The Rise and Fall of the Aid Effectiveness Norm

This article analyzes the rise and fall of the aid effectiveness norm, using the lens of Finnemore and Sikkink’s norm ‘life cycle’. I argue that, although donors and recipients endorsed the founding principles, the norm was only feebly internalized: they were unwilling to overcome their reluctance—and important disincentives—to substantially change their practices on the ground. After the norm cascaded, the donor-led process sought to gain legitimacy and diffuse the norm more broadly by trying to bring in a wider range of actors through norm substitution. These changes failed to convince emerging donors to engage and caused the norm to decay to the point where it ceased to constitute a norm. This case highlights the importance of refining the life cycle model to take into account weak internalization and the potential existence of a second phase in which norms decline and potentially die.


Introduction
Faced with often disappointing results from foreign aid, scholars, policy makers and practitioners have been trying for decades to determine what makes aid more effective. In most instances, the blame is placed overwhelmingly on the shoulders of the governments of developing countries. However, in the mid-1990s, a parallel vision of aid effectiveness began to evolve under the leadership of the Organisation for Economic Co-operation and Development and, in particular, its Development Assistance Committee (OECD/DAC), the main club of traditional Western donors. The most prominent result of this process, the 2005 Paris Declaration on Aid Effectiveness, was ground-breaking. It recognized that recipient countries' deficient institutions were partly to blame for a lack of effectiveness in the past. However, crucially, it also recognized that the disappointments and sometimes failures of past aid were not solely due to problems in recipient countries and that donors also needed to change how they worked. The Paris Declaration's principles quickly constituted the overarching norm for 21st-century development cooperation.
Although this international aid effectiveness norm was almost universally embraced, its principles were only slowly and partially put into practice. Faced with a rapidly changing development context, including the rise of non-OECD donors and non-state actors, as well as the growing importance of private financial flows, the international dialogue was reoriented in 2011 towards development effectiveness and responsibility for the process shifted from the Western-dominated OECD to the newly created 'multi-stakeholder' Global Partnership for Effective Development Cooperation (GPEDC). The broader agenda and expanded participation weakened the norm. Emerging donors were very reluctant to engage with the process, non-state actors remained rather peripheral, and more traditional donors lost interest. This process-the rise and fall of the aid effectiveness norm-has never been comprehensively analyzed. This article seeks to do so, using the most common definition of norms in academic literature (as elucidated by Jurkovich 2019) interpreted through the lens of the norm life cycle (Finnemore and Sikkink 1998).
I argue that, although donors and recipients theoretically endorsed the Paris principles, they were unwilling to overcome their reluctance-and important disincentives-to substantially change their practices on the ground. Faced with a legitimacy problem and a changing international development landscape, the donor-led process sought to diffuse the norm more broadly and deliberately tried to bring in a wider range of actors, which had the effect of diluting the norm while failing to convince emerging donors to engage. These changes, paired with shifts in the donor architecture, have caused the process to lose momentum and relevance. In norm life cycle terms, the aid effectiveness norm emerged in the 1990s and cascaded in 2005, but was only feebly internalized, leading to a lack of compliance. Subsequently, norm substitution and decay emptied it of substance and relevance. After 2011, it became increasingly unclear which actors had an obligation to undertake what action, with no stigmatization of non-compliance. The norm underwent a 'reverse cascade', and aid effectiveness ceased to meet the necessary conditions to constitute a norm.
My account of the evolution of the aid effectiveness norm is based mainly on primary sources, including official documents that resulted from international meetings on aid effectiveness and published accounts from some participants, as well as secondary literature. My explanations for the norm's fall are based on a review of the literature, supplemented by 3 months of fieldwork in Mali, Ghana andEthiopia, in 2015 and where I conducted 60 interviews on the topic of aid effectiveness with donor and government representatives, non-governmental organization (NGO) officials, academics and other informed individuals. I selected those countries for both their shared characteristics (location in Sub-Saharan Africa, the region that receives the most foreign aid; currently and historically significant aid programmes; large number of donors; and strongly expressed commitment to aid effectiveness) and the variation these provide in state capacity (lowest in Mali, highest in Ethiopia) and in wealth (Mali and Ethiopia are low-income countries; Ghana is a middleincome one).
The remainder of this manuscript is organized as follows: First, I define norms and present theorizations concerning the norm life cycle model, which serve as the analytical framework for this paper. Second, I establish how aid effectiveness constitutes a norm. Third, I trace the rise of the aid effectiveness norm from the mid-1990s to 2011 through the emergence, cascade and internalization phases of the life cycle. Fourth, I analyze the fall of the norm through the phases of norm substitution, decay and reverse cascade. Fifth, I provide the main reasons for the norm non-compliance and failure. The conclusions sum up my findings, highlight their significance and suggest potentially fruitful paths for future research.

Norms and Their 'Life Cycle'
Constructivist international relations scholars generally agree that a norm is 'a standard of appropriate behavior for actors with a given identity' (Finnemore and Sikkink 1998, p. 891). As Jurkovich (2019, p. 2) argues, such definitions imply three essential elements for something to constitute a norm: there must be a moral obligation or sense of 'oughtness', the responsible actor needs to be defined, and that actor must be expected to take specific actions or behave a certain way. If actors are not defined, it is not a norm but rather a moral principle, no norm violators can be meaningfully identified, and no form of sanctions, such as shaming, can be applied. If instead there is no sense of obligation, then the phenomenon merely describes an specific actors' behaviour, again precluding the discrediting of violators.
Not all scholars, however, agree that the term 'norm' must meet those three conditions. For instance Fukuda-Parr and Hulme's (2011, p. 19) analysis of the Millennium Development Goals considers each of the eight goals to be an individual norm. However, using Jurkovich's (2019) more restrictive definition, as I do in this article, the goals do not constitute norms, as they lack the explicit link between actors and actions. Although almost all individual goals are accompanied by explicit quantitative targets, no individual actor is compelled to act or can be shamed if the targets are not met, because goals do not define who is responsible for meeting each of the targets or how.
In this article, I am interested in the aid effectiveness norm. Whereas the idea that foreign aid should be used well and not wasted is a moral principle, a norm requires the specification of what measures need to be taken and by whom. This occurred in the 2005 Paris Declaration on Aid Effectiveness, which elevated this long-standing principle to a norm.
Norms do not just appear fully formed but progress through a series of sequential stages. Finnemore and Sikkink developed a widely cited norm life cycle model to apply to the behaviour of states. According to the model, a norm passes through three stages: first, it emerges, propagated by 'norm entrepreneurs'; second, it spreads to a critical mass of states, and once it reaches a 'tipping point', it cascades; and third, the norm is 'internalized', that is to say, taken for granted and no longer subject to meaningful debate (Finnemore and Sikkink 1998, p. 895). I trace that phenomenon below in relation to foreign aid, labelling it the rise of the aid effectiveness norm.
Finnemore and Sikkink's life cycle model has had tremendous influence on the study of international norms. Subsequent studies have brought further nuance to the model by examining aspects and dynamics that it underplays or ignores. For instance, although Finnemore and Sikkink make clear that not all norms make it all the way through this cycle, they say little about how or why an emerging norm might never make it to the tipping point. Elgström (2000) fills in some of the gaps by analyzing 'norm resistance', the counterpart of norm entrepreneurship, and 'norm ignorance', both hindering norm diffusion. Rosert (2019Rosert ( , p. 1105) refers to Finnemore and Sikkink's life cycle as presenting the 'entire biography of a norm'. However, this perspective assumes compliance and ignores what happens after the internalization of the norm, including the possibility of 'norm degeneration'. As Wunderlich (2013, p. 28) points out, 'even internalized norms may be weakened, lose their taken-forgranted status, or eventually decay'. The longevity of norms can also be limited by a process of norm redefinition or substitution, whereby a norm is revisited 'in light of new challenges or opportunities' (Lantis 2016, p. 388). In addition, as McKeown (2009, pp. 7, 11) highlights, the internalization of norms can be 'shallow and fleeting', potentially leading to their expiration or death through a reverse cascade process during which 'normative stigma for breaking the norm is now significantly reduced'. The subsequent phase does not always present, but in my case study, I refer to this phenomenon as the fall of the aid effectiveness norm. This article thus traces the aid effectiveness norm's rise and its fall, using the concepts presented above.

Aid Effectiveness as a Norm
This article analyzes the evolution of the aid effectiveness norm, defined by the Paris Declaration on Aid Effectiveness. Adopted in 2005 at a high-level forum organized by the OECD/DAC, it contains five basic principles, which the OECD (n.d., p. 1) summarizes as follows: 1. Ownership: Developing countries set their own development strategies, improve their institutions and tackle corruption. 2. Alignment: Donor countries and organisations bring their support in line with these strategies and use local systems. 3. Harmonisation: Donor countries and organisations co-ordinate their actions, simplify procedures and share information to avoid duplication. 4. Managing for results: Developing countries and donors focus on producing-and measuring-results. 5. Mutual accountability: Donors and developing countries are accountable for development results.
These are what I refer to as the five Paris (or aid effectiveness) principles. This term is common usage and should not be confused with the concept of moral principles introduced in the section above. The declaration also contains 13 indicators of progress, with explicit targets for donors and recipient countries to be met by 2010.
In this article, I focus mainly on the first three Paris principles, as they have the greatest potential to affect aid relationships on the ground. The OECD sometimes represents them as a pyramid, with ownership, numbered 1, on top, alignment (#2), in the middle, and harmonization (#3), at the base (OECD 2012, p. 18). The diagram thus sets out a sequence of visually interrelated principles. The ownership-alignment-harmonization sequence presents an optimistic scenario in which developing countries are at the helm, while donors set aside their own priorities, self-interest and rivalry and work together-and with recipients. Henceforth, recipients would set their own development priorities and plans, donors would adopt them as their own and coordinate among themselves to implement them.
The other two Paris principles are harder to integrate into a coherent narrative. The OECD represents them floating above or lying beside the pyramid but not composing it. They are nonetheless significant. The Paris Declaration describes 'managing for results' (principle #4) as 'managing and implementing aid in a way that focuses on the desired results and uses information to improve decisionmaking' (OECD 2008, p. 7). It implies a shift of emphasis from inputs to outputs and outcomes. Attention to results is an important principle, but it has essentially been recast technocratically. The focus on results is primarily a management tool-albeit one with important implications and consequences-that aggregates project data into large-scale reporting and monitoring frameworks. As I am more concerned with the macro aid relationships, I do not analyze this principle here. However, there is rich literature that delves deeply into the area of results-based management and the broader emphasis on quantifiable results, including how it can undermine ownership (e.g. Sjöstedt 2013).
While managing for results is represented as floating above the pyramid, mutual accountability (principle #5) appears to be shunted off to the side. As described in the Paris Declaration (OECD 2008, p. 8), it comprises three things: recipient governments' involvement of parliaments and other domestic actors in the formulation and assessment of national development strategies, donors' provision of comprehensive and timely data on aid flows to recipient governments, and joint donor-recipient assessments. The first component can be addressed under ownership and the third under harmonization (when donor coordination includes government participation or at least dialogue with the government). I therefore do not separately analyze the mutual accountability principle here.
As demonstrated below, the Paris Declaration codified a norm that had been emerging for over a decade. 1 I use Paris metonymically to represent the principles that were adopted there and the norm that together they constitute. As the declaration was very widely endorsed, it marks the norm's tipping point. Finnemore and Sikkink's norm life cycle model and the elucidations it generated provide useful tools for framing and better understanding the norm's rise and fall.

Norm Emergence: Aid Effectiveness before Paris
None of the Paris principles was actually new. The norm had been emerging for a long time. For instance the need for ownership and its corollary alignment had already been part of the development discourse for decades, although it was often a polite fiction. It currently seems commonsensical: If the project or programme is to have any chance of success, the host government needs to want it to work. If the programme is imposed without the government's buy-in, it will not allocate the necessary resources to implement it or will otherwise drag its feet, almost guaranteeing failure.
The end of the Cold War lessened Western donors' desire to use aid out of strategic self-interest, opening the door to an initiative for greater aid effectiveness. Though the promised 'peace dividend' did not ever materialize and donors did not increase their aid budgets in the 1990s, the international context opened a space for aid actors to seek ways to improve the impact of their aid. The overall failure of structural adjustment to produce economic growth in developing countries also suggested that donors themselves, and not just recipient governments, should modify how they operated if they wanted to promote development.
The norm emergence accelerated after 1996, when the OECD/DAC adopted a document entitled 'Shaping the 21st Century: The Contribution of Development Co-operation', that outlined its vision for reforming foreign aid. The document argues that success requires an approach that, among other things, 'respects local ownership of the development process' (OECD 1996, p. 9). Likewise, it underlines the importance of donor coordination, even if it did not use the term 'harmonization'. The Monterrey Consensus that resulted from the International Conference on Financing for Development, held in Mexico in 2002, also contains important elements of what would become the aid effectiveness norm. For instance it recognized that 'Effective partnerships among donors and recipients are based on the recognition of national leadership and ownership of development plans' and that bilateral and multilateral donors need to improve the coordination of their aid (UN 2003, p. 14).
The use of the term harmonization to describe coordination emerged with the Rome Declaration on Harmonisation, adopted two years before the Paris Declaration at the previous high-level forum on aid effectiveness. While using the term 'ownership' only in passing, it highlighted the need to 'Ensure that development assistance is delivered in accordance with partner country priorities' (OECD 2003, p. 11).
Even if the concepts were part of previous policy documents, the Rome Declaration was the first official building block of the aid effectiveness norm in a process that continued with the 2005 meeting in Paris and the third high-level forum in 2008 in Accra, Ghana, which resulted in the Accra Agenda for Action. This series of summits culminated in Busan, South Korea, in 2011, at which point both the concept of aid effectiveness and the debate on it metamorphosed.

Norm Cascade: The Paris Declaration on Aid Effectiveness
The Paris meeting was the first one that focused on the five aid effectiveness principles described above. The Paris Declaration constitutes the aid effectiveness norm's founding document. It was signed by 138 donor and recipient countries, 28 international organizations and numerous NGOs and civil society networks.
Aid effectiveness, as codified in the Paris Declaration, meets Jurkovich's (2019) three requirements of a norm. First, the expected actions are spelled out, not only in the five principles but also in the indicators and targets. Second, the actors responsible for implementing the norm are clearly specified: bilateral and multilateral donors and recipient country governments, depending on the indicators and targets. Third, there is a sense of oughtness that signatories have made a common commitment to applying the Paris principles and that failure to live up to it would be a moral liability.
The Paris summit represents the moment the aid effectiveness norm reached its tipping point and cascaded. As Wood and Betts (2012, p. 196) argue, 'The Declaration's core principles and commitments have built on, reinforced, and disseminated the earlier good practices of different countries and donors and become widely accepted norms for good practice in development cooperation'. The Paris Declaration embodies the norm, even though this continued to evolve slightly during the internalization process.

Norm Internalization: From Paris to Accra
In the years following the Paris Declaration, donors and recipient governments internalized the norm, reflected to differing degrees in the way they formulated policy, delivered foreign aid and reported on their results. Although taken for granted after 2005 and not subjected to meaningful debate, the aid effectiveness norm was periodically subject to further refinement and efforts to accelerate progress. The Paris principles were tweaked three years later at the next high-level meeting, held in Accra, Ghana, in September 2008. Compared with previous meetings, it had significantly more participation from developing countries as well as civil society organizations. The resulting Accra Agenda for Action both clarified some of the concepts that were at the core of the Paris Declaration and adapted them to the rapidly changing aid landscape. It noted that progress was too slow and that all parties had to make a concerted effort to accelerate the changes being made. To do so, the document emphasized three major challenges (OECD 2008, p. 16).
The first challenge concerned ownership. The Paris norm had raised concerns that it overemphasized state actors on both the donor and recipient sides. By side lining civil society organizations and reinforcing the power of authoritarian rulers, the Paris principles could undermine democracy, international solidarity and grass-roots projects. The Accra Agenda for Action therefore clarified that ownership was not the sole preserve of the government or the executive, emphasizing the need for input from a wide range of actors, including parliaments, sub-national governments and civil society organizations.
Second, in response to shifts in the aid landscape, the Accra Agenda for Action expressed the need to include a broader range of actors, including 'middle income countries, global funds, the private sector, [and] civil society organisations' (OECD 2008, p. 17). It recognized that the growth of these actors' importance would increase aid fragmentation and further complicate coordination, and therefore encouraged them all 'to use the Paris Declaration principles as a point of reference in providing development co-operation' (OECD 2008, p. 18). Third, as part of the focus on results, the Accra Agenda for Action shifted the emphasis under mutual accountability to 'citizens and taxpayers' of both donor and recipient countries (OECD 2008, p. 17).
In refining the Paris principles, Accra constituted what Mawdsley et al. (2014, p. 35) called 'the high point of the aid effectiveness paradigm'. It further defined some of the content of the norm and made some changes to reflect the contemporary aid regime but without compromising its 'taken-for-grantedness'. Although the changes constituted a case of limited 'norm redefinition', they were not significant enough to be described as 'norm substitution'. However, the signalling in Accra about bringing in new types of actors opened the door to substitution a few years later, which set off a reverse cascade, as described below and explained in the article's final section.

Norm Substitution: The Busan Forum
The next high-level forum, held in Busan, South Korea, in 2011, was originally intended to be another meeting to review progress rather than challenge takenfor-granted principles. However, it unexpectedly became 'an important turning point in the history of development cooperation' (Atwood 2012, p. 1) and created a 'paradigm shift' (Mawdsley et al. 2014). Following the intentions outlined at Accra, the organizers broadened the invitation list, which included representatives of the private sector and significantly increased participation from civil society organizations, recipient countries and emerging donors. In parallel, also reflecting changes in the international development ecosystem-including the aftermath of the global financial crisis, the accelerating rise of China as a global actor and the impact of private flows on development-the agenda became much broader and was no longer limited to aid. The concept of aid effectiveness was dropped and replaced with development effectiveness. Even if the term had not been defineddeliberately, to avoid 'definitional conflict' (Eyben 2013, p. 87)-participants wanted to recognize that development could be the result of a number of factors beyond traditional aid. They therefore needed to take into account contributions from the developing countries themselves, suppliers of South-South cooperation and the private sector, as well as other financial flows such as investment, trade and remittances.
Several emerging donors, in particular China, India and Brazil, were reluctant to sign the Busan agreement. They saw the aid effectiveness norm as reflecting OECD perspectives incompatible with South-South cooperation and otherwise poorly adapted to the evolving development landscape (Abdel-Malek 2015, p. 180). Traditional donors realized that, without the participation of China in particular, the 'emerging development architecture' would lack 'both legitimacy and efficacy' (Abdenur 2014(Abdenur , p. 1889. To secure emerging donors' signatures, the declaration's wording deliberately de-emphasized some of the Paris principles, among other concessions, without completely abandoning them. Although the final statement included a reminder that 'those of us that endorsed the mutually agreed actions set out in Paris and Accra will intensify our efforts to implement our respective commitments in full' (Busan Partnership 2011, p. 4), it dropped any mention of alignment and harmonization as fundamental principles, while only briefly reiterating the importance of ownership. It added transparency-a major step down from harmonizationand inclusive partnerships as basic principles. Crucially, it also recognized that non-traditional actors should not have to engage in the same way as traditional donors and that adherence to the agreements' principles was voluntary in cases of South-South cooperation, thereby setting some explicit limitations to the applicability of the prescriptions.
Also as a concession to Brazil, India and especially China, the final document lacked some basic key elements of any such agreement, including 'not only a concrete timetable, but more importantly, objectives and measurable indicators' (Lightfoot and Kim 2017, p. 171), making implementation impossible to track. Moreover, 'it soon became clear that there was no shared understanding of what this agreement actually signified' (Constantine et al. 2015, p. 10). Thus, Busan both broadened and weakened the Paris principles. In fact, at Busan, the Paris targets were recast as ideals (Mawdsley et al. 2014, p. 31). As Eyben and Savage (2013, p. 457) noted, 'The old certainties about the norms of international development cooperation were disappearing in a rapidly changing global political economy, while new modes of behaviour were yet to strongly emerge'.
In theory, the increased vagueness of an evolving norm could favour its stability and longevity by increasing its flexibility and the ease of compliance (Panke and Petersohn 2016), as could the fact that it could be framed as a 'norm cluster' (Lantis and Wunderlich 2018). Instead, the norm substitution triggered its weakening, reversal and decay. In fact, neither the Busan outcome nor its constituent elements retained the three requirements of a norm. The applicability of the principles was expanded, but the list of actors became inchoate. The actions required also became quite unclear, while the prescriptive nature was restricted to traditional donors, and the language weakened the degree of oughtness of the principles, greatly diminishing stigma for not adhering to them. As Jurkovich (2019) has argued, optional actions constitute a moral principle, not a norm per se. By that definition, aid effectiveness could even be said to have lost its status as a norm in Busan.

Norm Decay and the Reverse Cascade: The Global Partnership for Effective Development Cooperation
The Busan document called for a new more inclusive Global Partnership for Effective Development Cooperation to replace the OECD as the body in charge of leading the process. Launched in 2012, one year after Busan, the GPEDC took over the agenda from the OECD. Led by representatives of development cooperation providers and recipient countries, local governments, civil society, the private sector, private foundations, the United Nations (UN) and the OECD, it is an amorphous platform with no permanent base of its own, depending on the OECD and UN Development Programme to support it. With a membership comprising 161 countries and 56 organizations, it sought to replace the binary vision of Northern donors and Southern recipients with an 'equator-less' (Eyben and Savage 2013, p. 467) perspective of a wide variety of actors, with often overlapping roles.
Like the OECD-led process it replaced, the GPEDC holds high-level meetings every few years. The first one, hosted by Mexico City in 2014, issued a statement that contributed 'little by way of concrete actions or commitments' that were new, mainly reiterating what had already been agreed in Busan three years earlier, as well as adding some discussion of other topics, such as climate finance, fragile states, middle-income countries and domestic resource mobilization (Abdel-Malek 2015, p. 324; Government of Mexico 2014, pp. 120-125). Its first annex consisted of a list of 39 voluntary initiatives. Neither China nor India participated in the meeting, while Brazil attended without actively participating (Keijzer and Lundsgaarde 2016, p. 2).
The members of the GPEDC met again in 2016, in Nairobi, Kenya. The agenda was extremely broad, with little time allocated to discussions of the reasons for the disappointing results (outlined below). The Nairobi Outcome Document contains calls for the inclusion of all actors in all sectors, from security to humanitarian assistance, the need to link up with the Sustainable Development Goals and many more things-an agenda full of good intentions but extremely diluted. With only one exception, namely reducing the transaction costs of remittances by 2030, none of the commitments was assigned a specific deadline or target date, which did little to spur signatories to implement them (Bena and Tomlinson 2017, p. 5). Again, China and India did not send representatives, nor did Brazil and South Africa this time, highlighting a lack of buy-in among emerging donors, the main purveyors of South-South cooperation (Klingebiel and Li 2016).
Many GPEDC members believe that the partnership 'has been significantly less effective than the previous structures and to date has had limited impact on the ground' (Janus et al. 2016, p. 3; see also Abdel-Malek 2015, p. 321). Moreover, since Busan, OECD aid actors have lost interest in the evolving agenda. At subsequent meetings, very few OECD countries have sent high-level representatives, unlike in Busan. On the ground in recipient countries, traditional donors are still operating to a certain extent under the Paris principles, with a focus on aid. Albeit somewhat anecdotal, comments made by Western officials I interviewed in Addis Ababa in 2016-on the eve of the Nairobi high-level meeting-appear indicative of the GPEDC's lack of purchase: The head of cooperation of a European country stated, 'We are aware of the GPEDC, but we don't have time to engage. Donors are not using it as a framework or in their reporting', unlike the Paris principles, and 'It is unclear how relevant [the monitoring] is'. 2 A senior aid official of another European country was even more dismissive: 'We have very little to do with the Global Partnership. It is very theoretical, pretty useless'. 3 A Western diplomat agreed that 'the Global Partnership is not perceived as an important policy priority'. 4 Bracho (2017, p. 3) argues that, 'the GPEDC has focussed on "discourse" rather than "action"… and has lost much of its political punch. It is no longer the GPEDC envisioned at Busan'. Instead, it has increasingly concerned the sharing of best practices for development. As the GPEDC shifted to a knowledge exchange platform, without any more agreements or official statements, the pressure on traditional donors to respect their prior commitments was reduced even more, and the remnants of the aid effectiveness norm further decayed. Having lost its oughtness and takenfor-grantedness, the erstwhile norm was replaced by, at best, a set of vague principles, while its institutional framework increasingly became primarily a means of sharing ideas and experiences. In the next section, I document the generalized lack of compliance and offer the main explanations for the failure of the norm.

Non-compliance
The OECD hoped that the aid effectiveness norm would 'significantly increase[] the impact of aid' (OECD 2006, p. 49). However, it recognized that 'serious and sustained political resolve at the highest level' was necessarily for it to succeed (OECD 2006, p. 54). Accra was meant to refine those principles and increase the probability of success.
Donors and recipient governments, however, proved unwilling (or perhaps unable) to change their behaviour. They achieved only 1 of the 13 targets they set for themselves for 2010, namely having at least 50% of 'technical co-operation implemented through co-ordinated programmes consistent with national development strategies' (OECD 2012, pp. 15, 19-21). OECD/DAC donors failed to concentrate their aid in a smaller number of countries, despite the consensus on its desirability, including as a component of the Paris principle of harmonization (Brown and Swiss 2013). Donor coordination actually worsened after the Paris Declaration (Nunnenkamp et al. 2013). Case studies suggest that, if anything, donors strengthened their capacity to impose their priorities on recipient governments during this period (Bidaurratzaga-Aurrea and Colom-Jaén 2012; Hayman 2009).
Even if some progress was made in some areas, the overall picture is one of a failure of the signatories to meet the targets they set for themselves and to put the norm into practice in substantively meaningful ways. Even if actors did not seek to repudiate the norm, its internalization was 'shallow and fleeting', as McKeown (2009) warned norms could be. In donor countries especially, 'the necessary political, bureaucratic, and public understanding and support for difficult reforms have been hard to secure and maintain' (Wood and Betts 2012, p. 108). Because of the limited internalization of the norm was widely shared, non-compliance generated very little stigma, which in turn provided little pressure for actors to modify their behaviour, creating a vicious circle.
While sidelining the aid effectiveness norm, the Busan process also abandoned the monitoring process for implementing the Paris principles. Instead, in 2012, the GPEDC adopted 12 new quantitative targets to be achieved by 2015. The indicators, unlike those adopted in Paris, applied not only to donor and recipient countries but also civil society and the private sector, at least in theory. Several of them were vague and broad (GPEDC n.d.).
The 2016 Monitoring Report was launched just before the Nairobi GPEDC meeting. In it, the results vis-à-vis the targets are not clearly presented, which de-emphasizes the fact that, although there had been some progress in most indicators, it would appear that not a single specific target was achieved, other than some of the unambitious ones that were defined solely as 'progress over time'. In fact, some trends have been going in the wrong direction. For instance, although the goal was for half the countries to improve their country systems, only 11 out of 60 surveyed did, while 14 others actually experienced a decline (OECD/ UNDP 2016, p. 62). This feeble performance points to the GPEDC's weak power to compel a change in both donor and recipient behaviour and to continued norm resistance.
At that point, there is no doubt that the aid effectiveness norm had almost fully degenerated. The changing development landscape, including the decline of the role of foreign aid, could have led to the substitution of a robust norm (Lantis 2016). However, after Busan, actors and especially expected actions became too broadly and weakly defined, and any sense of moral imperative decayed to the point where stigma for non-compliance had all but disappeared. A reverse cascade ensued, and only the disbandment of the GPEDC could more fully reflect the norm's demise.
The reasons for norm resistance and the feeble application of the Paris principles are numerous. I now turn to the main ones, addressing them according to the type of actors: recipient countries, traditional donors, emerging donors and other actors, as well as the GPEDC itself.

Recipient Countries
Based in large part on field research I conducted in Mali and Ghana (Brown 2017), I have identified three significant problems that undermine the meaningfulness of the application of the Paris principles. In both countries, the level of ownership is high, in the sense that their development plans-or at least the most recent ones-have been developed nationally, and the government sees them as overarching frameworks. However, first, each country has several competing and inconsistent plans, without any clear indication of which one actually embodies the owned development vision. Second, the plans lack strategic prioritization and resemble more of a wish list in virtually all possible sectors, also preventing clarity on what is actually owned. Third, follow-up is often inadequate, which prevents theoretical ownership from actually being translated into reality.
These deficiencies can be explained by the following main factors. First, a lack of government capacity (Wood and Betts 2012). Second, a lack of government leadership or will to prioritize. After all, there are great needs in all sectors. Moreover, concentrating on a smaller number of areas contradicts the goal of aid-revenue maximization. These funds are also very useful resources to maintain balance in neopatrimonial systems and avoid making hard choices that could alienate some domestic allies. Few governments will turn down free money, even if it does not contribute to their de facto priorities.

Traditional Donors
Most donors showed a basic lack of commitment to implementing the Paris Declaration (Wood and Betts 2012), and their latest aid policy documents barely mention any of the aid effectiveness principles, let alone the 'aid effectiveness agenda' itself (Lundsgaarde and Engberg-Pedersen 2019). Nonetheless, some of their practices on the ground do follow to a limited extent the spirit of the principles. Alignment with government priorities can hardly be considered a challenge in Mali, Ghana and other countries whose stated development priorities are all-encompassing. Donors can undertake virtually any activity that they like, including those that will please their political masters or public at home, and legitimately claim that it is aligned with one or more planning documents. Donors are generally reluctant to embrace the second component of alignment as defined by the Paris Declaration, the use of country systems, with the exception of a decade or so of significant general budget support, now virtually over-and a reasonable case can be made that they are right not to because of mismanagement and corruption.
In most recipient countries, donors are undertaking extensive harmonization activities, meeting in working groups and engaging jointly at times with government counterparts. Most of these activities, however, are limited to information sharing, which is useful but falls short of the division of labour or truly joint activities that would actually reduce transactions costs as foreseen in the Paris Declaration. Donors are quick to blame the host government for insufficient donor coordination, as it does not want donors speaking with one voice, preferring to interact with them bilaterally and potentially 'shopping around' a pet project. However, donors too have powerful incentives not to subsume the thematic focus and visibility of their work into collective efforts. Donor officials' career mobility, for instance, depends far more on responding to their own government's priorities and flying their country's flag. There is, however, one partial exception: instances of joint EU programming, which member states officially embrace but also resist for the same reasons and that has more to do with Brussels than Paris.
Finally, since Busan, traditional donors collectively have become less altruistic in their respective aid programmes. As Mawdsley et al. (2014, p. 29) argue, 'Shifts in political leadership in a number of DAC countries and the global financial crisis have stimulated a much stronger discourse of national interest and "value for money" within foreign aid', contributing to the rise of the explicit desire for 'mutual benefits' (Keijzer and Lundsgaarde 2018). In fact, some traditional donors have grown to resent the fact that the GPDEC commitments apply to them but not to emerging donors, giving the latter unfair advantages to pursue self-interest openly (Bracho 2017, p. 2).

Emerging Donors
Although, at Busan, emerging donors-notably India and China-initially demonstrated a willingness to engage with the evolving aid effectiveness institutional mechanisms and norm, most never participated meaningfully in the GPEDC. Even though they had obtained an opt-out for the Busan agreement's applicability to their own development cooperation activities abroad, they objected to being subjected to the same monitoring scrutiny as other countries-and worried that other forms of pressure would be applied to them in the future. They also did not want to participate in the construction of the new architecture in what they considered a continuation of the liberal order, with the DAC donors trying to contain and 'socialize' them into behaving like traditional donors rather than engaging in mutual learning and fundamentally modifying the international aid regime. They also believed that the UN, not the GPEDC, was the only appropriate and legitimate venue for these discussions (Bracho 2017;Constantine et al. 2015, p. 2, 10;Li et al. 2018). Whereas the Paris principles were not meant to apply to them in their capacity as providers of development cooperation (only as recipients), the most important emerging donors disengaged from the process after the GPEDC was created.

Other Actors
Civil society actors have been increasingly included in the aid effectiveness discussions, especially at Busan. They have a non-voting executive position at the GPEDC. They are probably the most enthusiastic members of the GPEDC, principally because they have a greater role at the multi-stakeholder GPEDC than they would at the logical alternative forum, the state-based UN. Within the platform, they have tried to advance agenda items that include strengthening of the principle of democratic ownership, greater focus on a human-rights-based approach and poverty reduction, the participation of civil society organizations as 'independent development actors in their own right' and reform of the development cooperation system (Mawdsley et al. 2014, p. 33). However, as the GPEDC flounders, the platform has not proved very useful for achieving their goals.
The private sector, too, has been included since Busan, albeit not as visibly as civil society organizations. Other actors hoped that it would 'bring efficiency and focus in a way that the aid-effectiveness agenda of the past 10 years had not managed to achieve' (Mawdsley et al. 2014, p. 34). However, these aspirations have not been met. The GPEDC has developed a clear concept of how to integrate the private sector but not of how to operationalize representation from such a disparate set of actors. Very few private sector actors have heard of the GPEDC, let alone seek to engage with its agenda and participate in its activities.

The Limitations of the Global Partnership
As a complex network of state and non-state actors with weak institutionalization, the GPEDC lacks a strong base from which to promote its agenda. Its rotating leadership has proved at best inconsistent, with changing priorities (Keijzer and Lundsgaarde 2016, p. 3). A particular challenge is that its mandate is extremely broad, as 'effective development … suggests an open-ended agenda that could encompass (almost) anything; in fact, themes as diverse as domestic resource mobilisation, middle-income countries, youth and the private sector, to name just a few, figured prominently in the first two [high-level meetings] of the GPEDC. This thematic expansion, in itself, seemed to weaken the action agenda, as it was unlikely that a forum such as the GPEDC could generate deliverables for these topics' (Bracho 2017, p. 33).
The platform has very little ability to influence its members and promote collective action. Both traditional and emerging donors have lost interest. Fundamentally, 'The lack of cooperation between traditional and emerging donors was helping to drive the entire development cooperation agenda towards a future of diluted standards, practices, commitments and responsibilities for all official donors, both Northern and Southern' (Bracho 2017, p. 25). The lack of stigma for non-compliance with any of the principles or ideals articulated since the 2011 Busan forum has facilitated the GPEDC's tenuous survival, but it also means that the vestiges of the Paris process no longer constitute a norm at all. Though some of the Paris rhetoric makes the occasional appearance and some Paris principles are still considered good practice, the aid effectiveness norm is effectively dead.

Conclusions
The aid effectiveness norm was the result a deliberate self-conscious process initially led by OECD donor countries. The norm emerged in the 1990s, cascaded in 2005 and reached its apogee in 2008, but was only feebly internalized. Its diffusion was successful in gaining formal endorsement but much less so in overcoming resistance and significantly changing behaviour, including among the norm's originators. The norm's failure was due in large part to a lack of political will and significant incentives of various types on all sides to maintain the status quo, including donors' desires to achieve self-interested goals and recipients' desire to maximize financial inflows and maintain policy space while manifesting agreement with the principles and performing compliance.
Starting in 2011, the norm was redefined, which caused it to decay and, as a result, be extinguished. The post-2011 norm substitution attempted to make the principles more relevant to the evolving global development context and engage a broader range of actors. However, the compromises made since then have gutted the original norm by broadening the agenda and diluting the prescriptions, yet failing to bring in the important new actors, not least emerging donors who saw no advantage in engaging. In doing so, it may have nominally extended the norm's longevity, but it also emptied it of its substance and influence, effectively killing the norm.
What remains of the aid effectiveness norm, one and a half decades after the Paris Declaration, is a loose grouping of various types of actors within the GPEDC, who hold periodic meetings, but whose main task appears to be monitoring a broad list of too-numerous and frequently vague indicators and sharing best practices. While traditional donors have lost interest and emerging donors have failed to engage, actors feel very little pressure to comply. The main promoter of aid/development effectiveness, the GPEDC, is a weak institution without a permanent secretariat and is still struggling to establish its legitimacy and carve out its place in parallel to the UN and other forums for discussing aid and development. It is thus unlikely to resurrect the norm or create a new one.
This article provides a comprehensive account of the rise and fall of the aid effectiveness norm, which constitutes an original contribution in its own right. The lens of the norm life cycle provides additional insight, highlighting in more analytical terms the emergence of the norm and its cascading in line with Finnemore and Sikkink's model, helping place the norm's evolution within a broader universe of cases, even if it does not conduct an explicit comparison. Moreover, the case of the aid effectiveness norm provides a useful example of how norms can deviate from the life cycle model, in particular through feeble internalization and significant de facto norm resistance in spite of a cascade.
Furthermore, the analysis of not only the rise but also the fall of a norm contributes to the sparse but growing literature that highlights the importance of extending the life cycle to a second phase in instances where the norm is not permanently internalized or taken for granted. As the case of aid effectiveness illustrates, subsequent norm redefinition and substitution to promote its diffusion can cause dilution, propel norm degeneration and, ultimately, lead to the norm's death.
Due to limitations in the data currently available publicly or from my field-based research, my explanations for the evolution and weak application of the aid effectiveness norm are partial. For a more complete picture, future research could analyze in greater detail the role and perspectives of the individuals involved in the international negotiations and the drafting of the documents that codified the norm, including people who would fall into the category of norm entrepreneurs or resisters. To be fruitful, such analysis would need to include the intra-and inter-organizational dynamics of norm internalization. In addition, more work could explore the impact of structural shifts, such as the global financial crisis and the rise of China as an international player, mentioned here as part of the context, on the aid effectiveness norm. A greater focus on both individual agency and overarching structure would thus provide complementary insights to the account and findings presented herein.