Introduction

Coordination can be regarded as one of the traditional topics in the debates on aid among academics and practitioners (see, for instance, Hoebink, 2004; Carbone, 2010b; Woods, 2011; Dearden, 2013). The need to coordinate aid was already a central rationale for the creation of the Development Assistance Committee (DAC) of the Organisation for Economic Co-operation and Development (OECD) in 1961, and the work of the DAC consistently focused on this issue (Herman, 2013, p. 10). More recently, the discussion on coordination took place in the context of the debate on ‘aid effectiveness’ (Rome High Level Forum [HLF] in 2003, the Paris Declaration (PD) in 2005 and the subsequent forums in Accra and Busan).

The need for coordination exists because aid is fragmented (Knack and Rahman, 2008; Knack and Smets, 2012). Official development assistance (ODA) is fragmented mainly for two reasons. First, aid comes from a number of different bilateral and multilateral sources. For example, the number of multilateral ODA providers reporting to the OECD’s reporting systems is around 230 (OECD DAC, 2013). In addition to those bilateral donors who are members of the DAC, several types of private (non-governmental organisations [NGOs], private foundations and so on) and public non-DAC donors (Arab countries, ‘new donors’ like China, India and Brazil) also provide aid. Second, aid typically provides resources through several modes of delivery (investments, loans or grants, advice, training and so on). These resources are spread among a number of instruments and interventions per donor (Acharya et al, 2006).

The increasing number of aid providers and aid activities has resulted in ‘aid proliferation’ or ‘fragmentation of aid’ (Klingebiel et al, 2016). Fragmentation of aid has a number of consequences, especially for aid recipients, as the direct and indirect transaction costs can be significant (OECD, 2011, pp. 4–5). Transaction costs are likely to increase because donors are engaged in a number of countries and sectors, and each donor intervention requires attention in terms of consultations, missions, reporting needs and so on.

The issue of coordination has gained new impetus over the years. Better coordination is considered crucial for overcoming negative consequences of aid fragmentation. The Busan Outcome Document encouraged developing countries to increase coordination efforts to ‘manage this diversity’ of donors, instruments, programmes and projects. Although all development cooperation actors have committed to increasing their coordination efforts in the context of the ‘aid effectiveness agenda’, little progress has been made (Woods et al, 2011, pp. 19–22).

While the term ‘coordination’ is used in a broad sense, there is a strong need to be specific about different areas where coordination might be required and what the potential consequences of ‘non-coordination’ are. The article contributes to conceptualising ‘aid coordination’ and provides a conceptual framework for coordination. In addition, despite the fact that aid actors agree on the need to have more and improved coordination, there is broad acceptance that lack of coordination remains a crucial challenge. With this in mind, we address the question of why aid actors might have a limited interest in aid coordination. We assume that ‘more effective aid’ through ‘aid coordination’ is just one objective of European aid actors, and that this objective might compete with other implicit or explicit donor interests.

The article mainly focuses on aid providers from the European Union (EU) as an important subset of the entire group of donors that is organised in the DAC. EU aid providers are European institutions (especially the European Commission (EC) and the European External Action Service (EEAS)) and the 28 EU member states (MS). The rationale for focusing on the EU is that the Union introduced several approaches aiming at aid coordination (examples are discussed below). In addition, we assume that the group of European aid providers is more homogenous and thus it is more likely to make progress on coordination than the DAC group as a whole. Nevertheless, the academic debate shows that the EU is far from being a single actor (Carbone, 2010b), and that MS adopt distinct aid approaches (with quite different consequences for aid coordination) (O’Riordan et al, 2011).

Within the EU, the 2007 Three-Cs evaluation, which looked at coordination, complementarity and coherence, concluded on the first area that the MS and the Commission had a ‘remarkably low level of performance when it comes to achieving value added effects and impacts of coordination efforts with each other in general, and specifically in the area of support for local development’ (ECDPM, 2006). The same applies to specific efforts for increased coordination, which mostly show little evidence of impact (see, for example, Delputte and Söderbaum, 2012; Bigsten, 2013; Delputte and Orbie, 2014). Thus, the main question posed in the present article is: How do we explain the discrepancy between the aid effectiveness rhetoric on the one hand and insufficient levels of coordination performance on the other?

This article uses two country case studies for a better understanding of assumed functioning aid coordination approaches. We focus on two partner countries where we assume a positive country context for aid coordination. The case of Myanmar was chosen against the background of a rare case of a country receiving aid only recently (since 2011). Our hypothesis in this case was that within the general aid effectiveness debate and with the introduction of EU-specific aid coordination approaches, the EU would apply high aid coordination standards in such a new country aid context. The basic assumption is that an aid-recipient newcomer might have strong advantages to apply ‘best practices’ from the international aid effectiveness debate. The case of Rwanda was chosen because of indications that aid actors are largely applying principles of the aid effectiveness agenda as a result of a proactive Rwandan aid policy. The country proactively applied the outcomes of the international debate on aid effectiveness: use of PD indicators; issuing an ‘aid strategy’ with clear guidance for donors; pushing donors even beyond the EU to implement a clear ‘Division of Labour’ (DoL) aid approach and so on. Thus, Rwanda is to a large extent a role model in partner country-driven aid effectiveness strategies. This is why we assume relevant insights in the political economy from aid coordination-enabling country cases. Thus, the article looks at the main driving forces for improved and better aid coordination in these specific cases.

In addition to the review of literature and documents, the authors conducted personal and telephone interviews with officials of the EU, particularly the Commission’s Directorate-General for International Cooperation and Development (DG DEVCO)/EuropeAid and the EEAS, several EU MS representatives and a number of experts. The case study on Myanmar is not only based on interviews in Europe and the analysis of relevant documents, but also on a research stay by one author in Myanmar (May 2013) in order to conduct a number of interviews with representatives of the EU and non-EU donor community, as well as of the Government of Myanmar (GoM). The case study on Rwanda is based on several research stays in the country (mid-2012, September 2014 and February to April 2015) by one of the authors.

Chapter 2 briefly discusses the concept, levels and benefits of EU coordination. Chapter 3 addresses the political economy of both donors and partner countries, and formulates hypotheses for improved coordination that are tested through the case studies in Chapter 4. The final chapter concludes.

EU Coordination: Concept, Levels and Benefits

In view of an international debate based on a multiplicity of understandings of aid coordination, we use the following definition of aid coordinationFootnote 1: Aid coordination occurs when two or more development partners – preferably under the leadership of the partner country – undertake activities intended to improve or harmonise their policies, programmes, procedures and practices, so as to maximise the development effectiveness and the efficient use of aid resources.Footnote 2

Coordination of EU aid actors includes both coordination among the EU MS and coordination between MS and EU institutions. Henceforth, these two types of coordination are both called ‘EU coordination’. Although our main focus is on ‘internal coordination’ among EU donors, there are also links to non-European donors.

We distinguish between several degrees of coordination. The lowest degree of coordination is just an exchange of information. The next degree focuses on interaction aiming at harmonising strategies and approaches (for example, sectoral or cross-country DoL). Finally, a fully integrated approach (for instance, the European Development Fund) can be regarded as the most ambitious degree of coordination.

In general terms, coordination aims at gains in efficiency and effectiveness. Efficiency gains are related to the extent to which time, effort or cost is used well for the intended task or purpose. Effectiveness gains focus on the degree to which objectives and results are achieved and the extent to which targeted challenges are addressed. Effectiveness, as distinct from efficiency, is determined by the impact without reference to costs.

Coordination is expected to (i) reduce financial costs for reaching a given target, (ii) improve impact or outcome with the same investment or (iii) both. The costs of coordination or the benefits of more and improved coordination might be relevant for either the aid recipient or the donor (or both).

We distinguish between three broad levels of coordination: (i) the policy level of development aid coordination, which is about principles and standards, strategic approaches and allocation patterns; (ii) the programming level, which is about approaches and strategies (such as a country-specific strategy) during the aid programming phase and (iii) the implementation level, which focuses on how aid is coordinated during the aid provision phase. In this context it might be relevant to look at joint or separate implementation arrangements.

The most evident, straightforward potential gains of coordination in development cooperation are savings in transaction costs, as they are not linked to any specific policy choice that other potential gains may require. Following Lawson (2009), transaction costs are defined as ‘the costs which allow an economic transaction to take place but which add nothing to the value of the transaction’.

But costs and benefits also include a broad range of positive implications of greater coordination, such as gains in efficiency and effectiveness in terms of attaining developmental objectives, governance, ownership, reductions in transaction costs, institutional capacity and so on (Klingebiel et al, 2013). The quantitative assessment presented in Bigsten et al (2011) and Klingebiel et al’s (2013) study makes a strong case that savings from increased EU coordination in the area of development cooperation could be substantial (in the order of several billion euros). However, it is the consideration of these numerical gains and the broader range of benefits beyond strictly quantitative gains that would appear to justify closer coordination.

At the policy level, EU coordination takes place externally as well as internally. Internationally, the EU has had a leading and active involvement in the four HLFs on Aid Effectiveness (Carbone, 2013) and the current Global Partnership for Effective Development Co-operation. The EU’s internationally leading role, however, has arguably been weakened by its slow internal progress in reducing fragmentation and aid proliferation, despite its commitments (Fejerskov and Keijzer, 2013). The Union’s landmark efforts on coordination, the Code of Conduct on Complementarity and DoL (2007) are the answer to two central problems in aid allocation: the proliferation of donors (too many donors per country or sector) and the fragmentation of aid (small amounts of aid from too many donors) (Mürle, 2007; Carlsson et al, 2009; Carbone, 2010a). The Code emphasises the importance of ‘cross-sector’, ‘cross-country’ and ‘within-country’ vertical approaches, as well as cross-modalities and the complementarity of instruments. This has been strengthened by the Agenda for Change approved by the Council in 2012 (EU, 2012), which provides additional directions concerning the geographical and sectoral orientation of EU development assistance.

At the programming level, sectoral DoL does seem to have a significant impact on aid fragmentation in most partner countries (Wenzel et al, 2010). A widespread perception among stakeholders is, however, that transaction costs have increased for partner countries (more meetings, implementation delays and so on) without being offset by significant increases in effectiveness. The EU intends to improve in-country donor coordination through the wider use of Joint Programming (JP). JP includes a single EU country strategy aligned with the partner country’s own national development plan. JP exercises were piloted in several countries in late 2013 and as many as 50 may be under way by 2020 (Furness and Vollmer, 2013).

At the implementation level, two main coordination mechanisms can be identified, namely, multi-donor budget support (MDBS) and blending facilities. The former is the most prominent form of programme-based approaches (PBAs), which can be implemented through different aid modalities, ranging from pooled (or basket) funding of specific activities or reform programmes to joint support of sector-wide approaches, as well as budget support (sector and general) (Koeberle et al, 2006). On the whole, MDBS provides a very strong framework for donor coordination with good potential for reducing overlaps and general transaction costs. Regarding blending, the EU Platform for Blending in External Cooperation was established in 2012 with the aim of reviewing the existing mechanisms and the development of a common results-based framework to measure impact. This could become an important field of cooperation with potentially high benefit depending on the level of funds it will help leverage.

It is worth noting, though, that from the perspective of a recipient, coordination of donors may result in the formation of a unified and strong position among the larger donor group. This pressure might be used to push a partner country government in a specific direction (reform decisions and so on). In this case coordination might reduce partner countries’ room to manoeuvre and thereby their ownership. From the partner country’s perspective, less donor coordination could also be regarded as a form of ‘risk sharing’. That is, partner countries may remain less subjected to particular donors’ leverage and therefore more independent if they have other donors they can turn to in the event of disagreements.

The emphasis on the crucial role of partner countries’ leadership in coordination processes has become more pronounced in the debate on aid and development effectiveness over the past 10 years. Coordination without guidance coming from the partner country can clearly be regarded as a second-best option, necessary – for example – in a situation where there is non-existent or weak commitment from the partner country government (for example, in failed states).

The Political Economy of EU Aid

There is extensive academic evidence for the financial and non-financial costs of weak aid coordination and the potential benefits of more and improved coordination (for example, Bigsten et al, 2011; Bigsten, 2013; Dearden, 2013; Klingebiel et al, 2013). At the same time, aid actors – in terms of development partners and partner countries – agree that more and improved coordination is desirable. The outcome documents of all four HLFs confirm the strong commitment of donors and recipients to an ‘aid coordination agenda’. Therefore, there is a broad consensus on the need for reduced transaction costs and greater impact through better coordinated aid. However, donors do not consistently follow a policy in line with clear coordination principles. Equally, on the side of partner countries, recipients also do not always push for more coordination. This raises the question: What are the main reasons why donors and recipients are not more committed to coordination?

Overall, we identify several important explanations related to the theoretical debate on coordination and the complex political economy of both donors and partner countries.

Regarding the former, even if the academic literature and aid agencies agree on the value added of more and improved coordination, it remains unclear what the right level of ambition is or should be. If a fully integrated approach (single window) can be considered as the highest level of coordination, a number of trade-off aspects might apply (monopoly position of a single aid agency and so on). For example, even if in principle just one donor representation in a partner country would allow high savings in transaction costs, MS would probably not consider this to be the best option because of a number of other aspects (for example, loss of visibility and aid as an incentive for ‘non-aid objectives’). This is why the best level of coordination depends on a set of assumptions and the specific context. Arguably, so far neither research nor discussions among aid practitioners have yielded good indicators for an ‘ideal case’, particularly at programming and implementation levels.

In addition, the political economy of donor coordination is complex. Coordination benefits are only partly realised despite the fact that the benefits of enhanced and increased coordination are recognised by donors. Therefore, any analysis of donor behaviour concerning coordination needs to reflect on competing and contradicting interests. First, donors behave, to a large extent, like private business actors; the ‘business of aid’ is dominated not only by the aid providers’ altruism. For example, when it comes to the best projects, political access to the host government, and public reputation in the partner and donor countries, other donors can appear to be competitors. Donors normally tend to prefer projects in social sectors. This might be an advantage to convince stakeholders at home in donor countries as aid strategies state such a preference for sectors with a high direct impact on poverty reduction.Footnote 3

Second, when it comes to the implementation level, donors may have a strong interest in implementing their own aid portfolio because of the ‘aid industry’ in their country of origin. National state and non-state implementing agencies may also discourage the use of, for example, budget support and other PBAs. In the case of Germany’s aid, for instance, the government needs to get approval from parliament in advance for budget support and other main PBA interventions. Implementing agencies have a strong incentive to highlight ‘coordination limits’ and to try to keep a significant share earmarked for them to implement. This in turn perpetuates in-country fragmentation because implementing agencies tend to be unwilling to disengage from sectors or areas where other actors may have a comparative advantage – even in cases when this can be clearly determined.

Third, each donor has specific requirements. These requirements may come from the national parliament, the auditor general and so on. All of this will result in specific regulations and expectations, which can limit the scope for coordination. Meeting these requirements may be legally binding and/or necessary to withstand public scrutiny or accommodate auditory needs. In the context of increased pressure to prove the value for money of every euro invested in development, it is particularly important for donors to be able to track their allocated funds and link them to direct achievements. Aid approaches like results-based aid may have an implicit bias for less coordination with other donors as the attribution of outputs and outcomes to a specific donor is quite often regarded as a key advantage. Therefore, a high degree of coordination at the implementation level may be in conflict with the attribution of results (Klingebiel and Janus, 2014).

Fourth, donors need ‘visibility’ (Vollmer, 2012; Delputte and Orbie, 2014); for example, in a high-profile political transformation context donors might consider visibility to be a high priority. Coordination at the implementation level in terms of pooled aid modalities and especially harmonisation may result in a reduction of donor-specific visibility. Joint donor approaches and the use of country systems as an indication for a high level of harmonisation with partner country approaches tend to imply a certain trade-off vis-à-vis the visibility of a donor’s profile.

Fifth, donors link aid to a number of foreign affairs and commercial interests. If, for example, aid also has to support the private sector of the donor country by tying aid formally or informally, aid agencies will be less keen on coordination. In this regard, coordination at the programming level may involve a sectoral DoL approach that would include the risk of being pushed out of a sector where the donor country has a self-interest.

Sixth, the perspectives of individual EU MS on coordination at all levels may be quite different. For example, MS with a strong aid portfolio in a given partner country might want to maintain strong national visibility and leverage, whereas MS with small aid portfolios might gain in case of a highly coordinated approach (free riders). In addition, there may be important differences in the issue of ‘EU identity’ where Eurosceptic countries may be more reluctant to ‘sail’ under the EU flag. Or, as Delputte and Orbie (2014) argue, the political shift to the right within EU MS has made them less conducive to agreement and more prone to pursue individual interests. In this regard, although the introduction of differentiation as regards country allocations has involved a certain degree of cross-country coordination, this has only affected EU institutions’ development cooperation, not MS’. In general, the lack of progress on cross-country coordination within the EU can be attributed to the MS’ view that country allocations are a sovereign decision as well as an instrument of national foreign policy. Finally, each European aid actor has to consider a number of different internal actors and interests. Member states’ interests are affected by several government players (ministries in charge of trade, foreign affairs, the environment and so on), parliaments, NGOs, private sector groups and individual companies. The same applies to the different services of the EC, the EEAS, the Council, the European Parliament and their respective internal interests. For example, initiatives of the European Parliament show a somewhat higher level of ambition for coordination than the government representatives in the Council.Footnote 4

Partner countries’ political economies are also complex when it comes to aid coordination at the programming and implementation level. This is because of several factors. First, coordination increases the leverage of donors. This might, for instance, result in decreasing ownership, poorly aligned strategies, micromanagement of donors or strong pressure for political reforms. This is typically not in the interest of recipients, which try to limit the interference of donor approaches with national policies for several reasons, such as avoidance of activities that might or could be perceived as undermining the sovereignty of the country; diluting national strategies; and transparency on the use or even misuse of resources/public financial management systems and so on.

Second, coordination results in decreased ‘flexibility’ for some recipient country stakeholders. For every line ministry, for example, it might make sense to be able to approach each and every donor, while finance ministries in partner countries are much keener on a coordinated and centralised approach that allows them to retain control over resources. Thus, donor coordination can support the coordination and planning power of partner countries if donors follow an approach under the leadership of the recipient.

Third, coordination may result in an ‘all-or-nothing’ dichotomy. Partner countries may perceive an increased risk of all donors reacting in unison and, for instance, jointly pulling out at once if an important issue such as disputed election outcomes or human rights violations, inter alia, arises. A typical case in this regard would be a coordinated response by donors on main political events in the partner country such as elections, unconstitutional behaviour of the partner countries’ government or human rights violations, inter alia.

In addition, greater donor coordination at the policy level on cross-country allocation may imply important shifts in funding, with a large number of recipient countries ‘losing out’ with the change. Indeed, a substantial number of partner country governments fear a decrease in aid funding and are hence not supportive of cross-country DoL.

Incentives by both donors and partner countries can be viewed as perceived advantages or disadvantages of coordination. A summarised representation of this is provided in Table 1, which includes a non-aid perspective, that is, perceived interests in areas other than aid.

Table 1 Potential advantages and disadvantages of EU coordination by aid actors and degree of coordination
Full size table

All of the above posits the main reasons behind the political economy of both donors and recipients pursuing different degrees of aid coordination at different levels. Coordination at the policy level is probably easier to achieve under some of these rationales. The EU has arguably been quite successful here except for in cross-country allocation. This corresponds to the generalised recognition that more coordination is desirable. When it comes to taking steps to modify country allocations or coordinate with fellow MS at programming level, determination seems to falter and decisive commitments seem more or less restricted to JP initiatives so far. At the implementation level, Klingebiel et al (2013) show that a strict application of the code of conduct should lead to a substantially enhanced sectoral DoL. This, however, does not seem to be significantly translating into reality on the ground, as Nunnenkamp et al (2015) report.

Some Hypotheses on the Degree of Coordination at Recipient Level

In view of the above, and recognising that the degree of coordination depends on a multiplicity of interrelated factors as discussed in the previous subsections, we posit that stronger coordination will take place the more the following circumstances are met:

  • There is strong demand, engagement or even leadership by recipient countries.

  • Recipients present adequate capacity to lead and handle donors and projects.

  • There exist perceived comparative advantages by donors – including history of engagement with and political access to recipient country and they are recognised by both donors and recipient. Although donors tend to overly justify their own comparative advantage, there may be situations where comparative advantages are clear because either the recipient establishes it that way or donors widely recognise it.

  • Comparative advantage, along with funding commitment, is the basis to determine coordination leading roles.

  • Donors have little direct interest in a strong bilateral/visible presence in the partner country.

In order to test these hypotheses, the following section presents the main findings of two case studies selected because of their strong recipient government involvement in coordination, comparatively high capacity and, in the case of one of them, relatively rapid and recent opening to development cooperation. As these are some of the key factors for genuine donor coordination, findings from Myanmar and Rwanda should be able to illustrate whether the above hypothesis can explain to some extent the larger political economy described herein.

Case Studies

A significant part of aid coordination is related to within-country aspects. While each country’s case is unique, we have chosen Myanmar and Rwanda under the assumption that we might draw conclusions from two cases where we expected the conditions for aid coordination to be favourable (see the section ‘Introduction’). Both present strong recipient government engagement, and in the case of Myanmar its opening up for development cooperation is fairly recent. With this in mind, we would expect close compliance of donors with the aid effectiveness agenda. In addition, Rwanda provides an example of ‘more’ and ‘improved’ aid coordination given its government’s practical commitment to the implementation of the aid effectiveness agenda.

Myanmar

Myanmar has undergone dramatic changes and reforms over the past couple of years. In March 2011, a fundamental transition phase started, led by former President Thein Sein. The political opening of the country has induced radical changes in all areas including development aid and, in April 2013, the EU lifted sanctions status on Myanmar (excluding arms).

Donor agencies have started to (re)engage with Myanmar. The GoM and other local institutions and actors are faced with a massive influx of donors. Many institutions are starting from scratch with a limited capacity for collaboration with aid providers. However, in general terms the GoM shows a high level of ownership related to development and a strong intention to apply lessons learnt on aid effectiveness from other countries (including Rwanda).

Thus, aid coordination has been one of the key challenges in Myanmar since the reform process started. EU donors’ engagement is increasing, but as of mid-2013 only the EC, the United Kingdom, Germany, Denmark, Finland and Sweden were actually present on the ground. Donors have a strong interest in setting up their own structures and projects, but well-coordinated joint approaches are only used to some extent (for example, the three operational multi-donor trust fundsFootnote 5 that were in place at the beginning of 2013). Since 2009, the Partnership Group on Aid Effectiveness, which included 45 donor agencies in 2013, has operated as the main donor coordination mechanism. The Partnership Group has also established a number of working groups and should be operating on a more formal footing in the near future. An important step by the GoM was to organise the First Myanmar Development Cooperation Forum in January 2013, which concluded with the Nay Pyi Taw Accord for Effective Development Cooperation. The outcome document explicitly refers to the Busan Outcome Document and provides specific commitments on the part of both the GoM and the donors as to how to make aid more effective in the context of Myanmar. The accord provides for specific guidelines to ‘align assistance with national priorities’ and ‘participate in and be guided by country-led coordination processes’. The 2013 event was followed in 2014 and 2015 by a second and third forum. The Forum has proved to be a crucial mechanism to follow up national reform efforts and donor coordination.

The EC and several EU MS are active in Myanmar, the United Kingdom being one of the most significant bilateral donors. The EU delegation in charge of the country was based in Bangkok (Thailand) until mid-2013, and only in April 2014 was a branch office inaugurated in Yangon. EU staff in charge of development cooperation shuttled between the Bangkok delegation headquarters and the offices in Yangon and Nay Pyi Taw. Initially, the relatively low profile and understaffing of the EU representation did not help it to lead EU donor coordination.

A European Council Conclusions of July 2013 set out a Comprehensive Framework for the EU and MS’ policy to support the country for 3 years. The framework specifies goals and priorities towards building a lasting partnership and promoting closer engagement. The goals aim at supporting political, social and economic development, while fostering respect for human rights and assisting the government in increasing its role within the international community.

A ‘Joint EU Development Partners’ Transitional Strategy for Myanmar’ was prepared and agreed at the local level by 13 Member States delegationsFootnote 6 and EU institutions (EEAS and EC) (eeas.europa.eu/delegations/myanmar/documents/eu_myanmar/joint-eu-development-partners-transitional-strategy-for-myanmar-2014-2016_en.pdf, accessed 7 August 2015). The strategy refers to main general coordination instruments of the EU and to Myanmar’s aid architecture. However, aside from a general compilation of EU engagement in main sectors and cross-cutting areas, it has little ambition, since the document does not provide guidance for aid coordination beyond information sharing (something that is envisaged to change after 2015).

Overall, the first phase of transition in Myanmar was characterised by non-harmonised and non-coordinated approaches. In general terms, donors pushed for coordination only to a very limited extent. Coordination among the EU donors is modest at best. This is proving a major handicap, not least for the introduction of a JP approach. DFID/UK is by far the most active bilateral EU donor in Myanmar, but is so far not perceived as a driving force for European aid coordination there.

In addition, other MS also raise concerns (overambitious synchronisation and so on) about the feasibility of a JP approach for Myanmar. Against this background, the most important EU initiative for aid coordination has received little support from European bilateral donors.

At policy level, Myanmar has received a great amount of attention from EU aid actors since 2011. Not least because of fundamental changes in the country, the EU tried to coordinate closely over a basic strategy (for example, on the sanctions). Looking at the overall assessment of the situation in Myanmar and the end to EU sanctions, it is apparent that there has been a harmonised approach to fundamental questions. There are also indications from the policy level to encourage a more closely coordinated EU approach, especially for programming. However, the perceived need for ‘visibility’ of MS vis-à-vis the GoM is limiting coordination at the same time.

At the programming level, the JP approach is intended to be the main vehicle, and preparations for the conceptual work started in 2012. However, EU aid actors on the ground are only partly committed to the agenda. Coordination meetings and activities are taking place but MS are not really ‘buying into’ the idea of preparing for JP. Myanmar’s most important bilateral MS partner is hesitant about the approach, and other MS are sceptical about potential benefits and feasibility. There seems to be little incentive for the main EU donor, the United Kingdom, to reduce its profile in favour of a unified EU engagement, and there is widespread concern about potential trade-offs between getting things done quickly and the delays involved in coordination. In addition, the EC does not act as a strong ‘coordinator’ on the ground because of capacity constraints, and the link between the policy level and the programming level is rather weak. This is indeed confirmed by non-EU actors on the ground, who do not seem to perceive the EU as a coordinated unit in aid.

At the implementation level, there are Multi-Donor Trust Funds, which provide benefits by means of a harmonised implementation structure.Footnote 7 Although this is not specifically an EU coordination instrument, it is supported by several EU aid actors. With a medium-term perspective, PBAs and even budget support might gain in importance. PBAs, in particular, can play an important role for the new sector working groups.

To sum up, despite the opening up of Myanmar and the opportunity this provided to set up a ‘model’ coordinated landscape, developments since 2011 seem to indicate that coordination will require substantial further efforts and greater political commitment by donors and will encounter serious challenges. The GoM has taken decisive steps towards greater coordination, but its failure to materialise can be blamed on both donors and recipients failing to identify clear competitive advantages, plus donors’ pressing need to be visible given Myanmar’s strategic and economic potential.

Rwanda

For a number of reasons, Rwanda is a unique case. It is still confronted with the consequences of the 1994 genocide, which are apparent to this day. In addition, the region is experiencing a number of conflict situations that are directly relevant for Rwanda, which is also seen as a conflict party. Furthermore, discussions on Rwanda are quite often controversial because of governance-related issues. Finally, and contrastingly, Rwanda is viewed as an example of a ‘developmental state’ (Booth and Golooba-Mutebi, 2012; Abbott et al, 2013) that has a committed leadership on the one hand but significant limitations to participation in the policy sphere on the other.

Rwanda can be considered as one of the most advanced examples for aid effectiveness (see Hayman, 2009). In 2006, Rwanda introduced an aid policy that applied the PD to its own specific context. This aid policy identifies budget support as the first priority in terms of aid instruments/modalities. It also served as the main reference for country-led donor coordination, as well as placing emphasis on alignment with government priorities, the use of country systems and ownership.

In addition to these efforts, the Government of Rwanda (GoR) is pushing to implement other commitments of the aid effectiveness agenda. For example, the government and development partners have agreed to undertake an ‘inclusive joint planning exercise’. This effort aims at better synchronisation of the development partners’ programming/planning cycles with the national processes, and further improving coherence and allocation of development cooperation to support national priorities (MINECOFIN, 2013, p. 4).

Overall, for a number of years coordination of EU aid actors at policy level mostly worked in the context of issues related to budget support. In 2012, a report by a group of UN experts on the Great Lakes region stated that public institutions and the GoR still have an active military role (supporting proxy rebel groups and so on) in the conflict in the eastern Democratic Republic of Congo, which led to the suspension of budget support by main donors. In particular, the restriction and suspension of budget support was coordinated and to some extent the reactions of donors were harmonised in 2012; these decisions were followed up by reprogramming former budget support operations in subsequent years. Linking policy debates and budget support has been criticised by the GoR, however, a clear indication that recipients may realise some of the costs – political in this case – that greater donor coordination can imply for them.

At the programming level, Rwanda is to a large extent an example to others. In addition to former budget support forums (which no longer exist), an overall, less intense dialogue mechanism is in place that involves almost all bilateral and multilateral donor agencies. In 2008, ‘donor mapping’ was conducted; the analysis showed that some sectors were proportionally overfunded (such as health care), whereas others were underfunded (such as transport). In addition, the GoR highlighted high transaction costs against the background of a fragmented landscape of donor interventions. Furthermore, development partners were asked to give a self-assessment on their respective comparative advantages. On the basis of those inputs, in early 2010 the GoR presented to the group of development partners a proposal for ‘DoL’. The proposal took the EU code of conduct as a reference and proposed specific sectors to each donor. Development partners were requested to limit the number of sectors to three. An adjusted proposal was endorsed by 15 major donors, including four non-EU bilateral partners and three multilateral donors, as well as by 16 UN agencies. On the basis of Rwanda’s Economic Development and Poverty Reduction Strategy II (EDPRS II), the GoR revised the division of labour matrix in 2013 (see www.devpartners.gov.rw/fileadmin/templates/documents/DOL_Oct_2013.pdf).

At the implementation level, EU donors participated in budget support operations until 2012. In general terms, this was an important contribution to greater aid effectiveness and is in line with the Rwandan aid policy. The de facto reprogramming of former budget support operations has led to less donor coordination at the implementation level for all former budget support donors, but especially for the EU. Donors have fewer incentives for ambitious coordination efforts as ‘projects’ do not essentially require a high degree of coordination. At the same time, the GoR has responded with a relationship with donors that is characterised by less dialogue on key aspects of Rwanda’s development. Nevertheless, the GoR keeps using its policies and strategies to guide donor approaches.

The major positive difference compared with many other countries is the leading role taken by the GoR. The government organises the most important parts of aid coordination and pushes donors to ‘walk the aid effectiveness walk’. The Rwandan case has three main characteristics. First, the government has an aid policy in place that is implemented consistently by the entire government. Sector ministries cannot maintain special relationships with donors because all donor activities are strongly coordinated by the powerful Ministry of Finance and Economic Planning (MINECOFIN). Second, the GoR measures the progress of donor performance or lack thereof through the Rwandan Donor Performance Assessment Framework (DPAF), which is an effective tool designed and applied by the GoR (see www.devpartners.gov.rw/fileadmin/templates/documents/DPAF_FY_2012-2013.pdf). And finally, the government used budget support forums as its default approach to coordination (piloting and so on). As budget support came de facto to an end, the GoR is now trying to speed up its efforts to be less aid dependent, coordinating aid through other forums, and paying less attention to donor dialogue.

These two case studies confirm the importance of the recipient’s leading role in coordination and the tendency by EU MS to underachieve at the programming and implementation level on the ground when the former is not strong enough. Failure to clearly establish and recognise respective comparative advantages and to give up individual interests appears to be at the core of a poorly coordinated landscape, even in a country like Myanmar where there is recipient engagement and the cooperation infrastructure is being set up practically from scratch. The Rwanda case, however, shows a more successful scenario of coordination where the leading role of the recipient charts the course and donors follow suit.

Conclusions

The concept of aid coordination is complex in many regards. Coordination is not an end in itself, but a means of attaining an objective, in this case more aid effectiveness and greater efficiency. Indeed, coordination can even detrimental (for example, policy compromises) or ‘expensive’ (for example, higher transaction costs or delays) in some cases.

Nonetheless, coordination has the potential to increase the impact of aid, thereby contributing to the aid and development effectiveness agenda. At the same time, coordination can reduce transaction costs for both recipients and donors. Although some of these benefits of coordination can be quantified, we see great potential in outcomes that cannot be measured (for example, in the case of Rwanda): strengthening alignment, ownership or institutional capacity, to name but a few.

Aid coordination and development effectiveness remain rhetorically high on the European agenda. However, the level of ambition shown by MS in this regard varies greatly and depends on a multiplicity of interrelated factors (which were discussed at the end of Chapter 4). Overall, MS do not fully buy in and ‘walk the walk’. To a certain extent, the EU seems to be practising ‘pretended coordination’ (for example, in the case of Myanmar): a number of coordination efforts are kicked off without sufficient evidence for improved impact. As EU aid coordination happens on a voluntary basis, MS tend to accept more readily what aligns with their interests, developmental and otherwise, or at least does not go against them.

Some of this reluctance to pursue stronger EU coordination relates to the role of European institutions on the one hand and MS on the other. EU donors often see the function of EU institutions vis-à-vis themselves as a ‘primus inter pares’ where EU institutions are expected to lead. Increased EU coordination, however, does not always have to mean a leading role for the EC/EEAS, even though this appears to be the dominant assumption. A possible way to help MS ‘buy in’ to more ambitious EU donor coordination is, where appropriate, to put them in the driving seat with responsibility for collective action.

Issues of lack of identification with the EU project, political shifts within MS that have rendered them less prone to compromises, and lack of clarity on objective comparative advantages seem to be key obstacles to greater coordination. The EU has a number of instruments in place but the main challenge is the willingness in European capitals to prioritise those approaches and, on this basis, to give a high priority to coordination efforts within partner countries. Coordination, however, is a technical challenge only to a limited extent, as it is also hostage to the political commitment of aid actors. In this regard, the case of Myanmar shows that the objective of aid coordination might compete with other objectives of aid providers and that the final outcome is the result of a complex trade-off.

More generally, donor coordination without the leadership of the partner countries is only the second-best option. The role of partner countries in helping establish comparative advantage and use division of labour cannot be understated. As the Rwanda case study demonstrates, benefits can be optimised when there is successful donor coordination, full alignment with a recipient’s developmental objectives using the partner country systems and, crucially, strong government leadership.