A Prototypical Case in the Making? Challenging Comparative Perspectives on French Aid

The comparative literature presents a mixed picture of French development assistance, with qualitative studies portraying this as a ‘deviant’ case while variable-led analyses view it as broadly ‘representative’. Both sets of studies ignore the possibility that French aid might be ‘prototypical’, laying down ideas and practices for other donors. Drawing upon a conceptual framework and over twenty interviews, this article challenges these comparative perspectives and paves the way for wider consideration of French ideas on international development. It argues that French assistance has undergone ‘mainstreaming’ and developed prototypical features since Prime Minister Jospin’s reforms of the late 1990s. It attributes these prototypical characteristics to France’s continuing ambivalence towards international aid norms, its more strategic approach to promoting French ideas and the emergence of a more propitious intellectual climate for those ideas. It concludes by reconciling different perspectives on French assistance and exploring the implications of these findings for Northern aid.


Introduction
There is now an established comparative literature on overseas development assistance (ODA) or foreign aid. These writings look for cross-national trends (Lumsdaine, 1993), identify 'good' donor practices (OECD, 2009) or simply compare the governance structures (van Belle et al, 2004), volume (ONE, 2006(ONE, -2015, quality (CONCORD, 2014), effectiveness (Cox et al, 1997) or developmental goals (Lancaster, 1999) of different, mainly Northern, assistance programmes. This literature often considers the French case, which is unsurprising given that France has traditionally been one of the world's top four donors, home to the 'donor club'the Development  OECD, 2013) and from contributions by French officials to this literature (Gaulme, 1998;Pacquement, 2010). It follows that any misrepresentation of French aid is not purely a matter of academic interest.
Finally, there is no authoritative yardstick by which to measure the 'deviant', 'representative' or 'prototypical' nature of any programme. To provide a gauge, this study looks beyond the comparative literature and draws upon a long-run series of interviews with over 20 academics, international NGOs and officials from France's main agency, the Agence Française de Développement (AFD), the French Ministry of Foreign Affairs (MFA), the Welsh Assembly government, the Department for International Development (DFID) and the UK foreign policy establishment. It also employs a conceptual framework that examines actual, potential and possible uptake of French ideas according to whether they have, respectively, been adopted, drawn upon or given consideration by fellow donors.
There are inevitably caveats associated with any approach aimed at 'pigeonholing' donors. Clearly no donor can ever be 'deviant' or 'representative' across the board. Furthermore, the prototypical nature of any donor practice and its potential benefit to development can often only be evaluated qualitatively and over the long term, by which time it will either have been taken on board by some (usually not all) donors or will have run up against indifference or resistance, often from donors pushing different practices.
A final caveat relates to the taking of OECD norms as a baseline for 'measurement'. Such an approach is 'somewhat reductive' (Interview with Julien Meimon, December 2014) and implies that these norms still operate to form an aid 'regime' or 'set of implicit or explicit principles, norms, rules and decision-making procedures around which actors' expectations converge' (Krasner, 1983, p. 1). Clearly, in the early postcolonial decades, as Lumsdaine (1993) has demonstrated, a regime did emerge under the aegis of the World Bank, with its influential development prescriptions. This was backed by the OECD-DAC, with its peer reviews, and by around twenty 'Western' state donors, which adhered voluntarily to collective norms. However, such a regime is less relevant today given the emergence of 'new' donors operating to their own rules. Significantly, the BRICs have challenged the 'Western' development model, with the creation in 2014 of potential future rivals to the World Bank and IMF, in the form of a Shanghai-based development bank and reserve fund (Emirates, 2014). Importantly, too, the 'leading' non-DAC donor, China, does not adhere to many of the old norms, preferring loans to grant aid, non-interventionism to conditionality, tying to untying (the delinking of aid from the purchase of donor goods), low-cost contracts to strict environmental standards, and expatriate staff to local ownership. In this context, terms such as 'deviant', 'representative' and 'prototypical' are harder to pin down.

French Aid through the Comparative Looking Glass
Turning to the comparative literature, it will be argued here that, while some analyses focusing on the period after the French government aid reforms of the late 1990s do provide a more nuanced picture, most qualitative studies, particularly small-N comparisons, assume that French ODA is a deviant case and, at least by implication, not a template to follow. By contrast, variable-oriented comparative analyses consider France's assistance to be 'representative' but ranked in the bottom half of aid 'league tables'. These findings will be illustrated with reference to the structure, composition, effectiveness and function/purpose of French ODA.
To begin with comparative writings on aid structures, here qualitative small-N comparisons have highlighted the complexity and opacity of French ODA institutions and delivery mechanisms (Cunningham, 1974, p. 170). Kilby (2011Kilby ( , p. 1992) even found that France was the most fragmented donor, with double the average number of agencies involved in aid, 16 in total back in 1973. The multiplicity of France's institutions has continued to be remarked upon in more recent comparative studies but is deemed to be less pronounced (van Belle et al, 2004), particularly after the government's structural reforms in the late 1990s (Pacquement, 2010). By contrast, the singularity of France's aid structures and processes has gone largely unnoticed in large-N studies, which have found the French administration to be broadly representative, even if lagging behind other donors. To illustrate, Ghosh and Karash (2011, p. 25) ranked France 25 out of 31 in terms of the transparency of its development policy-making process. This was a low ranking, but only marginally behind Japan and the United States, and ahead of Luxembourg, Italy, the Asian Development Bank, Korea, Greece and the Inter-American Bank.
As to the nature of French aid, here qualitative, small-N studies have again emphasised deviance. Thus, Imbeau (1988) underscored the sheer size of French ODA, especially during the Cold War. Similarly, Cumming (2001) stressed France's exceptional focus on bilateral assistance to francophone Africa, while Schraeder et al (1998, p. 565) highlighted France's unique emphasis on the 'cultural imperative'. Others have compared the quality of French aid with that of other OECD donors, pointing to France's very low levels of concessionality and high levels of tying (Browne, 2006, p. 183). By contrast, comparative research with a strong statistical component has viewed French development assistance as largely representative. To illustrate, Lumsdaine (1993, pp. 79, 93) found that, while often ranked behind other donors in relation to OECD norms, France was aligning to aid regime trends over the post-colonial era. Other variable-led analyses have long painted a similar picture. Thus, Rao (1997), using a ranking system combining generosity and fairness, found French ODA to be just below average (ranking France 12th out of 17 donors) over the period 1980-92. Similarly, Knack et al (2011), using an index based on selectivity for the year 2007, rated France 25th out of 38 bilateral and multilateral donors in terms of aid quality. This latter ranking was broadly consistent with that of Roodman (2006Roodman ( , 2009, namely, 21st out of 36. As regards ODA effectiveness (taken to refer to developmental rather than cost effectiveness), the bulk of the literature in this area is not comparative or concerned with disaggregating the impact of different donors but is focused on whether aid hurts growth (Bobba and Powell, 2007), or is effective, at least when well targeted (Burnside and Dollar, 2000). There have nonetheless been comparative evaluations of effectiveness and, again, qualitative assessments have portrayed French ODA as a distinctive case. Thus, Cox et al (1997, p. 15), in a study of the management systems of six European donors, found that France was 'the most fragmented model and a priori the least appropriate for consistent and economic aid operations', with considerable aid dispersion having 'a negative impact on growth'. While the OECD and the French Parliament have recognised recent improvements in French aid effectiveness (Barrau, 2001, pp. 24-39;OECD, 2013, pp. 81-88), they have remained critical in their peer reviews and official reports. A similar stance has been adopted by Carol Lancaster (1999, p. 132), former head of the US aid agency (USAID), who concluded that 'French aid is one of the least effective in promoting development in Africa' and claimed that: In contrast to the aid of other donors, the French programme has never had a development doctrine to guide it. There is no general framework to provide a strategic focus for French aid, no means of prioritising the activities funded in the name of development, and no basis for holding aid agencies broadly accountable …'. (ibid., p. 130) While variable-oriented comparative analyses have also been critical, they have not found French aid to be a deviant case. As noted above, Knack et al (2011), drawing on the criteria defined by the Paris Declaration on Aid Effectiveness (selectivity, alignment, harmonisation and specialisation), ranked France 25th out of 38 donors, thereby placing it ahead of seven other bilateral donors (Japan, Spain, Italy, the United States, Canada, Austria, Greece). Easterly and Williamson (2011) reached similar conclusions when, with reference to best practice criteria (transparency, specialisation, selectivity, aid channels and overhead costs), they ranked France 26th out of 39 donors, ahead of eight other bilateral donors (the United States, Portugal, Sweden, Switzerland, Spain, Finland, Belgium, Greece).
As for the drivers behind French ODA, these have been found by qualitative, small-N comparisons to be unusual, even exceptional. Thus, Schraeder (1995, p. 542), using the comparative method, singled out 'cultural nationalism' as the key factor, while Lancaster (1999, p. 126) highlighted the extraordinary influence of personal friendships (cousinage et copinage) and Lloyd et al (2000) emphasised the importance of trade-related concerns in determining French aid policy. 3 These findings have been attenuated by qualitative studies that draw on statistical techniques. Thus, McKinlay (1979), using regression analysis to compare the United Kingdom, the United States and French programmes, argued that all were 'marked by foreign policy considerations', with French aid being less security-focused and less global than British and American ODA. Schraeder et al (1998, pp. 320-321), also drawing on regression techniques, found that French development assistance, although exceptionally focused on culture and indifferent to the ideological hue of recipient governments, had elements in common with American, Swedish and Japanese aid (for example, a post-Cold War emphasis on trade and limited focus on humanitarian need).
Larger-N comparisons have relied more heavily on statistical methods, including Boolean techniques and Tobit and Probit models, to demonstrate how most Northern donors, including France, have been driven by self-interest, whether in the form of strategic votes in the United Nations (Wittkopf, 1973) or trade concerns (Lloyd et al, 2000). 3 While the above studies have been firmly located within the 'realist paradigm', other statistically informed studies have advanced a more 'idealist' vision (Schraeder et al, 1998, p. 98), linking French aid to human rights concerns (Neumayer, 2003), the size of donor welfare states (Thérien and Noel, 2000) and the 'moral vision' of donors (Lumsdaine, 1993). 3 As a rule, variable-oriented analyses have portrayed the French case as representative, if at the more selfish end of the donor spectrum. These studies have recognised specificities, such as historical ties, but have contended that such factors apply broadly, if not equally, to other former colonising powers (Wittkopf, 1973, p. 887;Lebovic, 2005, p. 123). Some have even observed that 'proximity' need not be historical but may be geographic or ideological, as Japan prioritises its near neighbours while the United States gives more aid to capitalist democracies (Clist, 2011(Clist, , p. 1728. Other scholars have also downplayed the uniqueness of French motives. Thus, French aid has been found to be driven more than mostthough by no means alldonor programmes by concerns over bilateral trade Levin, 2006, p. 2044), domestic immigration and the quest for 'primary donor status' (Lebovic, 2005, p. 125). 4 In a similar vein, Berthélemy (2005, p. 17) shortlisted Francealongside Australia, Italy and the United Kingdomas one of a cluster of egotistical donors with a 'trade sensitivity parameter significantly higher than other donors'. Likewise, Harrigan andWang (2011, p. 1291) argued that, while France, much like other major donors (the United States, Canada, Italy, Japan, the United Kingdom), attaches little priority to recipient need or policy environment, it has not been the most selfish donoran honour reserved for the United States (Harrigan andWang, 2011, p. 1291).
Turning finally to the purpose of French aid, this topic is less the preserve of comparativists than of theorists and development economists seeking to establish universal truths about ODA. It is nonetheless possible to infer something from these writings about perceptions of French assistance. Thus, the fact that so many Marxist writers (Jalée, 1965;CEDETIM, 1980) have homed in on French assistance as a classic example of Western imperialism does certainly raise doubts about its developmental focus. So too does the fact that French journalists and academics (for example, Brunel, 1993;Péan, 1998) have frequently zoomed in on the neo-colonial bias of French ODA, while development economists have persistently criticised France for not doing more to address poverty and other developmental challenges (Boisdeffre, 1997;Naudet, 1997). On this latter point, Naudet (1997, p. 177) has argued that the 'priorities of French aid have been somewhat removed from the constantly changing international priorities. These are sometimes considered as "fashions" which are likely to alter or disappear as new ideas emerge'.
Whilst there is some truth in Naudet's observation, it would be wrong to overstate France's 'deviance' vis-a-vis World Bank-led development strategies. Thus, while France failed to adhere to the Bank's strategy of 'Basic Human Needs', it did increase the focus on poverty-reducing ODA in the late 1970s (Arnold, 1979, p. 55). Furthermore, although France rhetorically opposed structural adjustment in the 1980s, harbouring reservations over aid conditionality and privatisations, most French ministries ended up accepting that these neo-liberal recovery programmes were economically necessary, if politically difficult (Wilson, 1993). Similarly, while French policy-makers were initially reluctant to implement 'good governance', preferring stability over democracy promotion (Olsen, 1998) and according less importance to civil liberties than other donors (Pellicer and Wegner, 2009), 5 they subsequently embraced 'democratic governance' (Bellina et al, 2009). Finally, while France has harboured doubts about poverty reduction, as well as displaying 'low poverty sensitivity' in its ODA allocation (Clist, 2005, p. 1728), French policymakers do increasingly take 'various specific indicators of need into account when allocating aid' (Thiele et al, 2007, p. 622).
Overall, the comparative literature has clearly painted a fragmented picture. The explanation lies largely in the methodological challenges that comparativists face. The first is to reconcile the results of variable-oriented studies, which tend towards over-generalisation, struggle to generate hypotheses and ignore issues such as the 'opportunity cost' of spending on aid rather than other priorities, with the findings of qualitative, small-N analyses, which typically suffer from insufficient generalisation, are overly descriptive and attach too much importance to the hypothesis-confirming or hypothesis-infirming nature of a single case. Another challenge is to ensure comparison of 'like' with 'like'. This can be difficult in the aid context, where France's priority (francophone) African recipients tend to be poorer than, say, many of Britain's key beneficiaries in (anglophone) Africa, and where it is hard to find cases where two donors are of equal size and importance in the same sectors of a recipient country. A final methodological challenge involves overcoming the 'Galton problem'. This points to the impossibility of controlling for the effects of cultural diffusion (all donors have inevitably been influenced by globalisation). It also highlights the difficulties associated with comparing cases over protracted time periods: a comparison with, say, UK development assistance would look very different if conducted before or after the creation of the DFID in 1997.

Recognising the Trend towards Mainstreaming
It will be argued in the remainder of this article that, while these different perspectives are understandable and largely a function of different methodologies, they nonetheless misrepresent French aid by failing to capture the mainstreaming it has undergone (discussed below) and the protoypical features it has developed (discussed in some detail later) over recent years.
To begin with the mainstreaming of French aid structures, this can be traced back to Prime Minister Lionel Jospin's reforms of the late 1990s and the institutional reorganisations of the late 2000s (OECD, 2013, pp. 59-69). It began with the merging into the MFA of the Cooperation Ministry (widely viewed as the Ministry for francophone black Africa) and the subsequent loss or retirement of many Cooperation staff. It also involved the transfer of major budgetary responsibilities to a more professional agency, the AFD, which is now responsible for two-thirds of all bilateral ODA (ibid., p. 59). While these and subsequent reforms brought France's administration closer into line with the institutional arrangements of eight other OECD donors, which have a ministry with overall responsibility for policy and a separate agency for implementation (OECD, 2009, p. 30), they did not directly challenge the continued workings of the French President's advisory Africa cell, the shady dealings of the réseaux (informal linkages between French and African leaders) or the disproportionate influence of the French Finance Ministry, typically responsible for 40 per cent of the aid budget. Even so, these neo-colonial features are being eroded, as the Africa cell is now staffed by a career diplomat and a reformist AFD official, the activities of the réseaux have been curbed by media scrutiny and legal challenges, and the Finance Ministry's control over aid decision making has been reduced by the AFD's financial autonomy: it controls over 86 per cent of its own resources (Cambon and Vantomme, 2011, p. 45).
In tandem with the above reforms, France has carved out a better image for itself in development circles by setting up new coordination mechanisms: an inter-ministerial committee (Comité Finally, France has reassured fellow donors by ensuring that its policy-making process is more open. While the French aid administration is still not highly ranked in terms of transparency, 6 it has become more open to the ideas of civil society, not least via broad-based advisory bodies, such as the Haut Conseil de la Cooperational Internationale (1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008) and the Conseil National pour le Développement et la Solidarité Internationale, created in 2013. The policymaking process is, moreover, now more accountable thanks to the submission of a biennial report to parliament on the implementation of the Framework Document (MFA/ Finance Ministry, 2012) and to the passing, in May 2014, of an Act to allow closer parliamentary scrutiny of aid.
As to the composition of French ODA, here mainstreaming has taken place over a longer period. Thus, French aid, which averaged 1.3 per cent of GNP between 1960and 1964(Bossuat, 2003, p. 440) and 0.67 per cent between 1989and 1990(OECD, 2004, is no longer exceptionally large. 6 It fell to 0.32 per cent of GNP in 2000 and only returned to 0.50 per cent in 2010still well short of the 0.7 per cent target achieved by five OECD donorsthanks to massive debt cancellation and the inclusion of dubious expenditures, such as refugee costs and the tuition fees of developing country students studying in France. There has, equally, been a gradual erosion of other distinctive features of French development assistance. Thus, while French aid remains overwhelmingly concentrated on a single continent, Africa, the focus isin line with World Bank thinkingnow more selective, with France's Framework Document targeting 50 per cent of bilateral ODA on (initially 14 now) 17 mainly francophone African countries, 20 per cent on Mediterranean countries, 10 per cent on fragile states and up to 10 per cent on emerging economies. There has, moreover, been a drastic decline in the number of technical assistants (mainly teachers of French) from 20 000 at the start of the 1990s to 2200 some 20 years later (OECD, 2013, p. 15), with the focus shifting to shorter demand-led capacity-building missions (OECD, 2004, p. 71).
Finally, while France continues to channel budgetary support through the Franc Zone, a unique currency support mechanism for 14 African countries, the devaluation of the African (CFA) franc in 1994 demonstrated that France's support was not immune to market pressures.
Similar observations can be made with regard to the development effectiveness of the French ODA programme, which can no longer be singled out for criticism in terms of its wastefulness (Brunel, 1993), project failure rate (Péan, 1988), levels of tying (OECD, 1994, p. 19) or weak culture of evaluation (Lancaster, 1999). While recent reports by the Cour des Comptes (2012) and the MFA/AFD (2013) suggest that France still has room for improvement, French practice has begun to align to norms established at recent High-Level Aid Effectiveness forums in Paris, Accra and Busan. In particular, the AFD has integrated a 'results-based' focus into its aid contracts (OECD, 2013, p. 81) and embraced the concept of donor 'harmonisation', not least through its commitment to the European Code of Conduct, joint EU programming and pooled multilateral evaluations. It has also supported 'ownership' and 'alignment to recipient country efforts' by issuing aid tenders through African ministries, respecting local management systems and supporting triangular cooperation, involving DAC donors, South-South partners and beneficiary/third countries (OECD, 2009, p. 52). Equally, the AFD has recognised the importance of 'transparency', with some evaluations being filmed and most published (AFD, 2013a, p. 26). Since 2002, it has also stepped up its budget for evaluations and now evaluates one in three projects (Cambon and Vantomme, 2011, p. 78). It has, moreover, increased the number of expert consultants employed, associated aid recipients with evaluations and included local authorities in evaluation steering groups (OECD, 2004, p. 63).
Turning lastly to the purpose of ODA, there is of course no donor consensus as to what this should be. In France's case, this issue is clouded further by disparities between rhetoric and practice and by a tendency to proliferate aid policy goals (OECD, 2004, p. 20). Even so, it is clear that the French administration has aligned to the current development orthodoxy, namely, poverty reduction. Thus, France has signed up to the UN Millennium Development Goals (MDGs), improved its ranking, between 2008 and 2012, in the Commitment to Development index from 16th to 13th out of 27 (ibid., p. 26), and increased spending substantially on primary education and health (ibid., p. 53). France's commitment has, however, remained partial, as it has continued to allocate large sums to the promotion of French culture (€1.3 billion in 2002: see OECD, 2004, p. 31), while providing a lower share of ODA (86.2 per cent) in grant form and channelling less aid through NGOsgenerally assumed to be effective in tackling povertythan other established OECD donors (OECD, 2013, p. 114).
To sum up, French aid has clearly become less 'deviant' and more 'representative'. One reason for this 'mainstreaming' is the fact that France operates in an increasingly interdependent world and cannot escape pressures such as globalisation, which have opened up Africa to other donors, making it harder for France to maintain privileged aid relations. Another factor has been 'Europeanisation', which hasthrough the creation of the EU's Common Foreign and Security policy, Consensus on Development and Policy on Coherence for Developmentpushed member states to cooperate more effectively and made it problematic for French policy-makers to continue ploughing their own furrow on development policy. The pressure on France to align to international aid norms has, moreover, been stepped up by bodies such as the United Nations, with the MDGs, and the OECD, via its High-Level Forums. Another driver has been the new generation of self-confident African leaders, such as Rwanda's Paul Kagame, who have sought to replace neo-colonial ties with more diverse relations, particularly with emerging economies.
Domestic factors have also militated in favour of 'normalisation'. These include the election of reformist politicians and officials, such as Jospin and Severino, who were prepared to break with the 'Gaullist Consensus', whereby ODA was concentrated heavily on francophone Africa (Interview with President Sarkozy's cabinet adviser, May 2012). These modernisers were also acutely aware of budgetary pressures arising from the EU Stability Pact, from spending reviews (for example, the Révision Générale des Politiques) and from the global recession.

Zooming in on French Aid: A Prototypical Case?
While French aid has clearly undergone mainstreaming, can it also be said to have developed prototypical features in terms of its structures, composition, effectiveness and purpose? To answer this question, this study uses our conceptual framework to assess the actual, potential and possible uptake of French ideas.

A Flexible 'Business Model'
To begin with aid structures, the AFD in particular has developed institutional features that fellow donors would like to replicate or learn from. The first is its capacity to respond to diverse challenges. The AFD has used its status as a public-private, semi-autonomus body to 'combine the role of banker, development agency, consultancy and technical assistant'. This has given it 'considerable flexibility' and 'a vast range of instruments with which to respond to the challenges of developing countries' (Cambon and Vantomme, 2011, p. 36). It has used this flexibility to step up its activities fivefold over the last 10 years and increase its staff by 35 per cent over the last 6 years (OECD, 2013, pp. 64, 19).
Second, the AFD offers a viable business model. It turns an annual profit of €200 million, which part-subsidises 'loss-making' grant-aid activities (ibid., pp. 36, 84). Such a model is attractive to some donors at times of budgetary restraint. So too is the AFD's championing of the socio-developmental benefits offered by companies and private investment. The fact that the OECD now talks openly of 'the nascent transition towards private developing financing' (Chang et al, 1999, p. 97) suggests that the AFD, an agency with considerable autonomy to raise money on the open market and with no direct equivalent among 'Western' donors (Cambon and Vantomme, 2011, p. 32), may be uniquely placed to shape donor thinking in the post-MDG era. 7 Finally, the AFD has developed the capacity to generate cutting-edge field-based research. It has 'established itself in international development debates thanks to its research work […] active communications strategy … [and] recruitment of new expertise'. It is now an intellectual powerhouse with 71 overseas missions, as well as access to expertise from 13 regional offices linked to its private sector arm Proparco; 163 French embassies and aid missions; state-funded specialist research institutes, such as Office de la Recherche Scientifique et Technique Outre-Mer (ORSTOM); and academic networks, such as GEMDEV (development economists in Paris-based universities) (OECD, 2013, p. 61).
It is for these reasons, and despite criticism that the AFD is overstretched and prioritising emerging countries over Africa, that other donors have been keen to tap into AFD expertise. In this context, Germany's KfW and the European Investment Bank (BEI) have been working with the AFD since 2010 to ensure mutual recognition of their procedures. Other donors are tentatively exploring the viability of establishing a development bank (Guardian.com, 2013) and replicating the AFD model. The visit by 10 British MPs to the AFD in 2014 explicitly addressed this issue (Interview in Paris, May 2014). In 2015, Sweden also sent a mission to learn about the AFD's use of credit lines, while the OECD and World Bank both requested AFD briefings on blending (the combining of aid grants with loans) ahead of the 2015 Addis International Financing Conference (Interview with AFD, July 2015).

'Beyond-Aid' Financing
As to the nature of aid itself, here French policy-makers have played a pioneering role in redefining ODA; advocating greater use of 'blending' and other forms of concessional lending; and focusing donor attention on innovative financing. Perhaps unsurprisingly in a context marked by poverty reduction targets, there is evidence of actual donor uptake in all these areas, except for concessional lending where there is potential for future uptake.
On the definitional point, French aid practitioners Severino and Ray (2012)building on the work of Kaul et al (1999) and the United Nations Development Programmepressed the view that ODA was no longer fit for purpose and should be replaced by a measure that takes account of the real cost of contributing to the management of global public goods, such as the world's climate, biodiversity and global health. While this idea was 'insufficiently followed up' after Severino ceased to be AFD Director, it did strike a chord with other donors struggling to meet their aid pledges and fed into the DAC's decision, at its December 2014 High-level meeting, to include 'contributions to global public goods' (OECD-DAC, 2014) in its more expansive measurement of ODA, known provisionally as 'Total Official Support for (Sustainable) Development' (Interview with AFD, July 2015).
In the case of blending, the AFD has joined forces with Germany's development bank, the KfW, to champion the use of this controversial instrument. Arguing that grant aid should be used to leverage loans, which are quicker, better at promoting self-reliance and larger than grants, the AFD has helped to ensure a 'silent revolution' (Interview in Brussels, 2014), whereby the hitherto grant-focused Commission has now embraced this 'radically different modality', channelling 2 billion of its €56 billion budget to blending projects and opening up national as well as regional indicative programme monies for these operations (Interview with AFD, June 2014). It also pushed successfully for the creation of forums, such as the EU's Mutual Reliance Initiative, to help finance blending; regional funds, such as the Africa-EU Infrastructure Trust Fund; and a multi-sectorial regional facility on blending (Interview with AFD, July 2015).
As regards concessional lending, France has been, after Japan, the main exponent of these loans, which have a grant element of at least 25 per cent. While concessional loans can contribute to unsustainable debt levels, they are nonetheless recognised by the OECD-DAC (2014) as 'important instruments to finance long-term investments in economic and social infrastructure'. As one UK official told the author, even a grant-only donor like the United Kingdom is, in its International Development Committee reports, 'beginning to take an interest in the French/ German models of concessional lendingrecognising that loans can be helpful in some circumstances, particularly in the case of global public goods', and more specifically, 'climate change where … most of the financing requirements for mitigation (reducing emissions) are in emerging countries'.
Turning to innovative financing, here France's advocacy role has been less controversial. As a founder member in 2004 of the 'Initiative Against Hunger and Poverty', Francetogether with Chile, Spain and Brazildrew up the original definition of 'innovative financing', stressing that resources should be 'new', 'stable' and generated by those industries (finance, transport, communications) that have benefited most from globalisation (Interview with MFA, December 2014). The MFA subsequently assumed the permanent secretariat of the Leading Group on Innovative Financing and focused on the need for 'globalisation to give something back', as well as placing innovative financing on the agenda of the G20, a body that 'had not previously discussed international development' (ibid.). Importantly too, France has consistently supported innovative financing schemes, such as the UK-led International Finance Facility for Immunization, while leading on others, such as the Financial Transaction Tax (FTT). French policymakers ensured that the FTT was 'politely discussed by the G20' in 2011 (Interview with ONE, May 2012) and helped to secure a commitment by 10 EU states to introduce the tax by 2016 (Guardian, 2014). While the benefits of the FTT to overseas development have fallen short of expectations (€130-160 million annually; interview with UK official, July 2015), the same is not true of the 'solidarity tax' on air travel, which France also championed. This levy has won support from several OECD states, raised over a billion dollars since 2006 and enabled the development of new products, such as antiretrovirals for children (MFA, 2015).

An Eclectic Approach to Effectiveness
As regards development effectiveness, this is another area where France has provided or could provide leadership, whether through its field expertise, innovative instruments or evaluation techniques. On the first two issues, there has been actual uptake or at least movement in the direction of French thinking. On the latter, there is potential for future uptake.
To begin with expertise, a good example is the land tenure issue. With experience of establishing cartographies of land tenure systems across francophone Africa, setting up land observatory missions in places like Mali in 1991 and monitoring food security across the Sahel (OECD, 2013, p. 28), French officials were well placed to take over the secretariat of the Technical Committee on Land and Development in 2005. In so doing, they oversaw the production of key publications, including a 2009 White Paper (Le livre blanc sur les politiques foncières), a 2010 strategy paper on land-grabbing (Les appropriations de terres à grande échelle) and analytical guides to help operationalise these ideas. Crucially, these publications then formed the basis of voluntary guidelines on land settlement issues adopted by the French government, the EU, the Food and Agriculture Organisation and the Global Donor Platform (Interview with AFD, January 2015). This marked a significant shift in donor thinking towards the long-held French view that African customs and rights were the key to settling land-related questions and away from the World Bank orthodoxy, namely ,'titling'the delivery of individual private property titleswhich has exacerbated some disputes, not least between African farmers and pastoralists (Hall et al, 2012, pp. 34-43). As land law specialist Ambreena Manj (personal communication, June 2015) told the author, 'The French have taken a more progressive line on customary tenure and community land, though I am less sure of the process by which this has come to influence DfID and the World Bank, as I think it has' (Interview, June 2015). The same expert cautioned, however, that 'the Bank's recent embrace of customary tenure may be only skin-deep and its … assumption that land relations … must, evolve from communal to individual title remains unchanged, even if it is now more nuanced'.
As regards innovative instruments, one such mechanism was the AFD's tool, elaborated in the mid-2000s, for measuring carbon dioxide emissions. This instrument was made publicly available via the AFD website and was adopted by the International Finance Corporation, the private sector arm of the World Bank. The IFC acknowledged AFD's 'groundbreaking work' on its website (IFC, undated) and is said to have 'completely changed its procedures for measuring tons of carbon emitted' (Interview with AFD, December 2014). The CEET's success 'revealed how big announcements of extra money can be less valuable than concrete tools that allow developing countries to measure pollution levels and climate change' (ibid.).
Finally, with regard to evaluation procedures, France has been pressing both for greater rigour and a more holistic approach. In line with demands for rigour, Severino and Ray (2012) called for international harmonisation of evaluation norms, as well as verifiable and independent evaluations by public and private sector donors. In the same spirit, the AFD integrated its evaluation division into its research department and invested in costly cutting-edge impact assessments, whose findings are hard to contest empirically (Interview with AFD, December 2014).
To illustrate, in 2006, the AFD undertook a major study that provided a more nuanced understanding of microcredit projects, showing how these can reduce poverty when combined with professional training, improved savings schemes and structural interventions in the wider economy. Equally, the AFD conducted a large-scale evaluation of voluntary health insurance in Cambodia (2007Cambodia ( -2011 and demonstrated how such schemes only cut poverty if they constitute the first step towards a wider system of social security cover (AFD, 2013c).
The above approach contrasts with, but also complements, France's innovative attempts to make evaluations more holistic via the creation of the Fonds pour la promotion des Etudes préalables, des Etudes transversales et des Evaluations (F3E) in 1994. Financed jointly by the French government and NGOs, this unit undertakes evaluations which consider non-measurable factors (such as shared lessons and experiences, feelings of wellbeing, empowerment or solidarity between peoples) when assessing projects that metrics-based assessments might deem failures. While such 'soft' variables are unlikely to gain traction with technocratically minded donors, they do provide privileged access to the views of CSOs, as well as offering a new way of framing aid, which could be important in maintaining public support, foregrounding the Southern perspective on development and persuading emerging donors to undertake more evaluations (Interview with AFD, 2012). Significantly, too, the F3E's methodology is currently being considered by the Wales for Africa team within the Welsh government as well as the Wales for Africa Hub, that is, the confederation of Welsh NGOs (Interviews July 2015), as part of their wider reviews of monitoring and evaluation.

Advocating a New Role for Aid
Turning finally to the purpose of aid, France has traditionally been more of a reluctant follower than a pioneering setter of international development nostra. While France did seek to reframe World Bank-led strategies over the early post-colonial decades, its influence was at best limited. The difference now is that France has actually signed up to the current poverty reduction orthodoxy and, as a donor that has undergone considerable mainstreaming, finds itself well placed to challenge this development nostrum from the 'inside'. In this context, French policymakers have labelled the MDGs as 'technocratic' and 'utopian' (Interview with MFA, May 2012). They have equally drawn up influential strategy papers (for example, MFA, 2001) challenging the idea that poverty is purely a technical problem subject to technical solutions and arguing that it is also a socio-political issue, requiring socio-political solutions that include better regulation of globalisation, improving poor people's access to productive resources and wealth redistribution. French policy-makers have also argued that, rather than grand announcements on ODA, the emphasis should be on inclusive, shared and sustainable growth, as this is what will enable developing countries to tackle poverty themselves. France's reasoning chimes in well with the current donor emphasis on Sustainable Development Goals and has been praised by the OECD, even if French aid structures may not be geared towards delivering this inclusive approach to poverty reduction (OECD, 2004, p. 11).
To begin with mainstreaming, this process helped give France greater credibility as a donor. It addressed some of the persistent criticisms of French aid, notably its perceived ineffectiveness and neo-colonial bias, removing obstacles to a French leadership role and paving the way for greater cooperation with other donors: the AFD, for example, signed a Memorandum of Understanding and Joint Action Plan with the DFID in December 2010. Conversely, the fact that France, unsurprisingly given its history of exceptionalism, has remained ambivalent towards many international development norms has also given it the freedom to push in directions that other donors, such as the Nordics and the DFID, might not, due to their overriding focus on poverty reduction. French policy-makers have not been deterred by CSO attacks (CONCORD, 2014) from taking the lead on 'beyond-aid' financing, taking advantage of the fact that many donors are keen both to generate more ODA-like flows and to use concessional resources to tap into emerging markets. According to one UK official (Interview, July 2015), this ambivalence towards aid norms has helped ensure that 'France has been at the forefront of moving from ODA to development financing. While this has negative implications in terms of (for cynics) being used as a smokescreen for falling ODA levels, it also means that resource-generating mechanisms such as the Financial Transactions Tax … are carried by France'.
As to the more strategic approach by French policy-makers, this has taken many forms. The first involves a better communications strategy. Under Severino's directorship, the AFD appointed young economists and other new staff with fresh ideas and began preparing strategy papers to ensure that France's 'music [was] heard' in international development circles (Interview with ONE, May 2012). The second concerns institutional innovation: the MFA established, in 2008, a Unit for Partnerships and Innovative Financing (Interview with MFA, December 2014), whose creation 'opened up the French aid system to new ideas' from researchers, foundations, think-tanks and non-DAC donors (ibid.). Third, France has drawn on its eclectic range of partnerships with bilateral donors (for example, Germany's GTZ and KfW), local government bodies and CSOs to press different ideas in different forums. The jointly prepared AFD/European Commission briefing note on blending (entitled 'Financing for Sustainable Development') delivered to the 2015 Addis Conference is a case in point. Fourth, French policy-makers have shown an appetite for sharing ideas, as demonstrated by the CEET and their dissemination of impact assessments. Fifth, they have been prepared to bide their time until ideas (such as food security and customary land tenure) gain recognition. In so doing, they have recognised that 'good ideas often spend a long time in thinktanks' and usually need some kind of champion (Interview with DFID, January 2010). Significantly, too, French policymakers have been good at anticipating challenges in international development, investing €12 billion between 2005 and 2014 in combating climate change (AFD, 2013b, p. 7). In so doing, they recognised that 'development financing must be able to respond to key challenges which are beyond the traditional ODA remit (for example, education, health) in the most appropriate way' and that ODA was 'not the right vehicle to finance the massive cost of energy transition' (Interview with UK official, July 2015).
Turning finally to the intellectual climate, here several developments have created space for French ideas. First, the early optimism surrounding the MDGs gave way to a realisation that these goals were unlikely to be met even if ODA were scaled up, with all the demands that would impose on recipients in terms of absorptive capacity and donors in terms of spending power. Faced with this financing gap, the OECD-DAC (2014) recognised that 'a wide array of domestic and international resourcesboth concessional and commercial in natureneed[ed] to be mobilised from public and private sources and from all providers'. In this permissive context, private sector investment, concessional lending and 'beyond-aid' financing generallyall areas of French expertisehave been more warmly welcomed. So too has blending, which was hailed by the EU in 2015 as a mechanism that would leverage €100 billion for sustainable development by 2020 (Mogherini, 2015).
The second development is the increasing diversity of the development challenge. As Alonso (2014) has argued, 'aid was born as a response to a world that was characterised by a deep North-South divide … [However,] the North-South divide has given way to a more diverse … world where … there is a much wider and graduated spectrum of levels of development between countries' and where the location of extreme poverty 'has also changed, with the bulk of the world's groups of poor … found in middle-income countries' (ibid.). Against this background, there is a need for a differentiated approach to development, along the lines advocated by France's Framework Document, with grants focused on the poorest countries and concessional loans on emerging economies. Such an approach is all the more necessary now that some African countries (for example, Ghana, Uganda) are eligible for concessional lending and have, thanks to substantial growth during the 2000s, established targets for reducing their aid dependence (Guardian.com, 2011).
Finally, there has been a growing recognition of the importance of emerging donors as a source not only of concessional flows, now estimated at between US$11 and 41.7 billion (Aiddata, 2015), but also of new ideas: the OECD High-Level Forum in Busan welcomed South-South cooperation, established the Global Partnership for Effective Development Co-operation, and recognised that non-DAC donors could offer innovation and first-hand experience of being a recipient. These South-South donors have also come to be seen as a reservoir of potential partners, via triangular cooperation and information exchanges: France is, for example, helping one such partner, Thailand, to establish its aid agency (AFD, 2014). Finally, they have constituted a potential audience for French ideas such as blending that do not sit comfortably alongside existing OECD norms.

Conclusion
This article has challenged comparative perspectives on French aid that portray it either as deviant or as a representative case, ranked in the bottom half of ODA league tables. While these misunderstandings are to be expected, they miss out on the recent mainstreaming of the French programme and its development of prototypical features, such as its flexible business model and its pioneering role on 'beyond-ODA' financing. This prototypical dimension is assessed in terms of 'uptake' and is explained in terms of France's continuing ambivalence towards international aid norms, its more strategic approach to promoting French ideas and an intellectual climate that is more receptive to those ideas.
The above observations bring us back to our original questions about the nature of the French case. A few conclusions can be drawn here. First, French ODA clearly has elements of a deviant, representative and prototypical case. Second, while the French case has not shaken off its 'deviance' altogether and will not as long as cultural imperatives remain factors in France's ODA decision-making process, it is nonetheless increasingly marked by 'representative' features, such as the norms agreed at High-Level OECD forums. Finally, the French case does have a prototypical dimension and has scope to develop this further as the AFD strives to become a 'global reference point' for the donor community (AFD, 2002, p. 6). Ultimately, however, donor uptake of French ideas is always likely to be selective rather than systematic in an aid context where so many competing norms are being championed by OECD and non-DAC donors alike.
These findings open out on to wider questions. First, given that French ODA is aligning to international trends, are we heading towards a more homogenous donor community, at least in the case of OECD states? Or is this harmonisation superficial, barely concealing cleavages between Nordic-Plus donors (particularly the Nordics, Holland and Britain), who are attached to the 0.7 per cent aid target, and many other DAC donors, who are not? In either case, how do non-DAC donors fit into this picture? Given that they are not bound by OECD norms and have their own approach to international development, could they now turn the tables on the DAC and begin shaping Northern donor practices? The recognition that South-South cooperation enjoyed at Busan has paved the way for greater influence.
Second, given that French assistance has retained deviant features, is it not reasonable to assume that there are other outlier cases among OECD donors? Is such deviance problematic from the perspective of other donors (particularly where common positions are undermined) and recipients (whose absorptive capacity is stretched further), or might it benefit the former (by ensuring aid diversity) and the latter (by allowing greater scope for negotiation)? Clearly, the fact that most South-South cooperation does not conform to OECD 'rules' yet is welcomed by developing countries does suggest that deviance need not be a problem.
Finally, how does a donor develop prototypical features? Our conceptual framework, with its focus on uptake, is a useful starting point and could be applied equally to established donors, to emerging aid programmes or to particular sectors in which donors have taken the lead, say, in the context of EU joint programming schemes. Clearly a donor must have features that others will want to copy or learn from, because they are innovative, evidence-based and/or well suited to a changing aid context. Some ideas are unlikely to enjoy wider uptake, particularly where they undermine moves towards greater technical rigour (as do softer forms of evaluation), are extremely complex (as are, for example, French bilateral debt cancellation schemes) or demand an overhaul of the entire ODA architecture (as do some more radical interpretations of 'global public goods': see Gabas and Hugon, 2001). Other initiatives, such as the FTT, sit uncomfortably with the neoliberal view of development held by the World Bank and liberal donors. The schemes that are likely to gain traction are those that remain within the free-market framework (as do AFD private sector development projects), that offer new or better ways of freeing up resources (as do some innovative financing programmes), that are championed by respected institutions (the IATA was, for example, established following DFID pressure) or that bring less established donors into the fold (as does the AFD's eclectic approach to donor partnerships). The donor community really does need to invest more in approaches of this kind if it is going to help lift 'a substantial amount of the world' out of 'extreme poverty and stagnant growth' (Easterly andWilliamson, 2011, p. 1930).