Introduction

Since the 1990s, the number of development donors and projects has multiplied significantly. Donors operate in many different developing countries and in various sectors within those countries, resulting in aid fragmentation (Fengler and Kharas, 2011, p. 4). Aid agencies employ their own strategies, procedures and requirements, which translates into a huge burden for recipient countries (Birdsall, 2008, p. 523; Aldasoro et al, 2010). They have their own motivations, priorities and strategies and apply their own allocation criteria, resulting in overlaps or duplication, as well as funding gaps or orphan sectors (Barry and Boidin, 2012, p. 656).

In view of these challenges, strategies to improve the quality and the effectiveness of aid have risen to the top of the international development agenda, particularly since the dawn of the new millennium, culminating in 2005 with the Paris Declaration on Aid Effectiveness. Coordination is now a guiding principle in helping to resolve the problem of aid fragmentation. By lowering transaction costs, supporting policy dialogue, increasing transparency and strengthening management capacities, coordination should reduce the burden on recipient country administrations (Rogerson, 2005; Faust and Messner, 2007; Barry and Boidin, 2012).

The European Union (EU) has on many occasions expressed its ambition to foster the international agenda on coordination and to promote its implementation. Simultaneously, the EU has taken several steps to strengthen internal EU coordination, guided by the principles enshrined in the European Consensus (2006) and operationalized in the Code of Conduct on Complementarity and Division of Labour (2007) and the Operational Framework on Aid Effectiveness (2009, revised in 2011). However, the limited empirical evidence suggests that the EU might be unable to effectively coordinate in practice (cf. OECD-DAC, 2012, pp. 76–79; O’Riordan et al, 2011, p. 112). This article aims to explore and understand this gap.

Scholarship on the coordinator role of the EU in developing countries is still scarce. Although researchers have given considerable attention to international donor coordination (de Renzio et al, 2004; Rogerson, 2005; Bigsten, 2006; Menocal and Mulley, 2006; Hyden, 2008; Hayman, 2009; Aldasoro et al, 2010) and to EU coordination at the policy level in Brussels (Arts and Dickson, 2004; Gill and Maxwell, 2004; Hoebink, 2004; Bretherton and Vogler, 2006; Carbone, 2007, 2010; Mürle, 2007; Schulz, 2007; Orbie, 2012), few studies have looked into the EU’s coordinator role in the field.

The few existing studies (cf. ECDPM, 2007; Söderbaum and Stalgren, 2008) suggest that there is only limited ‘European’ coordination. Although it is recognized that there is increasing ‘Europeanization’ at the strategic level in Brussels, the situation in the field is possibly characterized by more division than unity among European donors (Bué, 2010). An evaluation of the implementation of the Maastricht Treaty found that EU coordination mechanisms had mainly served to improve the exchange of ideas and information on activities (Hoebink, 2004, p. 41). A 2007 evaluation found that the EU has a ‘remarkably low level of performance’ when it comes to coordinating (ECDPM, 2007, p. 35).

However, the field research on which these studies are based dates from the period in which EU policies were still in the design phase or had only just been adopted. Implementing the coordination agenda takes time, not least because of donors’ multi-annual programmes. Some more recent evaluations (cf. O’Riordan et al, 2011, European Commission, 2011a, Delputte and Söderbaum, 2012) partly address this concern, but they do not include in-depth investigations of the explanations for the findings. Consequently, literature on the EU as a development actor has left an important question unanswered: How are the EU’s ambitions translated at the country level and which factors enable and constrain EU coordination? This article addresses this question by investigating the EU’s coordinator role in Tanzania and Zambia.

The literature on aid effectiveness offers insights into donor coordination in general, which may help to understand the specific case of the EU. However, it tends to emphasize institutional constraints and considers coordination a bureaucratic process, in which different organizations decide to solve a collective action problem, which conflicts with the internal organization and corresponding incentive structures of individual donor agencies (Faust and Messner, 2007, p. 1; Schulz, 2007; Harrison and Mulley, 2009, p. 283; Sundewall et al, 2010). Although this approach has emphasized the importance of political aspects (cf. de Renzio et al, 2004; Rogerson, 2005; Faust and Messner, 2007; Aldasoro et al, 2010), it tends to neglect the importance of development traditions, identities and perceptions that might shape development actors’ engagement with coordination. Consequently, this article contributes to the growing research on the EU’s role as an international development actor (cf. Gänzle et al, 2012; Carbone, 2013) but the findings are of relevance to issues of donor coordination in general.

The article is structured as follows. The next section presents the conceptual foundations and clarifies our research approach. The second section constitutes the empirical analysis and shows that EU internal and external coordination in Tanzania and Zambia has been limited. The third section analyses the factors that enable and constrain EU coordination, pointing to not only institutional but also ideational and political factors. The last section presents some concluding remarks on the importance of identity and legitimacy for the EU, on the relationship between internal coordination, and on suggestions for further research.

Aid Effectiveness, Coordination and the Role of the EU

The International aid Effectiveness Agenda and Coordination

Simply increasing aid budgets would not solve the problem of aid fragmentation identified in the introduction to this article. The solutions put forward as part of the international aid effectiveness agenda concentrate on better coordination, concentration and specialization. Through better coordination, donors aim to increase transparency, rationalize their objectives and procedures, improve the quality of the policy dialogue and make better use of different actors’ competences (Barry and Boidin, 2012, p. 648). Through specialization and concentration on fewer countries and sectors they seek to tackle the problem of ‘too many cooks in the kitchen’ (Faust and Messner, 2007, p. 1), avoid overlaps and duplications, and reduce the accompanying transaction costs (Kjellman et al, 2003, p. 858; Faust and Messner, 2007, p. 13).

Aid modalities such as joint policy dialogues and conditionality frameworks aim to improve coordination. Harmonized approaches to budget support and sector support such as sector-wide approaches and pooled funding arrangements have been put forward as instruments to improve coordination and reduce fragmentation. It is important to note that improved coordination benefits donors as well. One study found that €2 billion per year could be saved by reducing the fragmentation of EU aid (Carlsson et al, 2009, pp. v–vi). Bigsten et al (2011) estimate the gains resulting from full coordination of country allocation among EU donors to be up to €8 billion.

Aid effectiveness was put on top of the international development agenda after the UN Conference on Financing for Development (2002) and coordination was portrayed as one of the leading strategies (Barry and Boidin, 2012, p. 644). The OECD-DAC has organized four High Level Forums on Aid Effectiveness, namely, in Rome (2003), Paris (2005), Accra (2009) and Busan (2011), where the international donor community, developing countries and civil society organizations agreed on international principles and a monitoring process. The Paris Declaration has served as the cornerstone of this agenda, as it was the first explicit agreement between donor and recipient countries to improve coordination and align aid to national development strategies.

However, coordination as such will not solve the problem of fragmentation. Without concomitant efforts to reduce the number of actors involved, it may even aggravate existing problems (Faust and Messner, 2007, p. 14). Hence, there has been a growing consensus on the need to go beyond mere coordination and to focus on specialization and division of labour whereby donors concentrate on those sectors and countries where they can bring added value (Schulz, 2009). Partly because of the EU’s efforts, the Accra Agenda for Action (2008) includes commitments for country-led division of labour (ActionAid/ECDPM, 2008; Keijzer and Corre, 2009; Schulz, 2009).

Although the aid effectiveness agenda has in general received broad support, there have also been some important points of criticism. It is dominated by the ‘traditional club’ of OECD-DAC donors (Eyben, 2013). Although the Working Party on Aid Effectiveness can be seen as a global policy space, it is closely linked to the OECD-DAC and the Paris principles were mainly established by the OECD-DAC donors. However, the emergence of new players, such as private actors and non-OECD-DAC donors such as China, India, Brazil and South Africa, has fundamentally changed the donor landscape (Fengler and Kharas, 2011, p. 6). Therefore, critics argue that ‘the OECD should give up control of the aid agenda’ (Glennie, 2011b), otherwise the Paris Agenda ‘risks reverting to a pre-Paris Declaration mode in which traditional players will be reasonably coordinated but the overall system will remain fragmented, resulting in sectoral and geographic misallocations of aid’ (Fengler and Kharas, 2011, p. 6).

Critics (cf. Burrall et al, 2006; Armon, 2007; Booth, 2008; Odén and Wohlgemuth, 2011) have also emphasized that the Paris reforms mostly concern technocratic approaches of efficiency in aid management and delivery. Given the political nature of aid and development problems, doubts have been raised about the feasibility of translating these principles into practice. Indeed, evaluations reveal that implementation at the country level has lagged behind. Although in general the results are ‘sobering’, the lack of progress on the implementation of the Paris Declaration is particularly notable for those indicators where the responsibility for change lies primarily with donor country governments (OECD-DAC, 2011).

The fourth High Level Forum in Busan had to give a response to these challenges. Apart from the discussions on aid and development, Busan would also be a test case for the emerging world order (Atwood, 2011; Glennie, 2011a, 2011b). It came as no surprise that China, India and Brazil only agreed upon the outcome document on the last day of the conference. The Global Partnership for Effective Development Cooperation puts more emphasis on ‘development effectiveness’. It builds on the principles of the Paris Declaration but has not further concretized the commitments on coordination or harmonization (UNDP-OECD, 2013).

The Federating Role of the EU and Coordination

The EU has been a major proponent of this international aid agenda. Internally, it has also envisaged taking a ‘federating role’ in development policy since the 2000s (OECD-DAC, 2007, p. 13), which means that it aims to coordinate EU member states’ (EUMSs) aid policies. Observers refer to a ‘new season’ (Carbone, 2008) or a ‘metamorphosis’ (Bué, 2010, p. 43) of European development policy since the turn of the millennium, characterized by a strong rhetorical commitment to coordination and by the gradual elaboration of operational initiatives in this area.

However, this ambition is not entirely new. Already at the outset of the European integration process, EUMSs discussed the possibility of a common development policy, although opposition of the national governments and aid administrations made it clear that full integration would never be an option (Carbone, 2010, p. 18). Throughout the 1970s and 1980s, the Commission launched several initiatives (for example, European Commission, 1972, 1975, 1984) to homogenize assistance practices and objectives but these received reluctant reactions in the Council (Grilli, 1993, p. 74). The Treaty of Maastricht (1993), which provided the EU with a legal basis for development policy and laid down three core principles of complementarity, coordination and coherence (Hoebink, 2004; Carbone, 2007), raised expectations of improved coordination. However, most EUMSs remained hesitant about a transfer of their sovereignty to the EU level.

Around the turn of the millennium, a combination of internal and external factors enabled the EU to manifest its coordinator role – at least at the level of policy formulation (see Bué, 2010, pp. 351–359; Holland and Doidge, 2012, pp. 101–109, 183–190). The increased international attention to donor coordination described above provided a window of opportunity for the EU to deepen and strengthen its coordinating role internally and to present itself as a major supporter of the new aid agenda.

In 2005, for the first time in 50 years of development cooperation, the Council, the Commission and the Parliament agreed on a common development statement: the European Consensus. The Consensus is based on the aid effectiveness agenda and symbolizes the EU’s commitment to providing more and better aid through better coordination and complementarity (Carbone, 2007). It prescribes that the Commission will facilitate coordination and ‘be one of the driving forces to promote EU delivery of its commitments made in Paris’ (European Union, 2006, p. 9). The Consensus did not include a thematic framework or a concrete action plan (Carbone, 2007), but served as the main reference point for several ambitious policy initiatives that would be launched by the Commission in the following years. More specifically, the Action Plan for more, better and faster aid (European Commission, 2006), the Code of Conduct (2007) and the Operational Framework on Aid Effectiveness (2009, revised in 2011) rely on working towards increased division of labour and joint multi-annual programming.

The Paris Declaration included neither indicators for Division of Labour nor a concrete strategy to divide the responsibilities among the donors. Observers therefore considered the EU as ‘fertile grounds for the application and feasibility test of new concepts like complementarity, because of its member states’ long-standing experience with the supranational modus operandi’ (Schulz, 2007, p. 2). The Code of Conduct laid down operational but voluntary principles for complementarity. In essence, EU donors committed themselves to focusing on a maximum of three sectors per country, in areas where they can bring added value, with a maximum of five EU donors per sector. Since 2008, the EU has implemented the EU Fast Track Initiative (FTI) on Division of Labour and Complementarity.

The EU’s Joint Multi-annual Programming Framework aims to reduce the transaction costs of programming and avoid duplication. Starting with a joint analysis of country situations by both the Commission and EUMSs, it intends to lead to collective response strategies to the challenges raised in the joint country analysis in the medium term, and in the long term to joint programming.

The Treaty of Lisbon, which came into force on 1 December 2009, also reflects the EU’s ambition to foster more coherence and coordination in its external policies. It confirms the shared competence between the EUMSs and the EU institutions in the field of development policy and reiterates the political commitment of EU coordination in this area. It also entails some institutional changes with implications for aid coordination such as: (1) the creation of the European External Action Service (EEAS), led by High Representative (HR) Ashton, which is now co-responsible for the programming of aid together with the newly established Commission’s DG Development and Cooperation – EuropeAid (DEVCO); and (2) the transformation of Commission Delegations into EU Delegations and the accompanying additional responsibilities. These new provisions generate opportunities and challenges for the future of EU development policy (see, for example, Van Seters and Klaver, 2011; Carbone, 2012; CONCORD; 2012; Furness, 2012).

In the absence of recent academic studies on EU coordination on the ground, assessments of the impact of the EU’s commitments rest upon a few policy evaluations commissioned or executed by the European Commission (cf. O’Riordan et al, 2011; European Commission, 2011b) and the OECD-DAC, which indicate that despite some progress the ambitions of the EU’s ‘federating’ role remain a ‘challenge’ (OECD-DAC, 2012, p. 76). Regarding division of labour, the European Commission (2011b, p. 1) concludes that while ‘there has been encouraging progress’, ‘Division of Labour is demanding and takes time to yield measurable results’. Similarly, although in some countries the EU has made progress in working towards joint programming, ‘there has been little enforcement of what has been agreed at the EU level to date’ (O’Riordan et al, 2011, p. 112).

Research Approach

Starting from this mismatch between the EU’s ambitions and the suggested weak coordination at the country level, as well as the limited number of recent in-depth studies of EU coordination on the ground, this article explores the federating role of the EU in Tanzania and Zambia, based on field research conducted during 2011. From a methodological perspective, Tanzania and Zambia constitute two ‘critical’ cases in the sense that they are ‘critical’ or ‘pivotal’ ‘to any understanding offered by the research’ (Ritchie et al, 2003, p. 80) – in this case the EU’s coordination role in practice. First, the EU has a clear ambition to play a central role in donor coordination and improve internal EU coordination in these countries. In 2006, Tanzania and Zambia were selected to pilot joint EU strategies and since 2008 they have been part of the EU’s FTI on Division of Labour. The EU as a whole is one of the main providers of aid to both countries. Moreover, both have been categorized as countries with more than five EU donors designating priority, which makes EU coordination all the more important (Carlsson et al, 2009, pp. 17–19). Second, Tanzania and Zambia are dependent on aid from a large number of donors. They have been portrayed as best practices for international donor coordination (Harrison and Mulley, 2009, p. 271; Söderbaum and Stalgren, 2008, p. 6; Odén and Wohlgemuth, 2011, p. 1). There are several operational coordination frameworks in place. The OECD-DAC donors have signed a Joint Assistance Strategy (JAS) and participate in a donor-wide coordination platform, namely, the Development Partners Group (DPG) in Tanzania and the Cooperating Partners Group (CPG) in Zambia.

Epistemologically, the research starts from the assumption that social conduct is not just shaped by actors or ‘by the environment or a structure’, but also ‘by the way that environment is defined or interpreted by the actors under study’ (Dessler, 2005, p. 598). Qualitative, in-depth semi-structured interviews were judged the most appropriate method to gather information on the views, attitudes, experiences and perceptions of the EU’s coordinator role at the country level (cf. Bogner and Menz, 2009). Thereby, roles are defined as ‘patterns of expected, appropriate behaviour’, and ‘encompass both an actor’s own considerations’ and the ‘expectations, or role prescriptions, of other actors’ (Bengsston and Elgström, 2011, p. 114). The interviews served to provide insights into and reconstruct the role of the EU by accessing ego and alter perceptions (cf. Arts, 2000, pp. 139–140). Ego perceptions are key players’ views with regard to their own role in key coordination processes. Alter perceptions are the views of other key players with regard to the role of ego in these coordination processes. Thirty-four representatives from EU Delegations, bilateral EU donors, multilateral and non-EU bilateral donors, as well as recipient government and civil society representatives, were asked to reconstruct coordination processes and the role of key actors, with a particular focus on the EU.Footnote 1 The fact that perceptions of the EU appeared to be fairly homogeneous across the two countries reinforces the validity of the findings.

The EU as a Coordinator at the Country Level

This section analyses what EU coordination means in practice by differentiating between its external and internal dimensions. The former concerns the multilateral perspective and refers to coordination between the EU and other international partners. The internal dimension relates to coordination between the EU and the EUMSs only.

External Dimension

To assess the external dimension, we focus primarily on the engagement of EU Delegations in the donor-wide coordination processes, and in particular in the areas of the general policy dialogue, division of labour and budget support. In Tanzania and Zambia, EU donors participate in local coordination processes around the DPG/JAS for Tanzania (JAST) and the CPG/JAS for Zambia (JASZ), respectively, as well as in their various sub-groups.

The EU Delegations have not played a major role in the Tanzanian and Zambian general aid coordination processes. Despite the size of their programmes and their active participation and contributions, the EU Delegations are mostly seen as ‘just another donor’ in the DPG/CPG: ‘the Commission is one of the other cooperating partners’ (Interview 1) or ‘one around the table’ (Interview 2). In Tanzania, aid coordination was initiated in the late 1990s by a group of Nordic countries, leading to the publication of the Helleiner Report, which called for more ownership, better aid management by the government, and more alignment and coordination of donors. On the basis of this report, in 1997 donors and the government agreed upon the Agreed Notes, which formed the basis for the current aid coordination architecture. The JAST was developed in 2006 under the lead of the World Bank (WB) and the United Kingdom in close cooperation with the government. Today, UNDP serves as the general DPG secretariat, whereas sub-groups are chaired by a rotating sector lead.

In Zambia, the path towards coordination was also paved by the Nordic Plus countries before the European Consensus and the succeeding initiatives were adopted. In 2003, the group of Nordic Plus countries initiated the Harmonization in Practice initiative (Wohlgemuth and Saasa, 2008, p. 5). The group aimed to harmonize their projects, align their support to the governments’ priorities, increase the use of budget support and better coordinate their inputs into policy dialogue (Fraser, 2009, p. 317). In 2004, this initiative was expanded to include all major donors, which led to the creation of the Wider Harmonization in Practice and eventually to a Memorandum of Understanding for Harmonization of Government/Donor Practices for Aid Effectiveness with the Ministry of Finance and National Planning (Wohlgemuth and Saasa, 2008, p. 5). The CPG is chaired by a Troika, which was chaired by the WB until 2012.

Apart from the Troika and rotating chairmanship systems, there is no clear hierarchy in the DPG and CPG architectures and, as a result, the EU Delegations have no formal role in the overall donor coordination structure, nor a mandate to enforce common positions or to speak for the donor group. In both countries, in line with the JAST and the JASZ and the division of labour in the policy dialogue, the EU Delegation has taken lead roles in sectors where it has a comparative advantage, such as in the road sector, public financial management and budget support. In the areas where the EU Delegation has acted as a lead, it is perceived as a donor that has helped the donor group move through the coordination process.

Regarding division of labour, the cornerstone of the EU’s aid effectiveness agenda, the EU’s FTI on Division of Labour has not acted as an engine for the donor-wide division of labour in these countries, as donors were already working on a donor matrix before the EU Council Conclusions were adopted. In contrast, the perception is that ‘the Commission jumped upon the band wagon’ (Interview 3). It should be emphasized that the division of labour process takes place in the context of the JAST/JASZ, which is not strictly speaking an EU affair. In Tanzania, the JAST included a chapter on Division of Labour, offering a strategy on how to proceed, including proposals on selection criteria and the process (Government of Tanzania, 2006). In Zambia, the elaboration of a donor matrix took place in 2005–2006 (Weingärtner, 2008, p. 16). The specific contribution of the EU consisted of developing a proposal on a matrix for EU donors but the exercise was mainly developed in the framework of the JAST/JASZ in cooperation with the government. Therefore, several interviewees (Interviews 4, 6, 7, 11, 17) did not attribute further progress to the efforts of the EU.

Both countries have been supported through large shares of budget support. In Tanzania, Poverty Reduction Budget Support (PRBS) has been initiated by the group of Nordic Plus countries and it has become the most important donor group in Tanzania in which both EU and non-EU budget donors participate. As the EU has been one of the largest providers of budget support, it has been a strategic partner for others, but overall the EU Delegation plays a role similar to that of the bilateral donors. In Zambia, the EU started with budget PRBS in 2001, and in 2005 other donors joined and signed a Memorandum of Understanding on PRBS. Furthermore, according to several interviewees, together with the WB, the EU bolstered the development of the Performance Assessment Framework, the monitoring system for PRBS. Additionally, when in 2009 corruption scandals in the Zambian health and road sectors led to a crisis in donor-recipient relations, the Commission was chairing the PRBS group and was responsible for keeping the donor group together while at the same time conducting the dialogue with the government. Although it was extremely difficult to take a unified and harmonized position in the divided donor group, under the lead of the EU Delegation, the group managed to agree on a way forward. Several interviewees (Interviews 3, 5, 9, 10) interpreted the coordination process under the lead of the Delegation as successful, despite the fact that there were different interpretations resulting in different disbursement decisions among EU donors.

To summarize, although in Zambia the EU Delegation has played a particularly proactive role in the area of budget support, the field research found that the EU Delegations are mainly acting as ‘just another donor’ in the overall donor-wide coordination processes: a donor that participates in coordination exercises and shares responsibility with the other donors for the success of the exercise, but no more than other participants.

Internal Dimension

To explore the internal dimension of the EU’s coordinator role, the following paragraphs look briefly at coordination in the programming phase but concentrate on coordination during implementation, considering the degree of institutionalization, the agenda and the main objectives of EU coordination.

Coordination among EU donors is limited in the programming phase. For the programming of the 10th European Development Fund (EDF), the Country Strategy Paper for Tanzania was based on the joint country analysis that was conducted in the context of the JAST, while in Zambia the EU Delegation used the national development plan and the JASZ as guiding documents but conducted its own country analysis. In both countries, the EU Delegations formulated their own response strategy just like all other EU donors did for their programming. Although in 2006 Tanzania and Zambia had been selected as pilot countries for the implementation of joint EU strategies (O’Riordan et al, 2011, p. 16), EU donors did not engage in joint programming for the 11th EDF.Footnote 2

In the implementation phase, EU coordination mainly serves as a communication mechanism, facilitated by both formal and informal practices. In Tanzania, EU coordination is institutionalized through regular meetings, which mainly concentrate on information exchange and ad hoc coordinated initiatives. The EU Head of Cooperation chairs these meetings, which have a clear agenda, namely, information sharing, the implementation of the EU FTI on division of labour and other EU-specific policies, issues of shared interest, and questions related to the boards of the multilaterals. In Zambia, EU coordination is not institutionalized by means of a fixed meeting schedule but there is event-based information exchange and consultation. Consequently, although a fixed schedule is absent, meetings have taken place to exchange information and consult on issues related to the EDF programme, on existing as well as new EU policy initiatives, or on specific problems. Apart from coordination meetings, EU donors also coordinate or share information on an informal basis,Footnote 3 via e-mail, and at bilateral level, as the EU Delegation is not always the ‘central hub’ of coordination among EU donors.

The following paragraphs will briefly discuss the three elements of EU coordination in practice: information exchange and consultation; the monitoring of the implementation of EU policies on division of labour; and the occasional adoption of common positions or joint initiatives.

First, in both countries, EU coordination is mostly about exchanging information about EU donors’ programming, modification of programmes or visits. However, information exchange is mostly going in one direction, from the EU Delegation towards the EUMSs, which is explained by EUMSs’ decision-making power in the EDF committee. EU donors thus see an interest in monitoring how the Commission is spending EUMSs’ taxpayers’ money. Although information sharing should help to avoid overlap and duplication and ultimately reduce the burden on the developing countries’ governments, contributing to development effectiveness, interviewees suggest that objectives are mostly donor-related, such as ensuring access to information in the case of smaller donors and facilitating control over EDF funds managed by the Commission.

Second, EU coordination meetings are used to monitor the implementation of specific EU policies such as the EU FTI on Division of Labour. This is mostly considered a technical exercise to monitor progress as there is a need to follow up on this Council Conclusion to which EU donors have agreed in Brussels. In Tanzania, the Commission is lead facilitator for the FTI on Division of Labour, while in Zambia, Germany is lead facilitator. The facilitating donor needs to monitor the process and report back to Brussels where the results of the different FTI countries are compared and recommendations are made. The monitoring of the FTI was described as being ‘obliged to fill in a questionnaire’ (Interview 4). The Code of Conduct was adopted at a time when coordination around the JAST and the JASZ was already advanced and a donor matrix had already been elaborated. These donor matrices and roadmaps have not led to major reallocations, although there are a number of delegated cooperation agreements in place with arrangements on lead and silent donors. For example, in Tanzania, the EU has a delegated cooperation agreement with Germany in the health sector and with the United Kingdom in the education sector. In Zambia, Germany delegated a part of its cooperation in the education sector to the Netherlands, and in the field of justice, the EU engaged in delegated cooperation with Germany.

Third, the EU framework is in general not used to speak with one voice, adopt common positions or launch joint initiatives, although this has taken place on an ad hoc basis. In Tanzania, EU donors occasionally discuss issues of common interest, which can then be taken to the DPG. For example, in 2010 the EU Delegation managed to coordinate an EU position on the draft text of the Tanzanian PRSP so that the EU had a common view and shared general comments on this. These EU comments were then incorporated into the formal response of the DPG on the draft PRSP. Another example is an EU consensus on allowances and per diems of training sessions and workshops for government officials, which was then proposed to the DPG. In Zambia, the response to the crisis following the revelation of the large-scale corruption cases constitutes an example of how EU donors tried to harmonize their messages towards the government.

In sum, EU coordination in Tanzania and Zambia tends to be limited to information sharing and ad hoc initiatives, while largely lacking a component that would require modification of EUMSs’ aid practices. Although Tanzania and Zambia are FTI on Division of Labour countries and there are a number of instances of delegated cooperation among EU donors, EU donors do not apply uniform allocation criteria and the donor matrices were developed under the JAST and the JASZ. Although joint programming is the ultimate level of coordination envisaged in an EU framework, this will not take place in Tanzania and Zambia in the context of the 11th EDF.

In Search of Explanations

Our field research in Tanzania and Zambia shows that the EU Delegations act mainly as ‘just another donor’ in (external) donor-wide coordination while (internal) EU coordination mainly serves as a communication mechanism. How can we explain this limited coordinator role at the country level? As mentioned in the introduction, the existing literature has mainly focused on institutional barriers. Our field research confirms that the existing aid architecture in Tanzania and Zambia and the EU’s organizational structure and procedures constitute constraining rather than enabling factors. The advanced existing donor-wide coordination processes influence EU donors’ readiness to engage in EU coordination. This risks creating parallel structures:

We don’t see very much value added of a strong EU coordination in general terms, because we have those other groups and it’s in a way more important to have the whole group behind you, rather than only the EU members. (Interview 2)

EU donors have not aligned their programming cycles, and the EU’s complex aid bureaucracy constrains the Delegations’ role of acting as a hub for EU coordination. In spite of reforms and steps towards decentralization, the Commission’s aid bureaucracy is still perceived to be too complicated, suffering from administrative delays, and subject to centralized checks within a burdensome hierarchy. The Commission is ‘a supporter of new and good ideas, but it’s quite difficult for them to act as an engine’ (Interview 5), as ‘by the time the Commission gives the green light, its ideas are already superseded by what we have already agreed upon in the sectors’ (Interview 6). It ‘needs to get a level of flexibility that is much, much larger and the best way of doing that of course would be to decentralize decision making’ (Interview 7; also Interview 8). In particular, its limited authority over disbursements of budget support and new horizontal or thematic initiatives that are launched at headquarters level hinder coordination efforts.

However, we also found that EU donors such as the United Kingdom, Denmark and the Netherlands, which are more decentralized and have some flexibility to coordinate at the country level (OECD-DAC, 2009; O’Riordan et al, 2011, pp. 23–24), are less interested in strengthened EU coordination (Interviews 7, 11, 12). In addition, existing evaluations have found that in theory all EU donors in Tanzania and Zambia are able to participate in joint programming (O’Riordan et al, 2011, p. 9). Our field research clearly indicates that in order to understand why EU donors will not engage in more advanced forms of coordination such as joint programming, we should take ideational and political factors into account.

First, EUMSs’ opposition towards EU-specific coordination has a clear ideational element. Coordination can be seen as a norm, which is introduced to maintain order in donor activities in the absence of a superior power that structures aid within a country to improve development effectiveness. In Tanzania and Zambia, no interviewee contested the need for better coordination among donors and the government. Nordic Plus donors and the Commission have also provided a large share of their aid budget through harmonized aid modalities such as general and sectoral budget support, which gives an indication of their ideational consensus on the importance of the aid effectiveness agenda.

However, donors hold different ideas on aid modalities and on the ‘right’ level of coordination, and they have internalized the ideas surrounding aid effectiveness to various degrees (cf. Béland, 2009, pp. 704–705). Accordingly, EU donors frame development problems and their subsequent response in different ways, which is visible in their use of aid modalities. Not only have some EU donors remained reluctant to engage in programme and budget support, even among budget support donors there are discrepancies, as exemplified by the use of technocratic versus political conditionalities, which has resulted in dispersed response strategies (Molenaers, 2012). In Tanzania and Zambia, EU coordination is severely questioned by the United Kingdom and other Nordic Plus countries, which are also the largest EU donors present in these countries. They promote a multilateral approach to aid coordination at the country level. Some Nordic Plus interviewees (Interview 2, 6, 11) suggested that more intense EU coordination conflicts with their identity, which corresponds more with ‘supporting the principle of multilateralism’ than ‘being an EU donor’. This opinion also rests on a perception that through EU coordination, the EU Delegations aim to strengthen a common European identity rather than making EU aid more effective:

if the objective is to reach better coordination and eventually better effectiveness of our joint efforts, we should be more pragmatic and ask for example Switzerland to join [EU coordination meetings]. But if it is actually more about portraying the EU as more politically visible and as a more visible player vis-à-vis for example the US, then improved coordination and aid effectiveness is not the main goal. (Interview 11)

The importance of visibility for the EU was also confirmed by non-EU interviewees (for example, Interview 9). In practice, this also becomes clear through initiatives such as the ‘Europe House’ in Tanzania and common references to the ‘European Family’ in Zambia. However, other identities compete with the construction of an EU identity, be it because of national identity (UK Department for International Development (DFID), a well-respected influential and responsible donor on its own) or a collective identity (like-minded – or Nordic (Plus) donors): ‘the like-mindedness is a much stronger basis for doing joint work than belonging to the EU’ (Interview 14).

Second, the domestic economic and political context in Europe, characterized by the eurozone crisis and a political shift to the centre-right, is not conducive to engagement in joint approaches that require a redefinition of EUMSs’ own priorities.

There is a lot of discussions taking place at several of the EU countries, and a lot of questioning about aid and we have conservative governments in many of the EU countries that are … evaluating how we’re working …. (Interview 5)

The window of opportunity after the turn of the millennium that led to the adoption of the European Consensus and the Code of Conduct seems to have now closed. Despite the political commitment for EU coordination, the operational initiatives have, to a large extent, remained voluntary, which makes it vulnerable to EUMSs’ domestic politics. Implementing a better division of labour or engaging in joint programming requires high-level political support. However, the governments that endorsed the Rome and Paris Declarations have changed and even donors that were recognized for their commitment to aid effectiveness (for example, the Netherlands, Sweden and the United Kingdom) have turned towards approaches that allow for more control and for using aid as a foreign policy instrument (Interviews 5, 11, 15, 16). The shift is especially visible in the EU donors’ growing scepticism towards budget support, the most drastic example being the Dutch decision (2009/2010) to stop budget support programmes in both Tanzania and Zambia. In several EUMSs, foreign aid budgets are declining and agencies in the field experience increased pressure to focus on short-term successes of national aid policies, to the detriment of less visible coordination processes. Although visibility has been a primary concern for individual EU donors in ‘well-performing’ countries such as Tanzania and Zambia, donors now feel even more pressured to maintain this image.

The budgets are under pressure, we all have to show what we have realized with our money, what are the results, so then we engage less quickly in the overall alignment, which reduces the visibility as a bilateral donor. (Interview 15)

Concluding Remarks

The exploratory field research presented in this article points to the limited nature of the EU’s coordinator role. There are few indications that reveal ‘a European approach’ to coordination at the country level. Although the Commission is committed to the implementation of global commitments on aid effectiveness, and has been able to play an important role in a couple of cases by acting constructively and setting a good example, its coordinator role is much less developed. Although the mere fact that the EU promotes coordination may ensure that issues such as duplication and fragmentation take centre stage in aid discussions, in the field the EU’s pursuit of coordination has also been interpreted as an attempt to strengthen its international profile, to assert a European identity and to ‘add legitimacy to development assistance by highlighting efforts to spend resources wisely’ (Lundsgaarde, 2012, p. 709). From this perspective, given the ‘differential Europeanization of national identities’ (Risse, 2012, p. 43), EU coordination may compete with EUMSs’ identity as donors in a way in which other forms of donor coordination do not.

However, there were no indications that attempts to reinforce EU coordination conflict with its ambition to support ongoing multilateral coordination. As EU coordination is mostly focused on information exchange and reporting on internal EU issues, it is perceived to be complementary rather than fostering or hindering donor-wide coordination. Paradoxically, within the EU, the countries that have been least enthusiastic about European integration of development policy have been most supportive of the Paris Agenda (Orbie, 2012). These – mainly Nordic Plus – countries have ‘emphasized that the EU should support existing coordination efforts, such as the PRSPs and Joint Assistance Strategies’ (Carbone, 2010, p. 23).

These insights suggest that the donor composition, and specifically the presence of Nordic Plus donors, as well as the advanced existing donor coordination frameworks in developing countries, are important factors for explaining the EU’s limited coordinator role in the field. To investigate whether these suggested explanations are valid, further research could look at EU coordination in settings with less advanced existing donor-wide coordination or with a donor landscape dominated by EUMSs such as France, which have aimed for a European solution to coordination.

In the future, the decreased power asymmetry between EU donors and Tanzania and Zambia, resulting from the declining aid dependency and the growing presence of China, may encourage more EU coordination. Several interviewees, mainly representatives of smaller and medium-sized donors, recognized the need to maintain leverage by better coordinating their responses towards the governments. Consequently, while the changing landscape, with the emergence of China, India, South Africa, Brazil and other non-OECD-DAC players, risks decreasing the EU’s leverage, the cost of non-implementation of the EU’s coordination ambitions might increase and possibly enable more advanced coordination in the future. However, so far there are no indications that the institutional, ideational and political constraints identified in this article can be easily overcome.

Interviews

Interview 1 with multilateral donor representative, Lusaka, January 2011

Interview 2 with EU Member State representative, Dar es Salaam, January 2011

Interview 3 with EU Member State representative, Lusaka, February 2011

Interview 4 with EU Member State representative, Lusaka, February 2011

Interview 5 with EU Member State representative, Lusaka, February 2011

Interview 6 with EU Member State representative, Lusaka, February 2011

Interview 7 with EU Member State representative, Dar es Salaam, January 2011

Interview 8 with EU Member State representative, Dar es Salaam, January 2011

Interview 9 with Zambian government representative, Lusaka, February 2011

Interview 10 with EU Member State representative, Lusaka, February 2011

Interview 11 EU Member State representative, Dar es Salaam, January 2011

Interview 12 with EU Delegation representative, Dar es Salaam, January 2011

Interview 13 with EU Member State representative, Lusaka, February 2011

Interview 14 with EU Member State representative, Lusaka, February 2011

Interview 15 with EU Member State representative, Dar es Salaam, January 2011

Interview 16 with multilateral donor representative, Dar es Salaam, January 2011

Interview 17 with multilateral donor representative, Dar es Salaam, January 2011