations to partner country governments in all PSA countries where we provide resources through government. R2. We will continue to publish a comprehensive statement of future resource allocations for the current CSR period in our Annual Report. R3. Where DFID s portfolio is mainly project-based and/ or contains a high proportion of technical cooperation, country offices will review the communication of their forward financial programmes for the current and future financial year to ensure that we provide the best and most realistic estimates of expected expenditure and that these are recorded in our partner government s budget (and that these estimates are updated regularly). R4. We will commission short case studies on country-led initiatives to increase predictability in 2-3 countries, possibly including Tanzania and Malawi. Improving the transparency of our aid (see Section 5.14 5.25) R5. We will continue to report annually on our performance on aid effectiveness, based on country and regional reporting. We will aim to be a model of good practice on transparency and accountability and make use of credible independent evidence on our performance. R6. We will further enhance the section on aid effectiveness contained in the Annual Report to cover progress in implementing all the Paris and Accra commitments, including areas outside the 12 quantified Paris targets. R7. DFID country offices will work with partner governments to identify what further measures they can take to ensure that a greater proportion of aid is shown on budget, particularly for technical cooperation and other non-poverty reduction budget support (PRBS) aid modalities. Where possible, this should be part of the workplan for a locally owned public financial management reform programme. R8. We will work with the OECD DAC to address outstanding methodological problems with the Paris survey, particularly relating to cross-checking of data provided by donors and government and also to address the issue of DFID funds that are channelled to government through other donors. R9. We will adapt Country Planning Guidance to ensure that public dissemination of information at country level in a form accessible to local stakeholders is mandatory. This should be in line with the common formats and definitions agreed through the International Aid Transparency Initiative, once those are developed. R10. We will revise the Blue Book to provide clearer guidance as to when and how country offices should communicate aid to partner governments and citizens. We will promote these changes and others contained in this report through the Blue Book and 4

related guidance, making clear which measures are mandatory and which are good practice. R11. We will make information about the projects we fund available to the public both here in the UK, and in the countries we work, through the launch of a new searchable database on the DFID website. R12. We will ensure that ARIES DFID s financial, procurement and project cycle management system reports aid effectiveness information to meet Departmental Performance Framework (DPF) reporting requirements on aid effectiveness. Regional cabinets will work with country offices to ensure that ARIES is effectively capturing quality aid effectiveness data in a timely manner in order to meet DPF reporting requirements on aid effectiveness. Increased use of mutual accountability mechanisms (see Section 5.26 5.29) R13. Country offices in all countries where a recognised mutual accountability framework/ mechanism does not currently exist (according to the DAC Paris monitoring survey) to start a dialogue with other lead donor agencies and partner countries on reasons for lack of a mutual accountability framework/ mechanism and what needs to be done to develop one, based upon international best practice. R14. We will promote good practice in DFID by circulating information about successful mutual accountability mechanisms/ frameworks in which DFID currently participates, highlighting the benefits that these have achieved in terms of improved aid effectiveness. R15. We will identify the current barriers to the development of such a framework e.g. lack of government interest, lack of donor cohesion, resistance to independent monitoring of donor behaviour? Ongoing recommendations where corporate action is needed (see Section 5.30) R16. We will monitor DFID s use of parallel implementation units (PIUs) to ensure that we meet our Paris Declaration target of a maximum of 12 PIUs in 33 countries by 2010 (down from 18 PIUs in 2007). R17. In addressing new global challenges such as climate change, we will promote the use of existing aid delivery channels before new ones. R18. We will promote increased use of local and regional procurement by ensuring transparent procurement procedures that allow local and regional firms to compete, thereby helping to improve local firms capacity to compete for aid funded procurement. R19. We will ensure that development policies are consistent with agreed international commitments on gender equality, human rights, disability and environmental sustainability. R20. We will support efforts to increase the capacity of all development actors, including non-state actors, to take an active role in dialogue on development policy. R21. Our funding allocations to CSOs will be based on more rigorous performance assessments that consider the capacity of the organisation to deliver and their proven impact on poverty. 5

R22. We will monitor our implementation of the DAC Fragile States Principles as part of the collective OECD-DAC process; improve our use of flexible, rapid and long-term funding modalities for supporting early recovery; and in line with the new White Paper, we will implement a new approach to working in fragile and conflict-affected countries focused on state-building and peace-building. R23. We will ensure that Public Expenditure Financial Accountability (PEFA) assessments are country led, and where available, are used to inform DFID s Fiduciary Risk Assessment. R24. We will undertake a self-assessment of DFID against good practice on incentives for aid effectiveness to identify whether further action is needed to strengthen incentives. R25. We will promote stronger DFID engagement with the United Nations Convention against Corruption (UNCAC) to frame better country level dialogue and support around corruption. The White Paper has committed us to triple investments in efforts to track, freeze, and recover assets stolen from developing countries. We will report on our progress against these obligations and will encourage others to do the same. R26. We will participate in discussions on division of labour, given the Accra commitment to start dialogue on this in 2009. This dialogue and any resulting agreements could have important implications for which countries DFID works in and country allocations. R27. We will undertake work to better understand the impact and costs (financial and time) of the work on aid effectiveness taking place across the organisation. R28. We will consider what additional data will now be required for future Annual Reports to be able to report on our progress in meeting both our Paris and Accra commitments. 6

1. Introduction 1.1. The paper provides an overview of implications for DFID of the commitments on aid effectiveness made in the Accra Agenda for Action (AAA). It also analyses DFID s performance in the 2008 Paris Declaration Monitoring Survey and highlights progress made since 2005. 1.2. Specifically, the paper: Reviews progress since the first Paris Declaration Survey of 2006 and highlights the indicators where we still have further work to do to achieve the 2010 targets, according to the latest 2008 survey results; Discusses the main findings of the recent independent evaluation of DFID s performance in implementing our Paris commitments; Identifies the internal and external bottlenecks hindering further progress against Paris; Presents priority areas for DFID on aid effectiveness emerging from the third High Level Forum in Accra, September 2008; and Sets out the key actions that need to be taken to ensure that DFID meets our Paris commitments and additional Accra commitments both corporately and at country level by 2010. 1.3. The leadership role played by the UK at Accra and continued commitment in the new DFID White Paper, demonstrates DFID s belief that the Paris Declaration provides the right framework for managing aid to deliver better results and value for money - by driving ownership and accountability at the country level. Improving the quality and maximising the impact of aid in line with our Paris commitments will be even more important given the current global economic climate. 2. Accra High Level Forum & Accra Agenda for Action (AAA) 2.1. The Third High Level Forum (HLF3) took place in Accra in September 2008 and resulted in the adoption by all parties of the Accra Agenda for Action. DFID played a leading role in securing agreement to a strong, ambitious Accra Agenda for Action with time bound, monitorable commitments to improve aid effectiveness. 2.2. In Accra, we pushed for and achieved agreement on new commitments in four priority areas, in line with priorities agreed in advance by EU development ministers. These were (a) promoting greater use of country systems (a key aspect of alignment), (b) improving predictability of aid, (c) enhancing mutual accountability between donors and partner countries, including greater transparency, and (d) promoting a country-led division of labour amongst donors. 2.3. The resulting Accra Agenda for Action (AAA) contains 48 new commitments for the development community, of which 34 are for donors to implement. Most of these commitments build upon those already contained in the Paris Declaration. Box 1: Summary of Significant New Commitments contained in the AAA: Aid predictability and transparency provide full and timely information on annual commitments and actual disbursements; provide partner countries with regular and timely information on rolling three-to-five year forward expenditure plans. Mutual accountability ensure that mutual accountability processes are in place by 2010 in all countries that have endorsed the Paris Declaration (currently only 14 7

countries out of 54 1 covered by 2008 Paris survey are assessed as having such mechanisms in place). Use of country systems the previous EU target to channel 50% of government-togovernment assistance through partner country systems has now been adopted for all donor signatories of the AAA. Where donors are using non-country systems (including parallel project implementation units), they must state the rationale for this transparently and review the position regularly. Capacity development donors committed to further support for demand-driven capacity development to meet country needs, with Technical Cooperation (TC) selected and managed jointly with the partner country. Review procurement systems to ensure that local firms are able to compete equitably and promote south-south cooperation. Division of labour Partner countries and donors to work with the Working Party on Aid Effectiveness to complete good practice principles on country-led division of labour and to start a dialogue on international division of labour by June 2009. Engagement with civil society various measures to deepen engagement with and build capacity of civil society organisations in development processes. Adapting aid policies for fragile and post-conflict states agreement to monitor the implementation of the DAC Principles for Good International Engagement in Fragile States and Situations; joint assessment of governance and capacity; donors and developing countries to agree peace- and state-building objectives that address the causes of conflict and fragility. Improved statistics and monitoring further support agreed for national statistical systems, budgeting, planning, monitoring and country led evaluations of development performance; further work needed to strengthen partner country information systems for managing aid. (see full text of AAA in Annex B) 2.4. In the margins of the HLF, DFID led the launch of a new International Aid Transparency Initiative (IATI) which aims to deliver a step shift in public availability and accessibility of information on aid flows and activities, globally. The IATI commits donors to work with partner countries, CSOs and other users of aid information to agree, by end 2009, a set of common information standards applicable to all aid flows. Seventeen donors have already signed up to the IATI 2, and it has strong support from partner countries and CSOs. 2.5. One of the main conclusions to emerge from the Accra process is that, although progress is being made, it is not happening fast enough for donors collectively to meet their Paris commitments by 2010. 2.6. DFID demonstrated a considerable amount of political level commitment in Accra, led by the Secretary of State and Permanent Secretary. As a result, DFID will continue to lead by example on embedding aid effectiveness in our operations. Ensuring that we are a model of good practice will also strengthen our credibility when influencing others to meet their Paris commitments. 3. Results of 2008 Paris Declaration Survey 3.1. The 2005 Paris Declaration agreed 12 quantified targets for improving aid effectiveness, to be met by 2010. Progress towards these targets is measured through a survey carried out every two to three years by the OECD-DAC. Progress is measured on 1 Mongolia had no result on mutual accountability in the 2008 survey. 2 Germany, Netherlands, World Bank, EC, UNDP, UK, Norway, Finland, Ireland, Spain, Denmark, Sweden, Switzerland, Hewlett Foundation, GAVI, Australia and New Zealand. 8

a country by country and donor by donor basis. The main source of information for the survey is a questionnaire filled out at country level by donors and partner governments. 3.2. For a list of the countries that participated in the 2006 and 2008 surveys for which DFID specific returns were submitted to OECD/ DAC, please see Appendix Table 1. 3.3. The 2008 survey showed that, in the subset of countries which participated in both the 2006 and the 2008 surveys, DFID is making good progress towards the Paris targets (see Table 1 below). We have already met 7 of the 10 targets which are relevant to donors in these countries, three years ahead of schedule. This includes the target on use of Programme Based Approaches (Indicator 9), an area in which DFID s progress appears to have been strong and, based upon the definition of PBAs used in the 2008 Survey, the target of 66% appears to have been met. 3 Of the remaining three targets, one (avoiding parallel project implementation units PIUs) is very close to being met. However, on two of the indicators, aid on budget and predictability, while we are on track, a concerted effort will be needed to meet these targets. The survey also shows that we have already met the new target, agreed at Accra, on use of country systems. 3.4. Performance in those countries which only entered the survey process in 2008 is less strong. In this group of countries, we have not yet met the targets on TA, programme based approaches or joint analytical work. We also retain a number of parallel project implementation units (PIUs). These countries are not counted in measurement of progress towards the Paris targets in the DAC survey 4, but we should nevertheless ensure we are improving aid effectiveness in these countries. 3.5. We also need a concerted effort to meet the EU aid effectiveness targets agreed in 2005, committing us to providing 100% of TA through co-ordinated country led programmes and not to develop any new PIUs. See Table 2 below. 3.6. The next sub-section sets out the summary results for the Paris indicators on aid on budget (indicator 3) and predictability (indicator 7.) It analyses the key factors explaining these results, drawing on quantitative analysis and on results from questionnaires sent to selected country offices. Recommendations for steps that should be taken by DFID, working with partner governments, to ensure we meet these targets by 2010 are contained in Section 5 of this paper. How is DFID performing on aid on budget and predictability? Aid on budget (Indicator 3) 3.7. As Table 1 shows, across all 32 DFID Country Offices (COs) participating in the 2008 survey, aid on budget was on average 58%. For the 20 countries participating in both 2006 and 2008 surveys, the average was 65%. This is an improvement from 45% in 2005, but remains some way off the agreed 85% Paris target. 3.8. It should be noted that changes to the methodology for calculating both targets since the 2006 survey make them more challenging for DFID. In 2006, progress was measured at corporate level (using a weighted average across countries). This meant that larger countries had more of an impact on the results than smaller countries. In the 2008 survey, this indicator is measured at country level using an unweighted average. DFID 3 We wish to be cautious at this stage in declaring that DFID has definitely met the PBA target as the definition of PBAs used in the Paris survey has been changed and so this result may overstate progress somewhat. 4 This is because progress is measured according to the 2005 baseline and because some of the targets are calculated according to that baseline. 9

scores considerably lower at the country rather than at the corporate level. The decision to move from the use of weighted to unweighted averages was taken collectively by DAC members at a meeting in June 2008. 3.9. It is important to note that aid on budget is not the same as use of country budget systems. This target measures only whether aid is shown on the partner country s budget, regardless of whether it is channelled through budget systems. The indicator is measured by comparing the DFID estimate of total aid actually disbursed in 2007 with the partner government s estimate of aid flows from DFID shown on their budget. 3.10. The aggregate results mask considerable variation between DFID offices. In 13 countries out of 32, aid on budget is shown at 0%, in another 3 it is less than 50%. Only 8 DFID offices 5 appear to have met the target (although as discussed below, methodological issues may understate progress). Countries that score very poorly tend to be those with very small programmes, and those with a high share of TA. Countries with a high share of aid on budget tend to have relatively stronger PFM systems, high use of country systems, larger aid programmes and greater use of budget support. 5 Bangladesh, Ghana, Krygz Republic, Nepal, Rwanda, Tanzania, Vietnam and Zambia 10

Table 1: Summary Paris Declaration Monitoring Survey results for DFID 3 4 5a 5b 6 7 Indicators Aid flows are aligned on national priorities Strengthen capacity by coordinated support Use of country public financial management systems Use of country procurement systems Definitions Aid for government sector in budget (USD m) Aid disbursed for government sector (USD m) Coordinated Technical co-operation (USD m) Technical co-operation (USD m) Use of PFM systems (USD m) Aid disbursed for government sector (USD m) Use of procurement systems (USD m) Aid disbursed for government sector (USD m) Number of parallel PIUs All countries 2007 33 b countries 1 269 1 620 252 519 1 072 1 620 953 1 620 Avoid parallel implementation 45 structures Number of countries 32 Aid is more predictable Aid recorded as disbursed (USD m) Aid scheduled for disbursement (USD m) 1 357 1 851 Untied aid (USD m) 2 379 8 Aid is untied Total bilateral aid (USD m) 2 379 Programme-based approaches (USD m) 1 257 9 10a 10b Use of common arrangements or procedures Joint missions Joint country analytic work Total aid disbursed (USD m) 2 024 Number of joint missions (number) 170 Total number of missions (number) 291 Number of joint analyses (number) 100 Total number of country analyses (number) 164 Indicator values 2005 2007 33 c countries All d countries Illustrative 2010 Targets 45% 65% 58% 85% 56% 66% 48% 78% 77% 66% 78% 68% 59% Target of 50% achieved 50% (EU Target) 50% (Eu Target) 37 18 45 14 46% 60% 54% 73% 100% 100% 100% Target achieved 61% 71% 62% 66% 46% 61% 58% 69% 69% 61% Notes to Table 1: a. Green = 2010 target already met; Amber = on track to meet 2010 target (based upon continuation of progress achieved 2005-07). b. The 2006 Paris Monitoring Survey covered a total of 33 countries. In 2006, 22 of these countries had a DFID programme, accounting for 48% of country programmed aid. The 2008 Survey covered a total of 55 countries. 32 of these had a DFID programme, accounting for 61% of country programmed aid. c. The DAC calculate progress towards the Paris targets on the basis of the countries that participated in both surveys (i.e. against a sub-set of the 33 countries in the original baseline). Within this sub-set of 33 countries, there were returns from 20 DFID country offices, 16 of which were PSA countries. See Appendix Table 1 for the full list d. 55 countries participated in the 2008 survey, 32 of which are recipients of DFID aid. The results for all countries refer to this group. It is worth noting that performance in this wider group of countries is generally weaker than in the 20 countries for which a baseline was set in 2005. e. For all indicators apart from 3 and 7, indicator values are measured on the basis of a weighted average. This means that larger programmes are given greater weight. f. For indicators 3 and 7, indicator values are calculated on the basis of unweighted averages. This means that each country is given equal weight, regardless of size. f. The figures reported in this table are those issued by DAC in November 2008, and are consistent with those reported online by the DAC in March 2009. These figures update those used Target of 40% achieved Target of 66% achieved 11

in Accra and included in DFID s 2008 UK Progress Report. There are therefore some small variations in the final revised survey results as published by OECD/ DAC after the Accra meetings, particularly for indicators 3 and 7. See http://www.oecd.org/dataoecd/21/13/41564954.xls Table 2: DFID Performance against EU Aid Effectiveness Commitments Target 2005 (33 countries) 2007 (33 countries) 2007 (all countries) Providing 100% of 56% 66% 48% capacity building through coordinated programmes Channelling 50% of 78% 77% 66% govt-to-govt assistance through country systems Avoid establishing new PIUs N/A Not known. The total number of PIUs has increased in 2 countries (Nicaragua and Kenya), but stayed constant or fallen in N/A Reduce uncoordinated missions by 50% 54% missions not joint all others. 39% missions not joint 42% missions not joint Notes: a. Green = 2010 target already met; Amber = on track to meet target by 2010; red = off track to meet target by 2010. Predictability (indicator 7) 3.11. As Table 1 shows, across the 32 countries where DFID had programmes, on average 54% of aid that was scheduled by DFID for disbursement in 2007 was recorded by partner countries as having been received. Across the 20 countries where DFID had programmes and which participated in both surveys, this figure was 60%. This was an improvement from the 46% registered in 2006. DFID has not yet reached the 73% target for 2010. 3.12. In the Paris survey methodology, predictability is measured by comparing aid scheduled by donors for disbursement in 2007 with disbursements actually recorded by government. A low score can therefore be either a result of DFID not disbursing what it scheduled to disburse, or of poor data capture by the partner country government. This means that there is some overlap with indicator 3. Our analysis showed that in 9 cases, what appeared to be a low score for predictability was actually a result of poor data capture by government, rather than DFID failure to disburse on time. Comparing only DFID aid scheduled for disbursement with actual disbursements, DFID s score rises to 87.5% 6 The main priority in meeting the Paris target is therefore to work with partner governments to improve data capture. However, we will also improve performance by better matching scheduled and disbursed aid. 6 This is based upon aid scheduled for disbursement of US$1,851,000 against aid actually disbursed of US$1,620,000 in 2007. Calculations exclude Liberia. 12

Why do some countries score poorly against these two targets? 3.13. In seeking to better understand the Paris survey results, AEAD requested DFID country offices with low or unexpected scores against the Paris targets to complete a short questionnaire explaining the findings. The results are summarised in Appendix Table 2. Using these surveys in combination with our own analysis, we have identified six primary reasons for low scores: o o o o o o Methodological issues (both indicators) Poor data capture by partner governments (both indicators) Small programmes (aid on budget) Aid modalities, in particular high share of Technical Cooperation (aid on budget) Poor forecasting by DFID (predictability) Breach of conditionality by the partner government (predictability) Methodological problems 3.14. Methodological problems in the Paris survey were identified by a number of country offices. A common problem is aid channelled through other donors (e.g. through delegated co-operation agreements) or through joint trust funds. Partners will often include this in their budget as being channelled from other donors rather than DFID, but DFID will record it as having been disbursed. This explains low scores on aid on budget in Cambodia and Vietnam, for example. Exchange rate changes were also cited as a problem. Poor data capture by governments 3.15. Progress towards both indicators depends on actions taken by partner governments as well as those taken by DFID. A common response (raised by 4 country offices) was that the partner country did not break down aid within the budget by donor. Thus, while DFID aid may be included in the budget on aggregate, it is not identified as being from DFID. Partner countries therefore entered a zero score in the Paris survey return. Problems in accurately recording aid by the government were also cited in three cases. Small programmes 3.16. DFID s overall results on aid on budget are lowered by the large number of very small programmes where there is very little aid on the budget. Of the 11 countries where DFID aid was less than $5m, 9 (82%) had less than 10% aid on budget. Conversely, of the 20 countries with aid of more than $5m, 16 (80%) had more than 10% of aid on budget. In some of these smaller programmes, DFID does not have an office (e.g. Colombia, Philippines). Use of Technical Assistance 3.17. As we might expect, we found a positive correlation between use of both general and sector PRBS and aid on budget, although a number of countries with low shares of PRBS also managed high shares of aid on budget 7. Conversely, countries with very high levels of Technical Cooperation (TC) appeared to have very low shares of aid on budget. In particular, the five countries with more than 40% of aid as TC all had zero aid on budget scores. 8 7 Kenya, Bangladesh, Nepal, Sudan, Afghanistan 8 Ukraine, Nigeria, Albania, Burundi and Kosovo. 13

Poor forecasting for project aid 3.18. Predictability of budget support appears to be high, according to DFID s Annual Report. The main challenge appears to be with project aid. Four of the ten DFID offices who responded to our questionnaire cited delays in DFID funded projects as an explanation for divergences between actual and expected disbursements. Two countries observed that there had been unanticipated increases in funding during the year in question. Conditionality 3.19. DFID has a responsibility to ensure that our funds are used for the purposes intended, and we can and do withhold funds where partnership commitments are breached or targets are not met. Feedback from country offices suggested that this explained lack of predictability in five cases. This was fairly evenly spread between breach of the partnership commitment on good governance, anti-corruption and PFM (2 cases), human rights (1 case), poverty reduction (1 case.) Another cited general underperformance on conditions agreed with the partner government in the Performance Assessment Framework. Main Findings 3.20. The main finding from our analysis is that making further progress against these particular Paris targets will require us to work with governments to address the constraints to effectively capturing aid data on budgets both expected and actual disbursements. Where possible, this should be undertaken as part of broader, country-led, public financial management reform programmes. 3.21. DFID Country Offices also need to consider what further measures they could take to strengthen dialogue at country level to ensure that a greater proportion of aid is shown on budget, particularly for TA and other non-prbs aid modalities 9. This should include all DFID aid to a particular country, including resources that are managed by central spending departments which are delivered outside of the country programme via trust funds or other instruments. DFID s Departmental Strategic Objectives 3.22. DFID s Departmental Strategic Objectives (DSOs) for 2008/11 now entrench aid effectiveness as a corporate priority and include targets related to the implementation of Paris Declaration commitments by 2010. 3.23. Two of DFID s seven departmental strategic objectives are about aid effectiveness: to make all bilateral and multilateral donors more effective (DSO 5) and to deliver high quality and effective bilateral development assistance (DSO 6). Indicators to measure achievement of these objectives include improved global performance against Paris Declaration commitments (DSO 5.16); and Paris Declaration commitments implemented and targets met corporately and in country offices (DSO 6.22). Divisional Performance Frameworks (DPFs) set specific Paris-related targets for divisions to achieve as their contribution to meeting the DSO s. 9 A useful source of further information and detailed recommendations on improving the comprehensiveness of budget coverage is the CABRI and SPA Study Putting Aid on Budget, April 2008. 14

4. Additional analysis of DFID s Performance in Implementing Paris 4.1. There have been a number of recent reports and evaluations that have examined DFID s performance in implementing the commitments that we entered into under the 2005 Paris Declaration. These form an important part of the evidence base about where we need to do more in future. Evaluation of DFID HQ s Performance on Implementing Paris Declaration 4.2. As part of the lead up to Accra, an independent evaluation was undertaken of DFID headquarter s performance in implementing the Paris Declaration since 2005. The evaluation assessed DFID s institutional capability to meet its Paris Declaration commitments. It was part of a global evaluation of Paris Declaration implementation. 4.3. The evaluation found that the core enabling conditions (commitment, capacity and incentives) are strongly developed and embedded across DFID. Policies, systems and procedures introduced over the last decade mean that DFID approaches the Paris Declaration not so much as a set of external obligations, but as a tool to achieve its own corporate objectives. For this, DFID was commended. 4.4. However, the evaluation found that current systems did not deliver complete consistency across all country offices, nor against all aspects of the Paris Declaration. Recommendations were put forward to allow DFID to do so, including: Introduce explicit transparency objectives for corporate, country and programme information; Strengthen internal and external reporting on implementation of the Paris Declaration commitments; Undertake explicit consideration of the implications of any new global spending targets and initiatives for the Paris agenda; Clarify how the Paris principles will apply to new international initiatives and global public goods, including climate change; Prepare an annual report on aid effectiveness, and include summary findings and data in DFID s Annual Report to Parliament. 4.5. A second phase of the global evaluation of the Paris Declaration is planned for 2009 and will focus on measuring the impact of implementation of the Paris commitments on development outcomes. IDC report on DFID s performance on aid effectiveness 4.6. The International Development Committee (IDC) of the House of Commons produced a special report in July 2008 in the lead up to Accra entitled Working Together to Make Aid More Effective 10. 4.7. Their report recognised that the UK has performed well against almost all of the targets in the Paris Declaration on Aid Effectiveness. It also recognised that the OECD DAC had acknowledged UK leadership on aid effectiveness issues since Paris. 4.8. However, the report also highlighted a number of areas where the IDC felt that further efforts were needed by DFID. These included (a) developing an evidence base to 10 DFID s input to the IDC enquiry is contained in a UK Government Memorandum for the International Development Committee, Co-ordination for Aid Effectiveness, February 2008. 15

demonstrate whether transaction costs are actually falling for recipient countries as a result of donor alignment and harmonisation efforts; (b) working more effectively in partnership with other donors in country and adapting our approach to country circumstances; (c) developing more effective mechanisms to monitor progress against a range of targets linked to country ownership; (d) defining ownership more broadly than government ownership to reflect democratic processes and greater involvement of parliaments, civil society and citizens; (e) using aid more effectively to support statebuilding and peace-building in fragile and post-conflict states; (f) further work with the European Commission to pilot their Code of Conduct on Division of Labour in selected countries. 5. Post-Accra priorities for DFID on increasing aid effectiveness Introduction 5.1. In this section, we recommend actions that DFID needs to take both at country and corporate level to meet our commitments. Our priority remains country-driven systems for increasing the effectiveness of our bilateral assistance. This will involve working with our partner countries to put in place locally developed and owned systems to ensure that all donors meet the Paris and Accra targets. We believe that the changes required to DFID corporate systems will have the greatest impact if they facilitate such demand-driven country level reform processes. 5.2. In recommending the actions that DFID needs to take to ensure implementation of Paris and Accra commitments in country, we need to consider what other donors are doing at country level, as DFID does not act in isolation. As far as possible, we should identify actions that help both DFID and other donors to achieve their targets, given our DSO on making all bilateral and multilateral donors more effective. 5.3. Based upon the analysis in this report and recent evaluations of DFID s progress, we have identified three priority areas where DFID will need to focus to ensure that our Paris and Accra commitments are met. These priorities were identified on the basis that they are areas in which: We have not yet met the Paris/Accra commitment and are some way from doing so; We do not already have work underway (either bilaterally or with other donors) that will enable us to meet the commitment; Meeting the commitment will substantially improve the effectiveness and impact of DFID s aid; DFID pushed for an ambitious agreement at the Accra HLF, so failing to meet the commitment could affect our reputation and credibility; Analytical work such as the DFID Paris Declaration evaluation recommended further action. Meeting the commitment is politically and technically feasible. 5.4. There are also a number of other commitments which we need to meet, but which are lower priority according to these criteria. Priorities for DFID to Implement the Accra Agenda for Action and 2010 Paris Targets 5.5. Based on the above criteria, we have identified three priority areas in which DFID will need to take further action to ensure that we meet our Paris and Accra commitments: Improving the predictability of our aid, to enable governments to plan better; 16

Improving the transparency of our aid and ensuring all government-to-government aid is shown on partner country budgets; and Increased use of mutual accountability mechanisms at country level. Priority 1: Improving the predictability of our aid both in year and in the medium term. 5.6. The AAA states that beginning now, donors will provide developing countries with regular and timely information on their rolling 3-5 year forward expenditure and/or implementation plans, with at least indicative resource allocations that developing countries can integrate in their medium-term planning and macroeconomic frameworks.donors will address any constraints to providing such information. (Para 26c.) 5.7. DFID is a strong supporter of initiatives to improve predictability. The UK is committed to increase aid spending to achieve 0.7% of GNI on ODA by 2013 in order to meet our international commitment and we are on track to meet the EU commitment of 0.56% ODA/GNI by 2010. 5.8. As an important part of in-country dialogue, we already provide indicative rolling allocations to partner countries. This provides partner countries with useful information about likely future allocations according to each country context. It is often part of the response to requests from the partner government, IMF or country aid coordination processes. This is done on the basis that these indicative resource allocations may be subject to changes in the future. 5.9. Providing information on rolling 3-5 year expenditure plans is already the operational norm for many of our country offices, particularly where we provide budget support. 5.10. We use 10-year Development Partnership Arrangements (DPAs). DFID currently has DPAs with 10 countries.11 These provide an opportunity, through an overarching bilateral arrangement, to show a long term UK commitment, including delivery of more predictable and better aid. 5.11. Under the new country planning guidance, Country Offices (COs) are required to project likely resources over a multi-year period. When we publish country plans, we publish indicative resource allocations for the Current Spending Review (CSR) period and, in some cases beyond this. A Country Plan may extend beyond the CSR period where necessary, to align with national development strategies. 5.12. We publish forward projections for all country programmes for the 3 years of the current spending review period in our Annual Report. Depending on the stage in the Comprehensive Spending Review (CSR) cycle this for one, two or three years in future. 5.13. Internationally, we are working to support the EUs MDG contract, an instrument that will provide long term (6 years) predictable financing for the MDGs for an eligible set of countries. 11 Rwanda, Mozambique, Sierra Leone, Uganda, Zambia, Vietnam, Pakistan, Afghanistan, Yemen and Ghana. 17

Box 2 Estimates of the Costs of Unpredictability of Aid Aid is twice as volatile as domestic country revenues. Unpredictable aid increases macro economic instability and complicates medium-term expenditure planning. Unanticipated aid shortfalls can force developing countries to borrow more than they had planned to, often with a negative impact on economic growth. Unpredictable aid can increase the cost of public procurement in health in some cases by over 20 per cent, and decrease utilisation of primary facilities (in some cases by up to 70 per cent). Inability of donors to pre commit funds over 3-5 years has hampered the scale up of cost-effective and more equitable primary health services. Source: Hudson and Mosley (2007); Strategic Partnership for Africa (SPA) survey; IMF Independent Evaluation Office 2007 Recommendations: R1. We will continue to provide rolling 3 year indicative resource allocations to partner country governments in all PSA countries where we provide resources through government. R2. We will continue to publish a comprehensive statement of future resource allocations for the current CSR period in our Annual Report. R3. Where DFID s portfolio is mainly project-based and/ or contains a high proportion of technical cooperation, country offices will review the communication of their forward financial programmes for the current and future financial year to ensure that we provide the best and most realistic estimates of expected expenditure and that these are recorded in our partner government s budget (and that these estimates are updated regularly). R4. We will commission short case studies on country-led initiatives to increase predictability in 2-3 countries, possibly including Tanzania and Malawi. Priority 2: Improving the transparency of our aid and ensuring all government-togovernment aid is shown on partner country budgets. 5.14. Greater transparency of aid programmes and the donor practices around them is an important objective in its own right, given the need for DFID to be accountable for our actions to our own domestic audiences, including Parliament, non-governmental organisations and the general public. We also believe that making further progress in this area will have important spill-over effects, such as improving donor behaviour through greater peer pressure/scrutiny and improving the quality of the documentation that DFID produces. As such, we believe that greater transparency should underpin all of the other work that we are undertaking to make DFID s aid more effective and accountable. 5.15. The AAA includes three commitments on improving the transparency of aid: donors will publicly disclose regular, detailed and timely information on volume, allocation and, when available, results of development expenditure to enable more accurate budget, accounting and audit by developing countries. (Para 24a.) beginning now, donors will provide full and timely information on annual commitments and actual disbursements so that developing countries are in a position to accurately record all aid flows in their budget estimates and their accounting systems. (Para 26b) Donors will regularly make public all conditions linked to disbursement (Para 25b.) 18

5.16. These commitments are reinforced by those in the Accra Statement for the UKled International Aid Transparency Initiative (IATI), which emphasised the importance of transparency and committed donors to a number of steps to improve transparency, including sharing more detailed and more up-to-date information about aid in a form that makes information more accessible to all relevant stakeholders. 5.17. DFID is taking steps to improve our own transparency to meet these commitments. The UK s 2005 conditionality policy stated that DFID would publish all conditions on our website. DFID records conditions in project documents in order to comply with the 2005 commitment. We are working on publishing project documents on our website (see below). 5.18. Country Planning Guidance states that countries should produce a communications document, which includes information about expected and actual results and headline allocation figures. This is primarily for a UK audience and in English. The guidance states that parallel or similar versions may be created by country offices for local audiences. Publication of other elements of the country plan is at the discretion of country offices. 5.19. We need to develop reporting systems that enable us to report holistically about all of DFID s assistance to a particular country. There is currently a lack of clarity about how financing disbursed by some centrally and regionally managed programmes is captured in the recipient country s own budgeting and financial management systems. 5.20. DFID s Annual Report and Statistics on International Development (SID) provide more regular and up to date information on volume and allocation of aid. However, this information is not necessarily in a form that is accessible to all relevant stakeholders. 5.21. DFID provides guidance to country offices and other spending departments about when information on expected and actual disbursements should be provided to governments for non-prbs aid, which remains approximately three-quarters of bilateral programme spend. For PRBS, the policy states that we will provide information to governments about expected PRBS in a form required by governments, and that we need to be transparent about when and why aid is to be withheld because of breaches of partnership commitments. This should be sufficient to meet the Accra commitment for aid delivered as PRBS (which was 21% of total bilateral aid in 2007/08). 5.22. As part of our commitment to improving transparency, the Publishing Project Information (PPI) initiative will publish project information to the DFID external website from 2009. This will complement the existing project information published via the DAC website. 5.23. We will make information about the projects we fund available to the public both here in the UK, and in the countries we work, through the launch of a new searchable database on the DFID website. This will contain project information such as descriptions, purpose, dates, location, sector, aid effectiveness, financial information and an indicator to show whether or not specific conditions are attached. These items will subsequently be made available as a feed to Accessible Information on Development Activities (AiDA) and other external stakeholders for use in their analysis and information systems. 5.24. We need to improve our internal monitoring of Paris and Accra commitments in order to effectively and transparently report on progress and to show we are prepared to be held accountable for meeting the full range of aid effectiveness commitments. Strengthening monitoring of progress towards the more qualitative Paris commitments such as mutual accountability and ownership was a key recommendation of the DFID 19