A Decade Since Busan: Towards Legitimacy or a ‘New Tyranny’ of Global Development Partnership?

Abstract

In 2011, the Busan Forum on Development Effectiveness announced the creation of the Global Partnership for Effective Development Cooperation (GPEDC). The GPEDC promised to replace the old donor–recipient dynamic in international development with ‘an equator-less landscape of multistakeholder global partnership’ (Eyben & Savage, 2013). Multistakeholder partnerships (MSPs) have since become the global governance modality of choice for the UN Sustainable Development Goals. Advocates maintain that MSPs provide more legitimate and effective governance than interstate multilateralism. But critics contend that MSPs co-opt subaltern voices and reinscribe power asymmetries: MSPs may, thus, constitute a ‘New Tyranny’ within global development governance (Cooke & Kothari, 2001). Drawing on extensive empirical research, this article examines the extent to which the GPEDC provides more legitimate governance or effects akin to a ‘New Tyranny’. Against dichotomous portrayals of MSPs as either inherently ‘legitimate’ or ‘tyrannical’, the article finds that MSPs such as the GPEDC constitute spaces wherein normative principles of world order are negotiated. Nevertheless, the article finds that we are no closer to a global development partnership than we were at Busan.

Keywords:

• Foreign aid
• GPEDC
• new tyranny
• legitimacy
• multistakeholder
• global governance

1. Introduction

In 2011, the international development community met in Busan, South Korea. It was the OECD’s final Aid Effectiveness forum: a reform process that addressed two critical issues. First, the perceived ineffectiveness of aid. Second, the illegitimacy of international development governance: a system long dominated by Northern donors but was no longer tenable with the emergence of ‘new’ state and non-state actors. Busan constituted a shift from ‘Aid’ to a broader ‘Development Effectiveness’ paradigm, and it culminated in the creation of a new Global Partnership for Effective Development Cooperation (GPEDC) (Mawdsley, Savage, & Kim, 2014). The GPEDC implied a transition from the OECD-Development Assistance Committee (DAC) or Northern donor system (Eyben & Savage, 2013). It is instead a multistakeholder partnership (MSP): a governance modality wherein all stakeholders – Northern and Southern state and non-state actors – come together in equal partnership. MSPs are now the governance modality of choice for key agendas such as the Sustainable Development Goals (SDGs), while the GPEDC is perhaps the most ambitious – yet overlooked – multistakeholder development partnership (Schaaf, 2015, p. 69).

The proliferation of MSPs is among the most significant trends in the recent history of global development governance (Reinsberg & Westerwinter, 2021). But the recourse to MSPs is not unique to development. MSPs have proliferated across all global policy areas; MSPs have been described as a major alternative (and challenge) to interstate multilateralism (Gleckman, 2018). MSP’s associated values – corporate partnership, voluntarism, and informal governance – arguably constitute the defining characteristics of twenty-first century global governance (Buxton, 2019, p. 12). For advocates, MSPs offer ‘solutions’ to gridlocked multilateral governance, and ‘win-win situation[s] where effectiveness and the democratic credentials of global governance both increase’ (Hale & Held, 2012; Pattberg & Wilderberg, 2014, p. 18). But MSPs have their critics. Some maintain these ‘partnerships’ risk the co-option of subaltern or dissenting voices, the ‘privatisation of global governance’, and ‘imperialism by other means’ (Hill, 2014; McKeon, 2017; Tiwana, 2014).

To speak in a language familiar to development scholars, MSPs may reinscribe power hierarchies, thus, potentially constituting a ‘New Tyranny’ in global development governance (Cooke & Kothari, 2001). As Cooke and Kothari (2001) argued with regards to the paradigm of ‘Participatory Development’, there was a discrepancy between the laudable rhetoric – that offers emancipation and equitable outcomes – and the reality of these approaches. The latter, they contend, resulted in co-option, domination, and the replication of unequal power relations. Similarly, whether MSPs offer more legitimate governance – or effects akin to a ‘New Tyranny’ – is at the forefront of debate on these widely proliferating governance modalities.

This article, thus, asks whether the GPEDC – one of the most ambitious attempts at multistakeholder global development governance – constitutes either a ‘New Tyranny’ or a legitimate global partnership. To put this another way, the article asks whether the same hierarchical power dynamics characteristic of the ‘old’ international development governance manifest in ostensibly more equal and innovative governance forms.

This article contributes an empirical update to Eyben and Savage (2013) account of the shifting politics of international development as reflected within the 2011 Busan Forum. This article sheds light on the complex dynamics and diverse perspectives on the resultant GPEDC over the last decade¹. The GPEDC is a critical meeting point between both the ‘old’ and ‘new’ public, civic, and private actors that now constitute the field of global development cooperation. The GPEDC, thus, offers an unrivalled vantage point to explore perceptions of, tensions, and power dynamics within global development governance in the twenty-first century. To do so, this article draws on findings from 60 interviews with representatives who have engaged with the GPEDC over the past decade². This includes Northern and ‘South-South Cooperation’ (SSC) providers, recipients, Civil Society Organisations (CSOs), private sector actors, and UN and OECD staff. Findings were augmented with six months of ‘insider’ observations that I made as a trainee member of the GPEDC secretariat. Here, I was responsible for communications, liaising with stakeholders, secretariat support, and preparations for the 2019 Senior Level Meeting at the UN. These research activities uncovered how various ‘stakeholders’ understood the GPEDC’s legitimacy, and similar global governance mechanisms.

While the primary contribution of this article is empirical, the article also contributes a novel theoretical framework for assessing the legitimacy of Global Governance Institutions (GGIs) that combines both normative and sociological approaches. This approach builds on normative frameworks that have been used to appraise the legitimacy of GGIs, including development organisations (Kindornay & Samy, 2012; Schmidt, 2013; Vershaeve & Orbie, 2016). This combined normative-sociological framework is unique in that it engages with how different stakeholders evaluate the normative-democratic qualities of a given organisation, such as the appropriate level of inclusivity, representativeness, accountability, and so on.

I contend that we are no closer to a global development partnership than we were at Busan. Yet simply maximising legitimacy indicators – such as maximising the level of inclusiveness – does and will not provide for more widespread legitimacy perceptions. Rather, what matters is the interpretation of these values, and these differ across actors. The GPEDC case reveals there are intractable, conflicting perspectives over what legitimate governance demands. Against dichotomous portrayals of MSPs as either inherently ‘legitimate’ or ‘tyrannical’ processes, this article finds that MSPs such as the GPEDC constitute venues wherein normative principles of world order are negotiated: the outcome of these negotiations is contingent upon the prevailing balance of social forces, and their respective interests and strategies of engagement. The contingent nature of these governance modalities does not foreclose possibilities for more transformative governance. But in the context of Agenda 2030 – where MSPs are positioned as the sine qua non governance modalities for the attainment of the SDGs – I argue for more circumscribed expectations over what MSPs can achieve in practice.

This article’s remaining structure is fivefold. The next section briefly recounts the history of the ‘Aid Effectiveness’ process. The second outlines the theoretical framework, the third presents the findings, and the fourth discusses these. The final section concludes by outlining several policy recommendations for global development governance.

2. From ‘Aid Effectiveness’ to the GPEDC

International development governance has long been characterised by a hierarchical North–South binary (Six, 2009). Within international development organisations, it has been Northern donors who have determined the definitions, means, and ends of ‘development’. Another characteristic of international development governance has been its exclusively state-state nature. Although non-state actors could lobby for public policy goals, international governance was the preserve of nation-states. At Busan, these two longstanding characteristics were called into question (Eyben & Savage, 2013).

The ambition of the Aid Effectiveness agenda was ‘unparalleled in the history of aid’ (Ashoff & Klingebiel, 2014, p. 1). This DAC-led process was most visible through successive high-level fora (HLF) in Rome (2003), Paris (2005), Accra (2008), and Busan (2011). These fora ‘addressed the fundamental issue of why aid was not producing the outcomes intended and resulted in the GPEDC that is attempting to reshape the development landscape’ (Schaaf, 2015, p. 69). With each HLF, we see a broadening of the actors involved in international development governance. The Paris HLF ‘define[d] common principles and procedural rules for a large number of donors and partner countries, thereby significantly expanding the international regulatory framework for aid’ (Ashoff & Klingebiel, 2014, p. 7). Recipients and donors ‘both agreed to time-bound commitments and to hold each other to account’ (OECD, 2017). In Accra 2008, this framework was broadened to include civic and private sector actors, alongside parliamentarians, foundations, and local governments.

While the Aid Effectiveness process made strides towards inclusive governance, the DAC was unsuccessful in its outreach to Southern providers, namely China, India, and Brazil. These were not ‘new’ providers: they have engaged in development cooperation since the 1950s. But their influence and scale of activities were novel, and they aroused suspicion and intrigue within the aid establishment (Manning, 2006; Naím, 2007). In their affirmation that SSC was a distinct paradigm – rooted in post-colonial solidarity, a rejection of donor–recipient relations, non-Official Development Assistance (ODA or foreign aid) finance, and a shared developing country identity (Mawdsley, 2014) – these countries snubbed the Aid Effectiveness process, an agenda they felt was inapplicable to them as providers. They instead affirmed that the UN Development Cooperation Forum was the legitimate global development forum. Its broader mandate automatically encompasses ODA and non-ODA modalities, thus, introducing competitive pressure against the DAC’s Aid Effectiveness process (Vershaeve & Orbie, 2016). Securing the participation of these recalcitrant SSC providers was paramount for the DAC at Busan.

Participation at Busan was unprecedented. Over 3000 delegates from Northern and Southern states, CSOs, the private sector, and other non-state actors convened to forge a ‘new consensus’ (Abdel-Malek, 2015, p. 788). The forum’s goals were threefold: to review implementation progress of the Paris Aid Effectiveness principles³; to move towards a more expansive paradigm of ‘Development Effectiveness’, and; to ensure the new agenda was more inclusive (Kim & Lee, 2013, p. 788). On implementation, monitoring surveys revealed that although recipients had made significant progress, only 1 of 12 commitments had been fully met (Wood, et al., 2011). There was, therefore, a widespread perception that previous fora had failed to generate donor behaviour change. Areas where progress was weakest – the predictability of donor aid, and donor use of recipient country systems – were indicative of the persistent donor–recipient power imbalance (Chandy & Kharas, 2011, p. 745). At Busan, a ‘Focus on Results’ was a key theme, alongside efforts to ‘broaden the tent’ of development governance (Kindornay & Samy, 2012, pp. 5–7). Busan culminated in the Development Effectiveness principles: country ownership, inclusive partnerships, transparency and accountability, and a focus on results. These principles applied to all forms of development cooperation (beyond ODA), and all actors irrespective of their North/South, state/non-state designations.

For many commentators, Busan constituted a ‘paradigm shift’ in international development governance and discourse (Atwood, 2011). Development Effectiveness goes ‘beyond aid’: ODA merely serves a complementary and catalytic role alongside a suite of other public and private financial instruments. The ‘new’ consensus also solidified a shift away from poverty reduction and ‘rights-based’ approaches, as was development orthodoxy in the 1990s and 2000s, towards a greater emphasis on the private sector and economic growth as the means and ends of ‘development’ (Mawdsley et al., 2014).

Yet Development Effectiveness meant different things to different actors: it was ‘subject to so many possible interpretations that everyone could agree to the term, if not to its content’ (Eyben, 2012, p. 79). For donors, there were contestations over whether ‘Development Effectiveness’ was an auxiliary component – or a replacement – of the ‘unfinished’ Aid Effectiveness agenda (Esteves & Assunção, 2014, p. 1785). For CSOs, the notion retained a focus on human rights, while they presented concern over the elevated role ascribed to private actors (BetterAid, 2010). But despite this exalted role, to many it was unclear what the precise role of the private sector was in this ‘new’ agenda. At Busan, the private sector was approached as a ‘monolithic bloc’, with little attendance to sectoral distinctions and types of actors (Kharas, 2011, pp. 7–8). Moreover, getting the large SSC providers to sign the Busan Agreement was difficult. They signed only with the last-minute inclusion of Paragraph 2: ‘the nature, modalities, and responsibilities that apply to SSC differ from those that apply to North-South cooperation… principles, commitments, and actions agreed shall be the reference for South-South partners on a voluntary basis’ (BOD, 2011). The shared principles yet ‘voluntary and differential responsibilities’ proviso was regressive compared to Paris’ commitments-based approach to accountability. This new accountability approach risked becoming ‘a lowest common denominator [approach] without bite or focus by becoming more general and inclusive’ (Kharas, 2011, p. 7).

Busan nonetheless culminated in the creation of a new global partnership that sought to encompass both ‘old’ and ‘new’ actors. The GPEDC was to herald an era wherein ‘the old donor–recipient relationship is replaced by an equator-less landscape of multistakeholder global partnership’ (Eyben & Savage, 2013, p. 457). Some question whether the GPEDC’s creation marked ‘the end of the traditional OECD dominance in development financing’, while others suggest it might ‘dismantle the dominance of the DAC in the same way that G20 did to the G7/8’ (John, 2012, p. 17; Li & Wang, 2014, p. 2). But the GPEDC has been widely criticised. Some maintain it is 'business as usual', and the GPEDC perpetuates pre-existing power disparities of the DAC-dominated development era (Abdenur & Fonseca, 2013; Xiaoyun et al., 2018).

Besharati (2013, p. 22) contends that the GPEDC’s success will be dependent on ‘the extent to which stakeholders see the governing mechanism as legitimate in terms of its inclusivity and representativeness (input legitimacy), quality of decision-making processes (throughput legitimacy) and effectiveness in achieving outcomes (output legitimacy)’ (see Kindornay & Samy, 2012). Yet research into the internal dynamics and perspectives on the GPEDC is sparse⁴. This article now outlines a framework that can assist in exploring stakeholder perceptions on the legitimacy of this governance modality.

3. Framework

Legitimacy can be understood as a ‘generalised perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions’ (Suchman, 1995, p. 574). Yet the question of which legitimacy conception to adopt is one that ‘fundamentally divides sociology and political theory’ (Thornhill, 2011, p. 135). The choice is often presented as a binary: between ‘normative’ or ‘sociological’ approaches. Normative approaches develop an a priori set of well-reasoned standards, overwhelmingly derived from democratic thought, to evaluate an institution (Buchanan & Keohane, 2006). These democratic standards or values – such as inclusive or accountable governance – are understood as valuable in themselves towards assuaging the ‘democratic deficit’ at the global level (Moravcsik, 2004). But they also have instrumental importance insofar as their attainment can provide more effective global governance. Inversely, sociological approaches are guided by inquiries such as ‘to what extent are institutions accepted by the people, and what explains variation in their perceived legitimacy?’ (Dellmuth & Tallberg, 2015, p. 454).

Exclusive adherence to either approach has significant shortcomings. Normative-democratic approaches tend to treat the ‘problem’ of legitimate governance as self-evident: a lack of democratic procedures. Yet Peters (2013, p. 11) contends that simply evoking ‘democracy does not do much to clarify the [empirical] issues that global governance faces regarding legitimacy’. Democracy is not the sole legitimising force within global politics. Inversely, sociological approaches overcome this shortcoming by focusing on what different stakeholders understand as legitimate governance. But this approach also has its limitations. It is possible that ‘sociologically legitimate institutions can be normatively illegitimate’ (Westergren, 2016, p. 49). A stakeholder may regard an institution as legitimate simply because it serves their self-interest. Furthermore, this approach may yield an ever-expanding list of possible sources of sociological legitimacy. But this is not conducive towards informing reform trajectories in more progressive directions, nor does it encourage dialogue on realising worthwhile democratic values within key institutions. Moreover, completely abandoning normative-democratic criteria when exploring perceptions would be to ‘throw the baby out with the bathwater’. There may be other sources of (sociological) legitimacy that bear little relation to democratic thought, but ‘substantive publics now challenge the legitimacy of global regimes for their lack of democracy’ (Scholte, 2011, p. 115). Moreover, global institutions devote considerable resources towards augmenting and demonstrating their ‘democratic credentials’.

If the ambition is to adopt an approach that can engage with what stakeholders understand as legitimate governance, the tensions that exist between stakeholders, while exploring the realisation of democratic values at the global level, partisan adherence to either approach is neither necessary nor appropriate. I offer a combined framework wherein normative-democratic values can be used as heuristics to structure sociological research into perspectives. The approach is unique. Rather than adopting a normative or a sociological approach to legitimacy – or relegating sociological legitimacy to a single indicator (see Take, 2012) – the proposed framework investigates the normative qualities associated with input, throughput, and output legitimacy through the perspectives of stakeholders themselves.

We can also adopt well-worn heuristics, such as Scharpf’s (2001) ‘input’ and ‘output’ dimensions of legitimacy, and ‘throughput’ legitimacy as an intermediary dimension (Schmidt, 2013). The literature generally assumes a positive relationship between these dimensions: the more an institution satisfies each dimension, the stronger the overall legitimacy (Hatanaka & Konefal, 2012, p. 156). However, there may also be tensions between these dimensions, such as between having a highly inclusive (input) institution, and that same institution being capable of producing output legitimacy or effective governance (Lindgren & Persson, 2010) (Table 1).

Table 1. Framework for stakeholder assessment of multistakeholder partnerships

Dimension	Value	Inquiry
Input	Inclusivity	How do stakeholders evaluate the inclusivity of, and regard as legitimate participants within, the MSP?
	Representativeness	Are representative categories and structures deemed appropriate?
	Equality	Do stakeholders consider there to be both de facto and de jure equality within the partnership?
Throughput	Deliberative quality	Does decision making and debate proceed in a universal and rational manner?
	Transparency	Is information produced by the MSP considered both accessible and understandable by affected stakeholders?
	Accountability	How do stakeholders evaluate the monitoring mechanisms in place?
Output	Outcome	Do stakeholders regard the MSP as an effective modality towards encouraging behaviour change?

3.1. Input legitimacy

Input legitimacy is the oldest, most rigorously discussed legitimacy dimension. Linked to Republican ‘direct democracy’ ideals, input emphasises ‘government by the people’ (Bekkers & Edwards, 2016, p. 43). Input is concerned with participation: who ought to make decisions, and who ought to be represented within governance (Bäckstrand, 2006, p. 292; Scharpf, 1999). We can assess input legitimacy by asking stakeholders how they perceive the inclusivity, representativeness, and equality-based qualities of an institution.

Inclusivity is a ‘core element of any democratic theory’ (Dingwerth, 2007, p. 27). It is hypothesised that the more an institution is designed towards maximal possible inclusion of stakeholders, this will enhance the input legitimacy of a governance arrangement (Dahl, 1989). This concern holds normative weight in ensuring all those affected by a governance arrangement have a say in decision-making (Scharpf, 1999). It also holds instrumental value in that including all relevant stakeholders may foster a sense of shared ownership and, consequently, greater compliance (see output) (Bäckstrand, 2006, p. 295). Key questions to ask are whether stakeholders view a given GGI as appropriately inclusive, and who they regard as legitimate participants.

Representativeness is tightly linked to inclusivity. While the latter pertains to a general principle that runs through governance procedures, representativeness explores stakeholder views on the formal procedures that guarantee participation (Bäckstrand, 2006). Having all of those who are affected by a GGI present within it is impossible. In the absence of direct participation, structures can be put in place so representatives can ‘act for’ and ‘make present’ those who are not (Pitkin, 1967). This indicator requires reflection on whether the representative structures in place are regarded as ‘fit for purpose’.

Inclusivity and representativeness are not achieved when only those with the capacity to participate can do so (Guastafero & Moschella, 2012, p. 204). It is also necessary to examine whether asymmetric power capacities undermine equality while contributing to the ‘systemic underrepresentation’ of groups (Take, 2012, p. 224). For instance, the disproportionate capacity of powerful Northern state and private actors generates claims that they are over-represented within MSPs, while smaller state and non-state actors are underrepresented (McKeon, 2017; Pattberg, 2010). Here we can ask stakeholders the extent to which they perceive as equality – both de facto and de jure – among the participants of an institution.

3.2. Throughput legitimacy

Missing from most legitimacy assessments ‘is what goes on in the ‘blackbox’ of governance between input and output’ (Schmidt, 2013, p. 6). Proponents of throughput legitimacy are advocates of Habermas (1990, p. 185) ‘deliberative democracy’. They are concerned with how a governance process fares against the notion of an ‘ideal discourse’, and thus, the procedural quality of debate and decision-making procedures. Three indicators are useful: deliberative quality; transparency; and accountability.

Deliberative quality is a particularly useful throughput indicator in situations where conventional ‘democratic representation and/or voting mechanisms are not available options’, such as MSPs (Risse, 2004, p. 110). The ‘focus on deliberation helps to prevent relevant [stakeholder] information from being excluded… [and] the dialogue process from generating a… [limited] perception of the problem[s]’ (Kerkhof, 2006, p. 282). We can, thus, inquire into whether stakeholders perceive that decision making and debate proceeds in a universal and rational manner.

Meeting transparency implies stakeholders have ‘unlimited and timely access’ to information regarding decision-making and institutional processes (Keohane, 2011, p. 102). Hahn and Weidtmann (2016, p. 109) note that missing translations of documents can be a ‘potential impediment to democratic legitimacy’; it can result in the exclusion of ‘those experts who lack sufficient knowledge of English from participating effectively in direct communication’. It is also important that stakeholders are ‘able to do something with [the] information – that they understand it, evaluate it in their own interests and act upon it if necessary’ (Dingwerth, 2014, p. 1132). We can, thus, ask stakeholders not only how they evaluate their access to information, but also how they understand the accessibility of information.

Accountability refers to principal–agent relationships between actors (Bäckstrand, 2006, p. 294). Democratic accountability is a system whereby principals that are significantly affected by the actions of an agent can make demands on the latter to report on their behaviour. However, accountability relationships within MSPs are wrought with challenges: MSPs ‘are diffuse, complex weakly institutionalised collaborative systems that are neither directly accountable to an electoral base nor do they exhibit clear principal–agent relationship[s]’ (Benner et al., 2004, p. 3). Towards accountability, ‘it needs to be clear who is accountable for what and to whom, and mechanisms need to be in place to provide clear review procedures for monitoring’ (Dann & Sattelberger, 2015, p. 68). Exploring perceptions on monitoring frameworks is one pragmatic means to explore this indicator.

3.3. Output legitimacy

Output legitimacy points to ‘effective’ problem solving, or, that arrangements produce ‘government for the people’ (Scharpf, 1998, p. 11). Put simply, output legitimacy results from a stakeholder’s perception that an institution performs well in providing effective governance. Moreover, it is well established – within domestic politics – that high input and throughput legitimacy can make political institutions more effective, and this may also hold for GGIs (Sommerer & Agnè, 2018). Output legitimacy – or effectiveness – is, thus, both a cause and consequence of legitimate governance processes.

Wolf (2010) distinguishes between three indicators of output legitimacy: policy outputs as governance products (e.g. policy reports, declarations, etc); outcomes as behavioural changes that result from a stakeholder’s engagement within a process, and; impact as clear evidence that a governance modality has contributed towards overall goal attainment, e.g. poverty reduction. There are various approaches to disaggregating and assessing policy outputs. Tallberg et al. (2016), for instance, present a fivefold typology of sub-indicators to comparatively assess the policy outputs of GGIs. While this approach may be suitable for macro comparisons of GGIs across space, fields, and time, mapping perceptions on five subindicators would prove highly complicated and unnecessary for this study. Doing so would not only require that stakeholders have full knowledge of a given GGI's policy outputs. It would also require that stakeholders hold explicit assessments of these sub-indicators. Impact, on the other hand, requires judgments about the extent an institution relieves the problem it was tasked to resolve: an extremely complex measure that ‘involves comparison with a counterfactual state of affairs without the regime in place’ (Schmelzle, 2012, p. 2).

Outcome is the ‘ability of an institutional arrangement to generate acceptance and to motivate actors to comply with rules’ (Take, 2008, p. 3). Given the difficulty in studying perspectives on policy outputs and impact, outcome ‘provide[s] a valuable proxy for [overall] effectiveness’ or an institution’s output legitimacy (Mitchell, 1996, p. 23). As McEwen and Maiman (1983, p. 258) contend, ‘voluntary compliance [with an institution’s rules, norms or recommendations] is the fundamental observable indicator of legitimacy’. As an intermediary indicator, outcome can be used to examine the extent to explore the extent to which stakeholders regard a governance mechanism as effective in encouraging behaviour change, be it through enforcement mechanisms or successive monitoring reports⁵. Outcome also enables an examination into stakeholder justifications as to why they have – and have not – undertaken behaviour change. I now apply this framework to the GPEDC.

4. Perspectives on the GPEDC’s legitimacy

The GPEDC has an inclusive, multistakeholder governance structure comprised of three bodies: a 25-member Steering Committee⁶; a Secretariat based across the OECD and UNDP, and; four Co-Chairs⁷. The GPEDC is funded through voluntary contributions, mostly from donor countries and the incumbent donor co-chair (see GPEDC, 2018). The GPEDC boasts endorsement from 161 countries and 56 international organisations. It convened 1500 senior delegates at the 2014 Mexico High-Level Meeting (HLM1), 4600 delegates at Nairobi in 2016 (HLM2), and over 600 senior-level participants at its 2019 SLM. Its purpose is to advance progress on implementing the ‘Development Effectiveness’ principles, primarily through its biennial monitoring process, and through various working groups that advance particular agenda items (GPEDC, 2016). It does not have a formal role for SDG implementation, although it positions itself as the forum concerned with ‘how-to-do’ sustainable development (GPEDC, 2019). While some may regard the GPEDC as an MSP par excellence, it is a contested institution in terms of its perceived legitimacy.

4.1. Input legitimacy

4.1.1. Inclusivity

The GPEDC is purported to be a ‘uniquely inclusive global initiative in which non-state actors play a full role, alongside governments, in its governance’ (Bena & Tomlinson, 2017, p. 2). As a major DAC donor noted: ‘our aspirations for the global partnership was that we were kind of democratising the Aid Effectiveness discussion’⁸. Yet there is variation in perspectives over how inclusivity is understood and over who should be regarded as a legitimate participant.

Despite the aspiration for a global partnership, large SSC providers (China, India and Brazil) have not participated in the GPEDC⁹. These actors maintain the GPEDC is a ‘DAC-led process that could not be legitimated by actors outside DAC structures’ (Esteves & Assunção, 2014, p. 4; Li, et al., 2018). For DAC donors, the GPEDC is, thus, not ‘truly inclusive because it doesn’t have the SSC providers’¹⁰. Getting SSC providers involved in the agenda was of key importance to DAC donors at Busan. Hence, the absence of the former can – in part – account for declining donor interest in the partnership and agenda. Yet recipients do not regard this as an issue: ‘the absence of the BRICS in the GPEDC does not alter the legitimacy of the partnership’¹¹. For recipients, the GPEDC’s value lies in providing countries ‘which were rarely heard, a forum where their concerns and their priorities can be heard on such a global level’¹². Similarly, CSOs maintain SSC presence would not necessarily enhance the GPEDC’s legitimacy: ‘you can't call yourself a global partnership without having those players involved, but they would just weaken the accountability commitments of the partnership… well, good riddance then'¹³.

An additional line of contention concerns the position of non-state actors. CSOs argue the GPEDC’s inclusivity ‘set a precedent, one I think a lot of governments are uncomfortable with'¹⁴. The proposal to introduce a fourth non-state Co-Chair that would sit on par with the governmental co-chairs and would represent CSOs and other non-state actors has been met with consistent resistance by recipients¹⁵. Some recipients feel, for instance, that there are too many ‘CSOs in GPEDC meetings to the detriment of other [public] stakeholders’¹⁶. It was only after six years of sustained CSO canvassing that this position was implemented in 2019 at the behest of the German donor Co-Chair. Furthermore, CSOs express concern over how the private sector is included in the partnership. CSOs argue within the GPEDC and global development governance more broadly, ‘private sector engagement’ is a ‘euphemism for large corporations’¹⁷. With profit rather than ‘development’ as their primary directive, CSOs ‘strongly oppose why they [corporations] are there’¹⁸. Although a common and laudable rallying cry, tensions abound over what ‘inclusive governance’ should mean in practice: the ‘more-the-merrier’ principle is weak policy guidance (Weiss, 2014, p. 3).

4.1.2. Representativeness

The intention to represent all relevant development actors is reflected within the GPEDC’s constituency-based Steering Committee. But for most, this model is not working. Only those with pre-existing networks of representation and consultation in place – donors through the DAC and CSOs through their independent coordinating platform (CPDE) – regard themselves as well-represented. Despite the hope the GPEDC would better reflect the views of the beneficiaries of development cooperation (Carter, 2016), recipients are encumbered by fragmented and inadequate representative structures. Recipients could previously rely upon the representative Partner Country Caucus during the Aid Effectiveness era (2003–2011). Yet this caucus was disbanded upon the creation of the GPEDC due to ‘inadequate financial support [by donors] for such efforts’ (GPEDC, 2014, p. 3). SSC providers are also poorly represented. Despite being commonly subsumed under the banner of SSC, there are no pre-existing platforms suited for the collective representation of these stakeholders.

There are also huge difficulties in representing the ‘private sector’ as a global constituency. The issue is the ‘private sector is a hugely diverse and complicated non-state part of society’¹⁹ comprised of various sectors and organisations. As one representative for private foundations expressed:

It's difficult for me sitting on the Steering Committee to say, “I speak for foundations globally”, because I don't speak for foundations globally… I think that civil society would say the same thing… it's not always clear where voices are coming from. Is it just the voice of that constituency? Or is it an individual with an opinion?²⁰

As in Busan, there is a lack of clarity over the agenda’s relevance to the private sector, and the heavily procedural nature of multistakeholder discussions. Steering Committee meetings are subject to the ‘Development Problem’: ‘if you create a group you spend 90 per cent of the time talking about running the group, and 10 per cent on what you can do as a group’²¹. Hence, as a Northern business representative put it, ‘consultations were very, very limited… because what was there to be concerned with business? I don’t care if the next HLM is in Nairobi or Mogadishu… As long its not held during Davos week’²². These issues around representation are arguably intractable for governance at the global level: without a bounded community – or demos – in place, representative models are plagued with gaps and subject to criticism.

4.1.3. Equality

The GPEDC promised to be a ‘coalition of equals working towards shared goals’ (GPEDC, 2012, p. 2). Yet there was official recognition it risked being a ‘process driven by provider countries, and special efforts are needed to reach out to [recipient] partners’ (ibid.). For most, the GPEDC is an exemplary partnership of de jure equality for previously excluded state and non-state actors: ‘the GPEDC is one of the best examples where CSOs have a very well-defined role which is very close to that of governments on a purely formal basis’²³. Similarly, a Latin American SSC representative noted the GPEDC is ‘unique’ in that the Steering Committee offers ‘an equal footing of most development actors… it’s an exemplary multistakeholder process’²⁴.

But despite de jure equality, ‘what’s on paper differs from reality… ultimately one side has all the money and the other side doesn’t… even if you give everyone a vote that power dynamic will continue to exist’²⁵. Power hierarchies result from capacity differentials. This is recognised by DAC representatives: ‘the Global Partnership was meant to be a space where all actors could engage on an equal basis… and we try to mitigate against disparity in influence based on resources… But it’s the ‘Golden Rule’: the organisation with the gold makes the rule’²⁶. There, thus, continue to be stark de facto inequalities in the GPEDC’s governance. As a recipient representative noted ‘we are more passive… not because we don’t know what our challenges are, but we are so overwhelmed by the issues back home’²⁷.Those with the time and resources to participate tend to dominate, or as one CSO expressed, ‘it’s a partnership of equals, but it’s the most active equals’ who have the final say²⁸.

Furthermore, it is not only capacity disparities that undermine perceptions of de facto equality. There are also normative contestations over which actors should be regarded as equal participants. Some recipients argue non-state actors should not be on equal footing with governments, because ‘the governments have the lead, it is they who gives the money, and it is really the government that puts in action the programmes’²⁹. As noted by Wolf (2008, p. 237) on power relations in GGIs: ‘the shadow of hierarchy is still present, but the shade may vary’. Despite constituting an ‘exemplary’ de jure partnership of equals, the GPEDC’s de facto equality is encumbered by ‘a plethora of democracy-relevant structural inequalities that pervade global politics’ (Dingwerth, 2014, p. 1125).

4.2. Throughput legitimacy

4.2.1. Deliberative quality

In some respects, the deliberation enabled by the GPEDC’s informal nature is lauded by DAC donors: ‘the fact it is an MSP – so it’s not an intergovernmental or an official setting – it’s easier to engage in conversation, and discussions go beyond official positions’³⁰. Yet in the aspiration towards inclusivity, this inhibits the efficacy of governance, and depth in which substantive topics can be discussed. Speaking on major meetings, one DAC representative noted ‘…essentially there’s going to be a panel discussion in which you squish nine people into it, they get three minutes to say their little bit, and there is very little insight you can get out of that’³¹. For recipients, deliberation is ‘too global’, and this leads to ‘challenges in developing agreement or consensus’³². Consequently, these events are ‘largely back-patting’ performances, where ‘everything has been groomed well in advance’³³. The need to have all viewpoints present and in consensus, and the knock-on effects for expedient governance, has been a deterrent to private sector engagement: ‘international business organisations have made a strong effort to represent [business] interests in the Steering Committee… [however] shortcomings in efficiency and governance continue to fuel uncertainty and dissuade widespread business interest’ (BIAC, 2014, p. 1).

A tension concerns the procedural emphasis that should be placed on the GPEDC’s knowledge-sharing function. DAC providers are keen to transform the GPEDC’s primary function into a Davos-like knowledge-sharing forum (see 4.2.3). But recipients and CSOs maintain these efforts subvert ‘the main goal of establishing the GPEDC in the first place… to support and ensure accountability for the implementation of commitments at the political level’ (ibid.). These efforts to transform the GPEDC into an informal, knowledge-sharing forum are regarded by CSOs as ‘a political strategy of donor governments… If you’re attending, and you want to keep control, “brilliant, let’s have a technocratic discussion”. Because who can compete with you on a technocratic level? Nobody’³⁴. We, thus, see within the GPEDC attempts by powerful actors to ‘render technical’ governance spaces that are understood by CSOs and recipients to be fundamentally political (Li, 2011).

4.2.2. Transparency

All stakeholders regard the GPEDC as a highly transparent institution. There ‘might be more of a problem of information saturation [rather] than a lack of transparency’³⁵. The flipside, however, is that the amount of information provided to stakeholders is considered too burdensome. As one DAC donor noted, this ‘makes engagement more time consuming than it need to be… and that is a genuine disincentive to engage’³⁶. An additional contention for recipients, CSOs, and private sector actors concerns the accessibility of information. There are, as with many global initiatives, missing translations beyond English, French, and Spanish. But there are also issues over ensuring information is understandable to diverse public, private and civic actors: ‘the problem is that transparency, it's not just about giving out information… transparency is also making sure that the information is something different audiences can understand’³⁷. Transparency is a highly sought and praised normative goal, yet ‘realising transparency in a meaningful normative sense is indeed a far-fetched dream’ (Dingwerth, 2014, p. 1132). Here, we see how conflicting legitimacy demands pull the GPEDC ‘cart’ in different directions, between: those who demand information in more languages; those who want information to be accessible for a range of public and professional audiences; those who demand succinct and digestible information, and not least; a secretariat team with limited funds and capacity.

4.2.3. Accountability

The GPEDC has been described as a ‘major global level accountability mechanism’ within global development governance (Timo-Mahn, 2017, p. 22). As noted by former DAC Chair Brian Atwood, the GPEDC was created as a ‘way to hold governments’ feet to the fire’ through its biennial monitoring process (Abdel-Malek, 2015, p. iii). But there are significant tensions over the scope and nature of the GPEDC’s accountability mechanisms, and over who should be regarded as a ‘principal’ and ‘agent’.

Development cooperation providers – both SSC and DAC – have agitated against the GPEDC’s accountability function: they insist they are primarily accountable to their domestic constituencies. Bracho (2017, pp. 22–23) argues that although SSC actors initially signed onto the Busan Outcome Document, early attempts by DAC donors to enmesh SSC providers in the GPEDC’s monitoring process led to the perception of a ‘broken promise’. This, he argues, accounts for the subsequent lack of China, India, and Brazil’s engagement. DAC donors have since attempted to ‘downplay accountability in favour of less committal “learning exchanges”’ (Bena & Tomlinson, 2017, p. 3). Donors such as South Korea propose the GPEDC be transformed into a Davos-like platform for ‘various stakeholders to share development softwares and knowledge³⁸. The strategy here is that by downplaying the GPEDC’s accountability function, it might be possible to attain participation by recalcitrant SSC (and now DAC) providers.

In contrast to these negative perspectives on the GPEDC’s monitoring function, recipients and CSOs regard these accountability mechanisms as the ‘jewel in the crown’ of the process (Bhattacharya, 2017, p. 2). Both CSOs and recipients have collectively – and successfully – safeguarded the monitoring process as ‘the backbone and core product of the GPEDC’ (GPEDC, 2019). That said, recipients – and CSOs in particular – are concerned about the extent that providers have ‘filed the teeth’ of the GPEDC’s accountability bite. In attempting to secure the participation of both large SSC actors, and sustain the interest of DAC members, one CSO commentator noted that ‘accountability has been sacrificed at the altar of inclusiveness’ (Overseas Development Institute, 2016). Yet while recipients may regard the GPEDC as one of the few and highly valuable tools that they have for holding donors to account, they are not so keen on being held account by CSOs. CSOs also insist the private sector should also be held accountable through the GPEDC’s monitoring process, a suggestion private sector actors are keen to resist, while CSOs are not immune from criticism against their accountability credentials. Suffice to say, there are unresolved normative tensions over the accountability function of the GPEDC, and over who should be held accountable to whom.

4.3. Output legitimacy

In bringing all development actors together – and in pooling knowledge, resources, and expertise – the ambition was that the GPEDC would succeed over the ‘dismal results’ of the prior ‘Aid Effectiveness’ era (Abdel-Malek, 2015, p. 316). For most stakeholders, generating behaviour change through broad buy-in constitutes the GPEDC’s raison d’être³⁹. Stakeholders acknowledge that recipients – the focus of monitoring efforts – have undertaken significant behaviour change or reforms towards greater ‘Development Effectiveness’, and this is clear in successive monitoring reports⁴⁰. Moreover, we see significant increases in the number of recipient countries engaged in monitoring processes (from 46 in 2014, to 86 in 2018). Recipients ‘integrate [monitoring] into their national processes, they have their own incentive structures that trigger the whole process, and it doesn’t require any external pushing’⁴¹. While there are areas for improvement, particularly in ensuring inclusive partnerships and engagement with CSOs at national level, recipients have collectively undertaken significant behaviour change in line with the GPEDC’s recommendations.

But successive monitoring reports reveal ‘mixed progress’ towards implementing Development Effectiveness. Within the policy world, ‘mixed progress’ is a polite way of saying ‘failure’ (Leni Wilde in ODI, 2016). Progress is limited towards reducing transaction costs and improving recipient ownership: failings of the aid system that date back to the development ‘impasse’ of the 1990s (Whitfield, 2008). Crucially, it is regarding donor behaviour change that the GPEDC is widely perceived to have fallen short: ‘recommendations are largely unimplemented by donor countries and agencies’⁴². This is recognised by donors: ‘if we look at what happened to the development indicators since 2011, you could say that results are stable with some improvements, but not very significant or substantial improvements’⁴³. As one CSO representative observed, the ‘problem is that once the [monitoring] report is done, it’s thrown into the ether, and the expectation is that someone will grab it and do something with it… the follow up is largely non-existent’⁴⁴. The reasons underlying the GPEDC’s inability to affect donor behaviour change are twofold. First, as a voluntary and informal MSP dependent upon donor funding, the GPEDC does not wield disciplinary capacity to enforce behaviour change – nor would a ‘naming and shaming’ approach be politically congruent with its ethos of ‘equal’ and ‘horizontal partnership’. Moreover, one DAC representative noted the following on the GPEDC’s capacity to encourage donor behaviour change:

What is the actual influence of the development effectiveness process? [Now] I think it has less. But that's not necessarily got to do with the quality of the GPEDC: it fits within a much broader political agenda. The process will only be as good as the broader environment allows it to be… It is not going to be a counterweight to the more negative self-interest trends that we see globally…⁴⁵

Insofar as the GPEDC is out-of-touch with the broader political environment and donor incentives, the GPEDC will not be able to affect donor behaviour change. In the years following Busan, donor priorities have shifted away from ‘old’ Aid/Development Effectiveness concerns towards the use of aid to catalyse private investment (Mawdsley, 2018): ‘the attention is [now on] innovative financing and partnerships with the private sector’⁴⁶. With the lack of DAC interest in Development Effectiveness, we see a familiar and clear hierarchical power imbalance: recipients are compelled to take on reforms, while successive reports and recommendations are shrugged off by providers, both DAC and SSC.

5. Discussion

The GPEDC sought to overcome legitimacy deficits within global development governance resulting from the emergence of ‘new’ actors. As Alonso (2018, p. 22) notes:

…as the diversity of the actors increases, so does the heterogeneity of interests and visions at play, thus opening the space for new conflicts… it has become necessary to design an inclusive and legitimate structure for development cooperation governance.

While laudable, the above perspective fails to acknowledge that it is within ostensibly ‘inclusive’ and ‘legitimate’ structures where these ‘new conflicts’ manifest. Differences emerge from stakeholder preferences over the procedural function – and normative structures – that governance should adopt. Yet the ‘idea that actors may have divergent views of what constitutes a legitimate institution has received surprisingly little attention’ (Peters, 2013, p. 5). This article’s framework and findings challenge a key assumption put forward by normative-democratic approaches, that the ‘more indicators a global governance arrangement satisfies, and the higher the degree to which it satisfies them, the more legitimate is the organisation’ (Take, 2012, p. 223). By examining normative values through the eyes of stakeholders themselves, simply maximising ‘indicators’ does not lead to more positive legitimacy perspectives. Against the view that ‘the more inclusive a partnership is, the more legitimacy it will have’ (Dodds, 2015, p. 12), augmenting CSO inclusion within the GPEDC would antagonise countries who agitate against enhanced CSO presence. CSOs would likewise resist efforts to ‘bring in’ more corporate voices. Rather, what matters is the interpretation of values associated with legitimacy, and this differs across actors.

Differing perceptions result in intractable governance demands. It is not – for instance – possible to reconcile the demand from CSOs to be recognised as full and equal partners, with the view by some states that CSOs should not possess equal voice. Moreover, a key tension emerges over the primary function of the GPEDC: should it serve primarily as an informal knowledge-sharing forum? Or as a stringent accountability mechanism? I will return to what these intractable perceptions mean for governance and reform trajectories in the conclusion.

The question remains on the extent that MSPs like the GPEDC constitute a ‘New Tyranny’, or whether they provide more legitimate governance. Here we must ask: ‘legitimacy for whom?’ (Cerutti, 2011, p. 122). Although the lack of engagement from large SSC providers undercut the GPEDC’s claim to global legitimacy and partnership, the GPEDC provides a more inclusive, equitable platform of governance than its DAC predecessor. Yet it is also the ‘inclusive’ aspiration of the GPEDC’s accountability framework that – inter alia – dissuaded the large SSC providers from participating. That the GPEDC (and MSPs) assume the characteristics of loose networks of representation, soft forms of accountability, and limited compliance mechanisms is not a ‘natural’ response to the limits of formal, state-based multilateralism. It is a deliberate effort by powerful actors to ensure governance mechanisms act for, or not against, their interests. Insofar as the GPEDC relies upon the financial support of DAC donors, the partnership will – in aggregate – reflect their procedural preferences.

For donors, the GPEDC’s inability to encourage donor behaviour change may be regarded as a strength. In creating a ‘separate autonomous space’ where the DAC is just one of many stakeholders, the GPEDC serves as a buffer that becomes the object of criticism rather than the DAC (Eyben, 2012, p. 79). It is, thus, telling that since the GPEDCs creation there have been fewer criticisms against the DAC’s existence than there were during the earlier ‘Aid Effectiveness’ reform process. Concurrently, the lacklustre enforcement capacity of the GPEDC means there is little pressure on DAC donors to reform in a manner that is against their national and collective interests. The GPEDC can, thus, be considered, like the Aid Effectiveness process, a means through which the DAC has ensured its survival (Chandy & Kharas, 2011).

MSPs such as the GPEDC can, therefore, serve as a means of co-option and domination: a ‘New Tyranny’. But they also offer space for ‘weaker’ actors to cooperate and make demands. Previously disempowered stakeholders such as recipient countries and CSOs are the greatest supporters of the GPEDC process. Moreover, through their collective agency they have sustained the GPEDC’s focus on its accountability function against providers who would otherwise have this diluted. The fact that they have been able to instead protect, and augment, the monitoring function of the GPEDC is testament to the greater strength and influence that these so-called ‘weaker’ actors have gained through the partnership. Within the GPEDC the ‘shadow of hierarchy’ is present (Wolf, 2008, p. 237). Yet coordinated action by other stakeholders can lighten the shade of this influence. While the ‘mainstream’ SDG agenda is moving away from an ODA-centred development system, foreign aid remains a vital resource for many developing countries. Although the GPEDC may be limited in its capacity to affect overall policy trends, or in transforming deeper structural inequalities, ‘weaker’ actors are keeping elements of the ‘old’ agenda alive and sustaining mechanisms that address accountability gaps within the field. Against dichotomous portrayals of MSPs as either innately ‘democratising’ or as ‘tyrannical’ processes, these platforms are, thus, best understood as venues for normative contestation over the procedural governance of policy fields. The outcome of such governance processes is contingent upon the prevailing balance of social forces, and the procedural and substantive priorities that are pursued by these actors. In the context of the 2030 Agenda, wherein MSPs are positioned as the primary governance modality for attaining the SDGs, we now turn to how this balance can be tipped in a more progressive direction.

6. Conclusion

The GPEDC suggests the world is no closer to a global development partnership today than it was at Busan: there is no global platform where large SSC providers, recipients, DAC donors, and non-state actors can convene to advance progress on development. DAC donors continue to dominate within the architecture while recipients are pressured to take on reforms. Concurrently, CSOs are present yet increasingly side-lined within core institutions of development governance. The private sector is an ever-more embedded actor within the field, yet its exact contribution and role in this assemble remains unclear. Although we can say there have been de jure improvements in equality between actors, to suggest we have moved beyond the field’s long-standing power dynamics would be premature.

The diversity of interests and intractability of legitimacy demands suggest it is not possible to have a governance mechanism that fulfils the normative and procedural preferences of all stakeholders. Rather, a choice is required over whose preferences should be prioritised. Here we can take inspiration from the SDGs in that the ambition should be to ‘Put the Furthest Behind First’ and prioritise the views of those with the greatest ‘stake’ in realising Development Effectiveness. This means recipient and CSO demands for a global accountability mechanism ought to be the primary function of the GPEDC, and reforms towards augmenting this function (such as through a post-monitoring follow-up process and by safeguarding against accountability dilution) must be put in place. While this article provides a bleak assessment of the current condition of global development partnership, this analysis does not foreclose the possibility for more transformative governance in this direction.

Although the disproportionate capacities of DAC donors are likely to reproduce hierarchy within MSPs such as the GPEDC, this is not to suggest this influence cannot be countervailed. The first step towards countering provider power is to establish a representative mechanism for recipient countries: re-instating the Partner Country Caucus of the earlier Aid Effectiveness era would be a pragmatic yet powerful driver towards more cohesive and structured recipient engagement in the GPEDC and global development governance writ large. Yet moving towards more robust representative structures and accountability mechanisms would be to move away from the principles associated with multistakeholder governance, such as informality, voluntarism, and ad hoc approaches to participation. Doing so may be necessary for the realisation of a legitimate – or at least a less tyrannical – global development partnership.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Given its empirical focus, this article adopts an ideographic case study approach; it analyses the dynamics of a single yet important case (Eckstein, 1975, pp. 104–108). But it is nevertheless structured by a ‘well-developed conceptual framework that focuses attention on some theoretically specified aspects of reality’ (Levy, 2008, p. 4). This article may, thus, offer ‘transferable’ – rather than directly generalisable – insights for similar contexts and studies.

2 Most of these interviews took place on videoconferencing software between October 2018 and September 2019. A handful took place in London, Paris, and New York. I refer only to general identifying information to protect respondent anonymity.

3 Recipient Ownership, Donor Alignment and Harmonisation, Focus on Results, and Mutual Accountability.

4 Of exception is the work of Xiaoyun et al. (2018), but they only examine large Southern providers.

5 It is difficult to trace causal links between a stakeholder’s participation within a global institution and behaviour change at country level. Nonetheless, given that the GPEDC’s monitoring process is set up to assess stakeholder behaviour change against commitments, this is a pragmatic approach to determining outcome effectiveness.

6 Representatives from: DAC, SSC, and recipient countries; OECD and UN bodies, and civic, private and other non-state actors.

7 Comprised of one ministerial representative from a donor, recipient, and an SSC (‘dual category’) country, and one non-state co-chair.

8 Interview, 30/10/18.

9 Yet ‘small’ SSC providers such as Mexico and Indonesia participate actively.

10 DAC Representative, Interview, 06/12/18.

11 Central African Representative, Interview, 22/05/19.

12 South East Asia Representative, Interview, 15/01/19.

13 European CSO Representative, Interview, 16/10/18.

14 Ibid.

15 North American CSO Representative, Interview, 11/10/18.

16 Central African Representative, Interview, 22/05/19.

17 International civil servant, interview, 17/12/18.

18 Interview, 11/10/18.

19 CSO Representative, 16/10/18.

20 Interview, foundation, 07/11/18.

21 DAC Representative, 25/09/19.

22 Interview, 07/11/18.

23 European CSO, Interview, 20/10/18.

24 Interview, 11/10/18.

25 Interview, CSO, 05/10/18.

26 Interview, 30/10/18.

27 South East Asia Representative, 15/01/19.

28 Interview, 16/10/18.

29 East African Representative, Interview, 16/11/18.

30 European DAC Representative, Interview, 06/12/18.

31 European DAC Representative, Interview, 25/09/19.

32 Central African Representative, Interview, 22/05/19.

33 European CSO, 05/10/18.

34 Interview, 21/09/18.

35 European CSO Representative, Interview, 05/10/18.

36 European DAC Representative, Interview, 25/09/19.

37 South East Asian CSO, Interview, 24/10/18.

38 Observation, DAC Meeting, Paris, March 2019.

39 This was explicit across the interviews and my observations.

40 In 2014, 2016, and 2018.

41 Secretariat, Interview, 16/11/18.

42 Central African Representative, Interview, 22/05/19.

43 DAC Donor, Interview, 6/12/18.

44 Interview, 05/10/18.

45 European DAC, Interview, 05/11/18.

46 European DAC, Interview, 05/11/18.