Mozambican experiences with Paris Declaration
The Paris Declaration on Aid Effectiveness was endorsed in March 2005 by more than 100 countries and international organizations. The aim of the Paris Declaration (PD) is to improve the quality of aid and its impact on development. The PD was organized around 5 key principles of aid effectiveness; Ownership, Alignment, Harmonization, Management for Development Results and Mutual Accountability. Specific commitments and targets for 2010 were established for each principle.
Monitoring and Evaluation (M&E) of the PD was built into the declaration from the start, with a commitment to "periodically assess, qualitatively as well as quantitatively, our mutual progress at country level in implementing agreed commitments on aid effectiveness". There have been two monitoring surveys focusing on the indicators, and the first phase of the evaluation in 2007-2008 which aimed to assess how and why implementation had proceeded since adoption. The synthesis report of this first phase of evaluation was presented at the High Level Forum (HLF) in Accra in 2008 and aimed to draw out key emerging issues, initial lessons nd areas for concern as inputs for the high level discussions.
This report constitutes the Mozambican country report for the Second Phase of Evaluation which aims to document, analyze and assess the relevance and effectiveness of the Paris Declaration and its contribution to aid effectiveness and ultimately to development results, including poverty reduction.
The methodology followed was based on a standard Evaluation Matrix developed by the International Reference Group. The matrix was organized into three core questions, related to A) the PD in context; B) Aid Effectiveness Results and C) Development Results. A desk review, and data analysis was carried out, as well as semi-structured interviews carried out by the KPMG team, covering senior government officials, at central and sector level, donor and N representatives, civil society, academics and parliament.
There are certain limitations to the report and what can be achieved in terms of linking development results, and often even intermediate aid effectiveness results, to the PD. Firstly, many of the aid effectiveness initiatives in Mozambique pre-date the PD, and there is evidence that the Mozambican experience actually informed the drafting of the PD. Secondly, the PD was implemented in a highly dynamic context, with many other drivers of development results.
Thirdly, there is no counterfactual. Therefore the team followed the guidance of the International Reference Group and focused on "plausible linkages" and "possible contributions" of the PD rather than attempting any form of attribution.
The structure of the report closely follows the country matrix, to enable easy comparison across countries. Headings correspond to those in the matrix, and each section heading also includes the code from the matrix for easy reference.
Mozambique has long been regarded as a success story, and a donor darling. Impressive rates of economic growth since the end of the civil war that followed independence, the implementation of numerous reforms, a stable, democratically elected government, good progress on a number of social indicators and a seemingly large reduction in poverty rates have resulted in extremely large aid flows to the country. However, there are increasing concerns that the picture may be more complex, driven in part by the results of the household survey (2008/9) released in the last few weeks of this evaluation, which suggest that poverty on average may not have moved in the last few years, and has even got worse in some provinces. Results for social indicators were however in general fairly positive. There are also increasing donor concerns over governance, which led to a temporary suspension of General Budget Support (GBS) in the first quarter of this year.
In terms of the context in which the PD was implemented in Mozambique the main key finding is that there was already by 2004 (i.e. prior to Paris) a number of initiatives being carried out that already embodied what would become the principles enshrined in Paris. These seem to have been driven largely by internal factors, in particular a group of "like minded" donors and government officials, taking budget support forward, and a 2001/2 banking crisis which led to a suspension of budget support, which led to both government and donors recognising a need for a more transparent and structured approach to budget support. This resulted in a 2004 memorandum of understanding between Government and Budget support donors, which reads very much like a "PD-type" document. Other contributing factors were concerns with off-budgets, predictability, and the transaction costs associated with negotiating with individual donors.
By the 2005 baseline, there were already 18 donors providing General Budget Support (GBS) and there were already well-established Sector Common Funds in health, education, agriculture, water and HIV, some of which had been operating since the late 1990s. The proportion of programme aid (GBS plus sector programme aid) in 2005 was already 46% according to the Baseline survey. There was a nationally owned development strategy in place, and there was already a system of working groups to harmonize donor actions and promote government-donor dialogue. A joint government-donor Budget Analysis Group was working on the issue of offbudgets, and there was increasing investment in and use of government systems. There was already in place a mutual accountability framework, with a donor performance assessment framework (PAF) with indicators in Predictability, Alignment & Harmonization, Administrative urden, Transparency and Administrative Burden.
One aspect of the aid architecture which can already be seen from the early 2000s is the creation of an extremely strong and influential donor grouping around budget support, which became and remains the most influential donor group. These Programme Aid Partners (PAPs) have driven the aid effectiveness agenda in the country from the beginning. However, recently the US and UN were admitted as associate members of the PAPs, which further strengthens this group but continues to leave out vertical funds and non traditional donors, both of which are increasingly active in the country, as well as Japan who were invited to join and declined, partly on the basis of the high transaction costs associated with participation.
In terms of Ownership, by 2005 Mozambique was coming to the end of the first Absolute Poverty Reduction Plan (PARPA I) and starting to draft the second (PARPA II). Ownership was rated as "moderate" by the baseline survey in 2005 and the monitoring survey in 2007, and the assessment would remain accurate today. There is a functioning MTEF which is linked to annual budgets. There are examples of where the government takes ownership on particular issues, however there is a feeling among government officials that there are limits to the extent to which donors are willing (or able) to allow ownership, particularly where there may be conflicting approaches. Given this skepticism by government, which does seem to be somewhat justified by recent examples, it is perhaps reasonable to suggest that ownership remains partial.
All donors report that there is strong alignment between their country programmes and the government's overarching poverty reduction plans. However, PARPA II was fairly broad in scope, and donors participated amply in the formulation of the document, such that alignment of country programmes is fairly easy to achieve. There has been a steady improvement in government systems and increasing use of these by donors, mainly driven by programme aid but more recently donors have started to channel funding for projects (including the funding of this report) through national systems. As projects remain a large proportion of aid, this is an encouraging development. Predictability is good for GBS and common fund contributions, with firm commitments made in time for annual budgeting for the following year, but predictability remains a challenge for projects.
There has been from early on a strong degree of harmonization among GBS donors, in terms of policy dialogue and common reviews, and a large and cumbersome structure of working groups has been set up. There is evidence of fatigue associated with the large transaction costs involved in maintaining such a structure, although government does report positive effects of being able to deal with a large number of donors as a group. Division of Labour initiatives have been underway for a number of years, but have largely not taken off due mainly to lack of buy in from government, who see benefits in spreading risk by having a large number of donors in sectors, and who are nervous that exits from sectors would not be done in a coordinated way, thereby reducing funding. In terms of the PD indicators on missions and joint analytical work, progress has been slow. The PAP group reach the target of over 66% of aid flows provided in the context of Programme Based Approaches, but this is not the case for the country as a whole.
There is a well established results oriented framework attached to the PARPA II and the forthcoming PARP, which is used as the basis for a Performance Assessment Framework with PAPs. Programme budgeting has been introduced, although there remain sever capacity constraints especially at sector level in formulating indicators and also in terms of data quality.
With regard to mutual accountability, as part of the annual reviews between government and PAPs, the performance of PAPs with regard to PD principles is assessed. There is a Performance Assessment Framework with targets which are either the same or more ambitious than Paris. However, the results of the assessment, which scores each donor against each target, is more used by donors (in dialogue with HQ, or in terms of peer pressure by more "progressive" donors on others) than by government. It should be noted that this process only covers PAPs and associates, thereby leaving out Japan, vertical funds and non-traditional donors.
In terms of the impact of the PD on Aid Effectiveness, the principle role the PD seems to have played in Mozambique, given that all elements of the Aid Effectiveness agenda were in place prior to 2005, seems to have been to maintain momentum, and keep the issue on the agenda for both donors and government. This is an important role, however, as the implementation of PD and in particular the structures and processes related to the various modalities of aid have been dynamic, being refined and adapted over time. There are examples where both government and donors have used the commitments made under PD as arguments for particular courses of action, such that the PD can be seen as playing a legitimizing and supportive role. However, the interpretation of PD in Mozambique has been highly focused on GBS, with GBS being seen as the "purest" way of achieving PD principles, even though GBS itself is not mentioned in the text of the Declaration. This has led to a narrowing of the debate, with a focus on GBS and ABCD sector common funds, and only now are there signs of greater interest in making projects more effective (inclusion of project-based donors in the PAP structures and some projects starting to be put on treasury for example). As projects remain the largest aid modality in the country, this is crucial. There is no reason why a project cannot be made to follow all PD principles - Government can have ownership of a project, a project can be part of a SWAp, they can be aligned with government plans, and funds for projects can be channeled via government systems. Donors could in theory harmonize around projects, and there is no reason why projects cannot be managed for results and why donors cannot be held accountable for their performance within projects.
Development Results in the country have been mixed, despite high levels of economic growth. The recent 2008/9 household survey suggests a worrying lack of poverty reduction on aggregate, albeit with significant regional variations. This is extremely worrying both for government and also for donors who have poured aid in to support poverty reduction in recent years. In general there has been more progress in social sectors than in economic sectors, and one of the key areas identified as contributing to lack of progress on poverty is the failure of agriculture. Mozambique is considered to be likely to meet four out of the 21 country level targets of the Millennium Development Goals (MDGs), to potentially meet a further ten, and to e unlikely to meet one. Six do not have sufficient data to assess.
The sector studies highlighted some interesting issues which may be related to this differential in performance. In health, there has been a good performance on achieving indicators, and there is a general consensus on the approach to the sector, shared by government and donors. There is reasonably strong government leadership, and a well functioning SWAp and common fund, with mutual accountability mechanisms and use of government systems. This is not to say that there have not been challenges in implementing aid effectiveness initiatives in the sector, and transaction costs of doing so have been extremely high, but in general good progress seems to have been made in implementing Paris. In contrast, the situation in Agriculture is one where the donors and government have disagreed on the way forward, and where donors have been withdrawing from, or reducing support to, the common fund, which was repeatedly described by donors as "a mess". Of course this is not to suggest that where the PD is implemented at sector level, development results will flow. There are many other factors, some of which may impact on both development results and the implementation of the PD. For example, where government has a clear and strategic vision and is able to take leadership, donors may be more likely to support the sector with Programme Based Approaches. In fact, given the withdrawal from the common fund of some donors in agriculture, it could be suggested that donors may only be willing to provide programme aid when they agree with the approach taken by government.
- Extracted from the executive summary to: Final Report - Paris Declaration Evaluation Phase 2, Mozambique. The 98 page report can be accessed here.