Swedish Development Cooperation and Ownership of African Regional Organizations

Abstract This article deals with the ownership of African regional organizations (ROs) in the context of an overreliance on foreign funding. It focuses specifically on whether and how Swedish development cooperation contributes to the ownership of ROs on the continent. Sweden constitutes a particularly pertinent case study for several reasons: it is one of the world’s largest and most important donors to African ROs, it places great emphasis on the ownership principle and it is often considered to be a role model for effective development cooperation. The empirical material consists of an in-depth study of Swedish regional development cooperation with sub-Saharan Africa (2010–2015), policy reports and reviews, aid evaluations and a range of semi-structured interviews with Swedish and African aid officials. This study reveals a series of challenges regarding what is to be owned and by whom, and Sweden’s stress on building the capacity of the secretariats of ROs as the main way of strengthening ownership.


Introduction
External powers and development agencies have supported regional integration and regional organizations (ROs) in Africa for many decades and for many reasons. In the past, most external actors were critical of state-led regionalist projects in Africa, in part because of their interventionistic approach and poor results. This attitude has changed dramatically in recent decades and many donors have now become enthusiastic supporters of state-driven regionalist projects. Particularly popular are the African Union (AU) and the so-called regional economic communities (RECs), such as the East African Community (EAC), the Economic Community of West African States (ECOWAS), the Intergovernmental Authority for Development (IGAD) and the Southern African Development Community (SADC).
The EU is one of the world's most prominent region builders. It has several billion Euros with which to fund many ROs and regional projects, particularly the AU and the RECs (Herrero and Gregersen, 2016). While the World Bank used to be reluctant to support state-led regional integration in Africa, in recent years the Bank has instead become a strong supporter. Between 2008 and 2011, it increased its support to Sub-Saharan Africa by 75 per cent, reaching US$3.65 billion in 2011 (World Bank, 2011). Similar support is given to African regional projects and ROs by many multilateral agencies and development banks as well as by some of the largest bilateral donors, such as Germany, Sweden, Canada, the Netherlands, Norway and Denmark. 1 Although external funding to regions and ROs remains limited compared to country-based development cooperation, its importance cannot be overstated. External funding makes up a considerable proportion of the total budgets of many African ROs. In fact, the AU and most of the RECs (except ECOWAS) depend on external funding for between 50 and 90 per cent of their budgets (Engel, 2015;Vanheukelom, 2016).
Many external powers and international development agencies declare that their objective is to help African actors implement their 'own' regional integration agenda (Söderbaum, 2016, p.147ff). The problem with such beliefs is that external actors and donors are not simply neutral financiers of home-grown and independent African ROs. On the contrary, donors are deeply influencing these ROs in a number of ways. Although donors may not directly steer the policy goals and design of the African ROs, they tend to influence or even indirectly control them. This begs fundamental questions about ownership (Buzdugan, 2013;Gray, 2011;Krapohl, 2017;Muntschick, 2017;Söderbaum and Brolin, 2016).
While the issue of ownership was generally overlooked in the past, in recent years it has been brought to the forefront of the policy agenda. The so-called Kagame report on the institutional reform of the AU questions whether African member states can 'own' the AU when they are so dependent on external funding (Kagame, 2017, p.17). The Minister of Foreign Affairs of Zimbabwe, Sibbarashe Mumbengegwi, comments on this issue with regard to SADC: as long as we are unable to fund our own organisation, the future of our programmes and activities will remain uncertain and SADC will not be wholly ours … our continued dependency on their [the donors] generosity and benevolence constitutes one of the most profound weaknesses of our organisation. … This state of affairs compromises ownership as well as the sustainability of our programmes. (quoted in Musengeyi, 2015) Although the overreliance on external funding has started to attract attention in policy circles, there is still little research in this field. While there is some academic literature on external funding and the budgets of ROs (Engel, 2015;Gray, 2011), there is hardly any on the ownership problematique specifically. Moreover, there is virtually no research on other donors than the EU. This article focuses on one particular donor in order to gain an in-depth understanding of how foreign funding impacts on the ownership of ROs. Sweden has been selected for several reasons. Firstly, Sweden is one of the largest and most active donors to ROs in Africa (Stapel and Söderbaum, forthcoming). Secondly, the principle of ownership lies at the heart of Swedish development cooperation, and its regional development cooperation is officially designed to help African actors and governments improve the performance of their ROs and implement a home-grown African integration agenda (Brolin, 2017;Söderbaum and Brolin, 2016). Thirdly, Sweden is often regarded as a role model for effective development cooperation. This stems in part from its positive track record in ensuring national and local ownership (Brolin, 2017). Consequently, one might expect Sweden to exercise less steering and interference than other donors, particularly those driven by ideological, neo-colonial or geopolitical interests instead of enlightened interests.
While a study of Swedish regional development cooperation may be valuable in its own right, Sweden is also relevant for comparative purposes. Sweden is one of the most important donors in this field and there are many similarities between Sweden and other donors such as the EU, Germany, Norway, the Netherlands and the African Development Bank (AfDB), who also give considerable support to the AU, the RECs and other intergovernmental ROs (Söderbaum and Brolin, 2016). Although some donorsfor instance, the United Kingdom (UK) and the United States (US)prefer to support regional projects outside the framework of intergovernmental ROs, most of them still support the African integration agenda.
This article therefore concentrates on two sets of research questions. The first deals with the meaning of ownership and whose ownership is focused on in Swedish development cooperation with ROs in Africa. This question is important because it is often unclear what ownership actually means, what is to be owned and by whom. The second set focuses on the strategy Sweden employs to strengthen ownership. This is to enhance our understanding of whether, how and with what effects Sweden deals with the issue of ownership.
The empirical material consists of an in-depth analysis of Sweden's strategy for regional development cooperation with Sub-Saharan Africa (2010)(2011)(2012)(2013)(2014)(2015). This was based on a range of policy documents, donor evaluations and semi-structured interviews conducted with Swedish policy makers in Stockholm and in field offices. 2 The article is organized as follows. Firstly, I explore what ownership means in relation to donor-supported regional cooperation and ROs. Secondly, I provide a brief overview of Swedish development cooperation with ROs in Africa, with regard to objectives and reported results. The following two sections concentrate on the two main research questions: firstly, whose ownership is in focus, and secondly, the 2 Nearly 20 interviews have been conducted with Swedish representatives between 2014 and 2017. An additional 20 interviews have been conducted with other donors and development banks as part of a broader research project on 'External funding of regional organizations in Africa'. means and strategies used to try to ensure ownership. Finally, some conclusions are drawn.
Unpacking 'regional' ownership The concept of ownership makes little sense if viewed in isolation from external influence and development cooperation. Although ownership is one of the key principles of the Paris Declaration on Aid Effectiveness, many consider it to be an ultimate goal. This is somewhat misleading because ownership is primarily a means to achieve greater aid effectiveness.
Discussions over the past two decades have generated a wealth of insights into aid effectiveness and ownership. Some of the most important are (i) that development cooperation is only one of many factors influencing governance transformations and economic development; (ii) that development and foreign aid are politicized, so providing only financial and technical solutions often fails to deliver results and (iii) that the domestic political circumstances must be understood by donors so that they can create realistic goals and appropriate aid modalities, and coordinate long term with other donors (cf. Brown, 2017;Hasselskog et al., 2017;TWP, 2014;Whitfield, 2009).
The term ownership is generally used to refer to recipients' capacity and right 'to set their own development objectives and define their own strategies for achieving these objectives' (Brolin, 2017, p.1). However, it is also used with a variety of particular implications. Much of the discussion about the concept has been linked to 'national' or 'country' ownership (Brown, 2017;Hasselskog et al., 2017). The OECD has agreed on a definition of national ownership as The effective exercise of a government's authority over development policies and activities, including those that relyentirely or partiallyon external resources. For governments, this means articulating the national development agenda and establishing authoritative policies and strategies. … For donors, it means aligning their programmes on government policies and building on government systems and processes to manage and coordinate aid rather than creating parallel systems to meet donor requirements. (OECD/DAC, 2006, p.147) Evaluations conducted by donors reveal that ownership is as important for 'regional' development cooperation as it is for country-based assistance (AfDB, 2012;European Commission, 2007;IEG, 2007;Norad, 2014;UNDP, 2013). However, defining and establishing the ownership of regional development cooperation has proven more challenging than in the case of traditional country-based (national) development cooperation. There is often confusion about the meaning of regional ownership, particularly regarding 'what' is to be owned and who are the relevant 'owners'.
Regarding 'what' is to be owned, it is important to distinguish among ownership of the African integration agenda, ownership of an RO and ownership of a regional project. ROs (especially the AU and the RECs) are often viewed as the main drivers of the African integration agenda, often through donor-funded projects. Since policy makers often use these concepts interchangeably, it becomes unclear precisely 'what' is to be owned. This article therefore focuses specifically on ownership of ROs rather than on ownership of regional integration and the regional agenda.
A somewhat more complicated question concerns 'who' owns an RO. Although development cooperation with ROs is often channelled through RO secretariats, these are not the only or most relevant 'owners'. In Africa, it is largely governments and member states that decide on the policies that the ROs are to follow and they may even conduct their implementation. This means that the member states are effectively the 'owners'. The growing literature shows that most ROs in Africa lack autonomy from their governments (Bach, 2016;Welz, 2013). Even among the better resourced ROs, such as the AU and the RECs, whose secretariats have greater autonomy, ownership is in practice shared between ROs and the government.
The donor community is coming to realize that effective development cooperation depends on African governments feeling responsible for their own ROs. One donor report notes, 'regional co-operation will only be successful when the participants feel it addresses issues which are important to them, when they benefit from the co-operation and when there is sufficient ownership of the regional institution or activity' (Norad, 2003, p.1). Similarly, one of the largest evaluations of multilateral and regional projects observes that, 'success and sustainability depend on strong ownership of all participating countries' (IEG, 2007, p.xxvii) and a major report on the AfDB's support to regional integration stresses that 'key factors of effectiveness include country commitment and ownership, implementation and governance arrangements as well as a conducive policy environment' (AfDB, 2012, p.vii). These conclusions are echoed in various other evaluations (cf. Söderbaum and Brolin, 2016).
The creation of 'national' ownership is thus critical for the long-term effectiveness of development cooperation with African ROs, implying that African governments and stakeholders must be incentivized to engage in this. However, there is a variety of relevant 'national' actors, including the presidents' office, the government, the line ministries, businesspeople and civil society.
While the role of donors has been debated, there has until recently been little discussion of the negative effects of development cooperation with African ROs. Gray's (2013) study of how external actors promote regional economic organizations in the developing world concludes that African ROs are driven more by the interests of powerful external actors than by those of the member states. Other scholars concur, arguing that donors are not simply passive financiers who are motivated to 'do good' (Buzdugan, 2013;Krapohl, 2017;Muntschick, 2017;Söderbaum, 2004). While few African ROs would survive without external assistance, these studies explain why the organizations tend to be unstable and are not 'owned' by regional actors. Obviously, donors' motivations vary in the degree to which they explicitly or implicitly exert pressure upon recipients or attach conditions.

A snapshot of Swedish development cooperation with ROs in Africa
The general objective of the Swedish Cooperation Strategy for Regional Development Cooperation with Sub-Saharan Africa (2010)(2011)(2012)(2013)(2014)(2015) is 'to increase the capacity and the political accord among the African intergovernmental communities and the countries concerned to manage transboundary challenges such as regional stability, trade and economic integration and sustainable development' (Government Offices of Sweden, 2010a, p.1).
Intergovernmental ROs hold a special place in Sweden's strategy, which includes strengthening the mandates and capacities of the AU and the RECs, in particular EAC, ECOWAS, IGAD, SADC but also COMESA. Sweden may also support other regional bodies, such as other intergovernmental ROs, civil society organizations, economic organizations or other national or regional initiatives. Support is to be provided to actors and mechanisms that best help realize the African integration and development agenda (Government Offices of Sweden, 2010a, pp.1-2). Although the concept of ownership is rarely mentioned in this strategy, it is clear that the AU and RECs are considered to be the main drivers of African integration and development.
Sweden has prioritized four interrelated themes: peace and security, environment and climate, economic integration (trade, business and financial systems) and institutional development. These account for more than 90 per cent of the project portfolio (see Table 1). Although there is some variation in implementation of the first three themes, much support goes 'directly' to the AU and the RECs and 'indirectly' through so-called 'equivalent' or 'supporting' ROs (e.g. the AfDB, the UN Economic Commission for Africa and river basin organizations) (Government Offices of Sweden, 2010a, pp.1-2). The fourth theme, institutional development, is focused on the AU and the RECs.
The Swedish strategy is to increase aid effectiveness in a number of ways. Firstly, a greater proportion of aid is allocated to the support of programmes. Secondly, greater coordination with other donors is sought, including discussions about corruption and Source . The data for 2015 are Sida's own estimates of disbursements. Regional research cooperation (543 million SEK) was 'removed' from the regional strategy in 2013; regional health has a separate strategy and is excluded from the general regional strategy. capacity building but also supporting the cooperating partners' own priorities so as to strengthen their sense of ownership (Government Offices of Sweden, 2010a, pp.5-6).
No specific guidance is given on how these objectives should be met, apart from by increasing collaboration with other donors, particularly the EU, the AfDB and various UN organizations (Government Offices of Sweden, 2010a, p.12). Sweden aims to contribute to the EU's region-building activities since they share the goal of promoting African ROs and regionalist agendas and because Sweden contributes funding to the EU framework.
Results have been reported in a range or reports and reviews, especially in so-called 'strategy reports'. Many strategy reports focus on examples of interventions to illustrate implementation (Sida, 2011; or on the '1-3 most important results the Swedish assistance has contributed to achieve' (Sida, 2013b, p.6). The results presented in these reports tend to be anecdotal. They claim that Swedish support has contributed to development and improved performance of ROs in areas such as peace and security (Sida, 2012;2013b;2015a). The AU and the RECs, for example, are responding to the crisis in South Sudan, Boko Haram's atrocities in Nigeria, conflict in Somalia and the recent Ebola outbreak (Sida, 2015a). In the field of the environment, Swedish support to regional adaptation programmes of the AU, IGAD, ECOWAS, EAC, SADC, UNECA, AfDB is also showing results. Sweden's support to the Lake Victoria Environmental Management Project II has resulted in the establishment of a regional environment fund and a strategy for sustainable land use (Sida, 2012;2013b). Also in the field of regional economic integration, 'substantial progress has been made' (Sida, 2013a, p.20). Examples of results include harmonized rules, the establishment of institutions, regulations (Sida, 2012;2013b) and a certification programme for those involved in import and export business to reduce the time and costs associated with international trade (Sida, 2013b).

Whose ownership?
Who should own Swedish development cooperation with ROs in Africa? There is a lack of clarity in Swedish policy circles about what ownership of ROs actually means. Consequently, Swedish policy makers often disagree about who is supposed to own their interventions. Swedish development cooperation in fact prioritizes the secretariats of AU and the RECs although it is claimed that other ROs, NGOs and civil society actors can also be owners. Very little emphasis is placed on national level actors.
A considerable proportion of Swedish regional development cooperation is used to provide so-called 'direct' and 'indirect' support to the AU and the RECs (Sida, 2013a, p.7). This distinction is a Swedish invention that is not used by other donors. Interviews reveal that Swedish officials often disagree about its meaning and relevance. 3 'Direct support' to the AU and the RECs is straightforwardthe lion's share of it goes to the AU and the EAC, some to ECOWAS, SADC and IGAD, but none to COMESA. The confusion concerns 'indirect support', which is given to ROs and partners that Sweden assumes will 'indirectly' contribute African integration. There is also both 'formalized' and 'non-formalized' indirect support (Sida, 2013a, p.24). The former refers to organizations and partners that have a 'formalized' partnership with the AU and the RECs, usually a Memorandum of Understanding (MoU). The latter is more diffuse and refers to partners that have no such formal partnership but which are believed to make an informal contribution.
A large proportion of Swedish assistance is allegedly direct and indirect (two-thirds of total allocations for 2010-2012) (Sida, 2013a), although Sida's statistics do not strictly support this (e.g. Sida statistics, 2015a; 2015b) figure since they do not specify the organizations or what kind of support is included in each of their categories. There is debate in Swedish policy circles about the meaning and relevance of direct and indirect support. The emerging position is that: Cooperation with other organizations that support RECs agenda is often a more efficient way to achieve results in the medium-term than supporting the RECs directly, as these organizations normally have better capacity and are less affected by political constraints. Intermediary organizations can furthermore support the implementation of regional integration through close collaboration with the REC secretariat as well as the REC member states simultaneously and thus ensure that ownership for the process is shared between the main actors. Hence, the indirect support to AUC/RECs assists the processes towards regional integration and opens up the AU/RECs to external cooperation and scrutiny, and to practical direct engagement of the African civil society in their work. (Sida, 2013a, p.7) Despite efforts to clarify the concept of ownership, there are still many questions. For instance, it is unclear how so-called 'collaborating' organizations and agencies (such as UNECA and the World Bank) and 'equivalent' intergovernmental ROs, such as the International Conference on the Great Lakes Region and the Nile Basin Initiative (Government Offices of Sweden, 2010a), should be categorized. Similarly, it is not clear how various projects categorized as indirect support (for instance to Kofi Annan Peacekeeping centre) 'indirectly' strengthen the AU and the RECs. This makes it difficult to find the logic behind the selection of organizations and aid interventions and to understand whose ownership is to be built. 4 The question remains as to 'what' exactly is to be owned: the African agenda or a particular RO? 'Who' is supposed to own indirect support and support to collaborating ROs? And why are member states almost excluded from the Swedish regional strategy?

National governments and national ownership
The Swedish strategy is strongly geared towards strengthening the institutional capacity of ROs and their secretariats. As indicated in the above quote, Swedish documents continue to describe the interdependence between RO secretariats and their member states. The mid-term review of the Swedish regional strategy states that 'it would be a mistake to try to build AU/REC capacity for practical implementation of programmes at national level. Implementation must be done by the member states and regional partners to AU/RECs, which are active at the national level' (Sida, 2013a, p.7). While this is a plausible statement, Sweden pays little attention to national actors and member countries. Interviews confirm that the Swedish approach simply assumes that national actors should be integrated into the work of an RO through Swedish bilateral country-based development cooperation. This distinction between regional and national actors is problematic, particularly concerning African ownership. As discussed previously, most other observers and donors contend that ROs are 'owned' first and foremost by their governments and only secondarily by their secretariats.
The Swedish RO-centred 'top-down' approach contrasts with the 'flexible' and 'bottom-up' approach favoured by donors such as the AfDB, WB, GEF and the EU. While the top-down approach works from the ROs (top) to the national (bottom) level, a more flexible bottom-up approach promotes regional development and integration through national strategies and programmes. The main strength of the bottom-up approach is that it involves national governments in the work and ownership of ROs. While a bottom-up approach may not replace a top-down approach, Swedish development cooperation could certainly better integrate national actors and member countries into ROs.
As mentioned in unpacking 'regional' ownership' section, there is increasing evidence that regional development cooperation is usually considered unsustainable unless it is not well integrated into national agendas or locally/nationally owned. One of the most comprehensive donor evaluations ever conducted on regional and multi-country assistance identifies five design features that have proven critical for their success (IEG, 2007, p.27). Most of these features relate to the importance of national government and national/country ownership: (i) strong national commitment, (ii) the goals match national and regional capacity, (iii) clear delineation and coordination of the roles of national and regional institutions, (iv) accountable governance and (v) planning for sustainability (which depends on strengthening national commitment and benefits). These points about the relationship between regional cooperation and national governments appear to have largely been overlooked by Swedish policy makers. Important lessons may therefore be learned from studying how other funders have designed their regional support programmes in ways that are more sensitive to national agendas and national ownership (AfDB, 2012;IEG, 2007;WB, 2011).

Capacity building and ownership
A core element of Sweden's strategy is to strengthen the capacities of the AU, the RECs and regional actors so that they take responsibility for implementing integration and assume ownership of their ROs (Government Offices of Sweden, 2010a; Sida, 2015a, p.2). Capacity building (including institutional development) is thus the main instrument by which Sweden aims to strengthen African ownership. However, capacity building is not the same thing as ownership and the relationship between the two is ambiguous in the Swedish strategy.
It is widely recognized in both academia and the policy community that African ROs face numerous constraints. Problems are noted in Swedish reports and interviews with Swedish officials. One of the main problems interviewees noted was the lack of institutional capacity and a poor work ethic in the AU and the RECs. This has repercussions for both performance and ownership (Sida, 2013d;2013e;2013f;2013g;2013h;2013i).  (Sida, 2013d;2013e;2013f;2013g;2013h;2013i; see Söderbaum and Brolin, 2016).
of the AU and the RECs noted by Sida. Staff are often politically appointed and lack sufficient knowledge in their field (Sida, 2013f), the AU Commission (AUC) lacks technical staff (Sida, 2013d) and many RECs lack institutional competence and capacity. Indeed, according to Sida reports, ECOWAS is considered so weak that several donors avoid making direct financial transfers to the organization (Sida, 2013g). Heavy dependency on external funding is also described as limiting the ability of ROs to take responsibility for the design and implementation of their own programmes and thus ensure ownership. There is also no framework or sometimes even capacity for reporting results and evaluations of often ambitious development strategies. These factors undermine both RO performance and African ownership. The weaknesses of the AU and the RECs are also generally considered to reflect weak domestic capacity and political commitment to ROs in many member states (Sida, 2013b, p.4).
Despite the problems identified by its own staff and in critical reports, Sweden officially maintains that the regional strategy 'continues to be relevant in relation to African needs and priorities' and that it is on track for achieving its objectives (Sida, 2013a, p.13, 20;cf. Sida, 2012;2013b;2015a). Sweden also claims that its development cooperation is in line with African priorities and that it enables the AU and the RECs to play a leading role in achieving African integration (Sida, 2013b, p.2). The mid-term review of the Swedish strategy proposed that it should continue as it had during the first phase (2010-2012), albeit with greater emphasis on 'indirect support' (Sida, 2013a, p.6). Similarly, the Ministry for Foreign Affairs (2016-2020) stressed the continued relevance of the AU and the RECs (Swedish Ministry for Foreign Affairs, 2015). Overall, the Swedish position is that: The AUC and RECs continue to be weak, and capacity building is a slow and long-term process, but continued support to RECs is important for longer-term results of regional integration. These organizations are slowly improving, and they are the only ones with the political legitimacy to pull together the political decisions that are needed for improved regional political and economic integration. (Sida, 2013a, p.7) This attitude reveals that Sweden intends to continue on the same path even if positive results are not easily demonstrated. However, it is questionable whether capacity building is the most effective means of achieving ownership. In interviews, some Swedish officials argue that ownership may in fact be a precondition for effective capacity building. Although the AU and many RECs have achieved some results in the last decade, it may nonetheless be appropriate to question the Swedish assumption that capacity building is necessarily positive. There is clearly considerable uncertainty about the circumstances under which capacity building genuinely enhances the performance and ownership of African organizations. The crucial issues in this regard are how to deal with dysfunctional ROs and the problem of donor interference.

Dysfunctional ROs
Many African ROsincluding the AU and the RECsdo not function as portrayed in official documents. There is growing evidence that many show a wide discrepancy between agreed policies and their implementation. This is often referred to as the implementation gap. Studies also show that top-down ROs may be manipulated for narrow regime interests at the expense of broader collective interests (Bach, 2016;Söderbaum, 2016).
Ever since Sweden began supporting African ROs, Swedish aid officials have tended to ignore the dysfunctional organizations. However, discussion about this problem is beginning in the Swedish policy community. Sida recognizes that capacity building of the AU and the RECs has not yielded the anticipated results and Swedish donor officials based both in field offices and Sweden observe that supporting dysfunctional ROs may not only be ineffective but actually counterproductive. Some note that a problem they face in designing interventions is deciding whether or not a particular RO is effective. 5 There is disagreement within Swedish policy circles about how to respond when an RO proves to be dysfunctional. While the official Swedish position is that greater capacity building will encourage African ownership, this view is contested. Some Swedish officials claim that a degree of ownership is a precondition for capacity building to be effective. 6 Indeed, in some circumstances, capacity building may undermine ownership and an RO's performance. This is consistent with new knowledge in this field. A set of studies commissioned by the Swedish Embassy in Kenya and produced by the European Centre for Development Policy Management (ECDPM) point out that ROs must be understood in the context of the region's political economy . ECDPM claims that this requires greater knowledge about the politicized and often contested nature of regional cooperation and integration: Elite-defined national interests influence whether, when and how effectively countries engage on the regional agenda. This needs understood rather than assumed by regional policy-makers and donors. This implies a need to explicitly identify national and elite interests and their potential to undermine or support the role of regional organisations, and indeed how to adapt regional projects to contribute to attaining key political objectives. Identified interests in a specific policy area may suggest the need to await better circumstances or to avoid specific regional players or organisations for progress to be made. (Vanheukelom et al., 2016, p.20) Better understanding of context will enable policy makers to appreciate the relationship between capacity building and ownership. The ECDPM studies also underline the problems that arise when African ROs become overreliant on foreign funders.

Donor involvement and donor 'control'
It is often believed that donors must be involved in efforts to enhance the institutional capacity of ROs (DFID, 2010;UNDP, 2013). A Sida evaluation from 2015 claims that '[i]n a region characterized by weak institutional capacity and political risks, Sida should have had a greater presence in governing and monitoring' (Sida, 2015b, p.34). While there is no solid evidence that greater donor involvement necessarily reduces ownership, it does pose challenges.
The growing literature on the foreign funding of ROs shows that donors are frequently driven by their own domestic political and economic interests in controlling other states rather than by the desire to achieve the maximum development impact and local ownership (Gray, 2013; also see Vanheukelom, 2016, p.11). Although Sweden is often believed to be less intrusive than other donors, there are still tensions. For instance, the EAC has received considerable support from Sweden (and many other donors) in the last two decades. However, the EAC Secretariat has sometimes rejected support to prevent 'excessive donor interference'. Sida reports that this is 'complicating the overall cooperation' (Sida, 2013a, p.3). The problems relating to donor involvement are not simply about the self-interests of donors and apply also to donors with good intentions.
There is a variety of funding strategies and aid modalities (see Brolin, 2017). Most notable among them are project funding, programme funding, basket funding and core funding/budget support. The choice of aid modality has implications for aid effectiveness and ownership. This is frequently discussed by those involved in country-based development cooperation. Here it is well established that core and programme-based support is assumed to give recipients greater discretion and autonomy, and that this will usually lead to more ownership and sustainable development cooperation. By contrast, project-based support is generally given for shorter terms and is more tightly controlled by the donors. Somewhat strangely, aid modalities are rarely discussed as far as regional development cooperation is concerned. The Swedish regional strategy provides few guidelines about aid modalities apart from a brief statement that the share of programme-based support should be increased (Government Offices of Sweden, 2010a). Similarly, Swedish donor officials were unable to explain why any particular modality should be preferred in the case of regional development cooperation to ROs in Africa. It deserves mentioning that Sweden is more explicit on the need for programme-based approaches in its development cooperation with other regions. For instance, in its assistance to the Middle East and North Africa Sweden is explicitly expected to give 70 per cent of its support to programmes (Government Offices of Sweden, 2010b, p.2). Similarly, in the case of Asia, core and programme support is prioritized over projects. This is in accordance with the aid effectiveness principles of the Paris Declaration (Government Offices of Sweden, 2010c, pp.1-2).
Even if a large proportion of Swedish development cooperation consists of project aid, some two-thirds of total Swedish support is direct or indirect support to the AU and the RECs and other ROs (including collaborating and equivalent organizations). Compared to most other donors, Sweden works through the different ROs, the AU and the RECs in particular. However, Sweden also supports initiatives such as TradeMark East Africa (TMEA), which is a large donor-created, non-profit company supporting the East African integration process (including EAC itself). TMEA provides speedy construction of, for instance, roads and border posts. 7 Even if the EAC countries are reportedly enthusiastic about TMEA, there is a growing tendency within the donor community to channel funding outside the formal structures of ROs and the RECs. For Sweden as well as other donors, TMEA has been a way to bypass the EAC and circumvent the problems with direct support to this organization. 8 All of this raises questions about the choice of aid modalities and respect for home-grown African initiatives, African ownership and sustainability.
Comparing EAC and TMEA shows that there are trade-offs that must be made when selecting aid modality and these have consequences for ownership. Project support is not necessarily the worse option. Given problems faced by many RECs, including the EAC, many donors (and some recipients) tend to favour the quick and often efficient implementation offered by donor-created ventures such as TMEA. Interestingly, the non-bureaucratic implementation of such initiatives provides a 'strong country voice', which in turn enhances African commitment to the regional integration agenda (IEG, 2007, p.xvii). This is significant since 'project aid' contravenes the core principles of the Paris Agenda on Aid Effectiveness. More research is therefore needed to understand the conditions when project and donor-driven solutions may enhance African ownership. It is obvious that the politics around the EAC has become problematic. However, the solution may not necessarily be to replace direct aid to EAC with donor-controlled mechanisms either.

Conclusion
This article has focused on the link between development cooperation and the ownership of ROs. This issue has hitherto been largely ignored by researchers. Sweden was considered a particularly relevant case study for several reasons. Firstly, Sweden is one of the most important funders and supporters of ROs in Africa. Secondly, Sweden's policy for development cooperation is that it should be guided by the ownership principle. Thirdly, Sweden is often considered a role model for effective development cooperation.
The analysis shows that Swedish regional development cooperation with Africa has many advantages. It is designed to assist African actors in achieving their own development and integration agenda by strengthening the capacity of ROs and other African actors. Sweden works largely through and with African actors and recipients. These factors are all helpful for strengthening performance and African ownership.
However, two main weaknesses are apparent. Firstly, there is a confusion in Swedish policy circles about what ownership actually means and about who owns what. Swedish development cooperation focuses primarily on the secretariats and representatives of the AU, RECs and other ROs while paying less attention to national governments and non-state actors despite the fact that ownership of ROs is in fact shared between secretariats and their member states. Most observers go so far as to argue that ROs are owned first and foremost by national governments rather than by their secretariats.
Secondly, there is a lack of clarity about 'how to' effectively build ownership. Swedish policy makers tend to prioritize capacity building, particularly of RO secretariats, in the hope that this will enable African actors to design and implement their 'own' policies. However, there is little evidence of sustainable positive results. The relationship between capacity building and ownership is complex and is not mono-directional. Swedish policy makers need to acknowledge this and understand that a degree of ownership may be a prerequisite for effective capacity building.
The links between regional and country-based development cooperation are crucial to more effective development cooperation with ROs. This link has long been a topic of discussion among donors and some, such as the EU, have made progress. However, the disjuncture between regional and country-based development cooperation is particularly evident in the Swedish case. The Swedish 'top-down' and RO-centred approach reinforces the gap between national and regional processes, and this undermines African ownership and commitment in the ROs. This is in contrast to the 'bottomup' approach favoured by a growing number of other donors that promote regional development through national strategies and programmes rather than vice versa. The bottom-up approach appears to be more effective in strengthening national commitment to ROs and, consequently, also (national) ownership. Swedish policy makers may therefore learn from other donors and agencies, such as the EU, the AfDB and the WB, that have managed to harmonize national and regional foreign support programmes.
This article also reveals the importance of problematizing the way donors and ROs actually function. Donors are not simply neutral financiers who simply want to 'help' African actors achieve integration and build their own 'home-grown' ROs. Donors are an integral part of weak ownership and their modus operandi needs to be reconsidered. Similarly, ROs do not always function as we might wish or as stipulated in official treaties. There are many dysfunctional ROs in Africa and this presents donors with the question of how to respond. There are several reasons why African member countries may want to avoid the costs of assuming responsibility for designing and implementing their own regional programmes and policies. There are many cases in which African actors do not see the ROs as 'theirs' but rather as donor-controlled instruments or as part of the aid industry. There are also many dysfunctional ROsin Africa just like in other parts of the world (Söderbaum, 2016). Obviously, it is then difficult to build ownership. This is why research on regionalism is so important for providing insights both into and for policy making. It is valuable to critically examine the situation on the ground in order to find explanations for the implementation gap and the dysfunctionality of some ROs. Many donors, including Swedish officials, could benefit from considering different explanations and theoretical perspectives so as to gain a more nuanced understanding of how to effectively contribute to the African integration agenda and strengthen ownership of African ROs.