Abstract
This article provides a case study in the new and developing field of inter-organisational relations by looking at the position of the European Union (EU) within the Group of Eight (G8). It analyses the questions of why inter-organisational relations start, and how they develop over time. I argue that inter-organisational relations are determined by the dynamics within the international organisations that are involved: both the characteristics that define these organisations, as well as the preferences of their constituent parts. The G8-EU relation was a by-product of European integration resulting from the transfer of competences, but was initially thwarted by the strong and divided preferences within the EU. While the G8’s informality at first facilitated this internal division, it also enabled this relation to change over the following decades. This change cannot only be ascribed to rationalist and social-institutionalist factors but also testifies to a logic of path dependence.
Introduction
The Group of Seven/Group of Eight (G7/G8), 1 a summit consisting of eight of the world’s most influential state leaders, has taken place annually since 1975, remaining relevant even after the establishment of the G20 in 2009. 2 The European Union (EU) has taken part in this gathering since 1977. This makes for an interesting combination, with the world’s most elaborate supranational organisation, the EU, joining the negotiating table of the G8, an informal concert that can hardly be considered an institution at all. While this case has previously been examined from a G8 perspective, 3 as well as from an EU perspective, 4 few studies thus far have examined the dynamic and informal relation between the two international fora.
This relation constitutes a so-called dyadic inter-organisational relationship (IOR) between the EU and the G8. More specifically, it is a case of ‘nesting’, which occurs when a regional or issue-specific international organisation (IO) takes part in another. 5 While research on IOs in the field of International Relations has become more common since the 1980s, 6 most enquiries tend to focus on individual organisations and neglect the possible relationships between them. 7 Recently, however, the subject has drawn more scholarly attention, for instance, from Jönsson, 8 Young, 9 Biermann 10 and Brosig, 11 who have sought to develop theoretical building blocks that are to guide empirical analysis. Such theorising has primarily revolved around the questions of why inter-organisational cooperation takes place (and why not) and how possible successes and failures can be explained. Empirically, the EU increasingly serves as the main object of interest, due to its unique supranational characteristics and its extensive network of inter-organisational relations, including numerous instances of nesting, for example, in the World Trade Organization (WTO), the United Nations (UN) or the Food and Agriculture Organization (FAO). 12 Despite these efforts, both theoretical and empirical contributions on inter-governmental IORs as well as the challenges of nesting specifically remain sparse.
This article adds to this debate on both counts, offering an empirical case study that offers new theoretical suggestions regarding the circumstances under which inter-organisational relations can be successful. More specifically, I present a case in line with Biermann’s first and most basic explorative question ‘how and why does inter-organisational networking start?’. 13 To this basic question, I add an exploration of how inter-organisational relations develop over time, as well as potential explanations for such development. I argue that informal IORs offer the opportunity for flexibility and change according to the (possibly divided) preferences of the IOs that are involved while they also attest to a logic of path dependence.
Two characteristics make the case of EU participation in the G8 particularly interesting, both for empirical and theoretical reasons. First is the fact that the EU is a participant in the G8 while some of its own member states, France, Germany, Italy and the United Kingdom, also take part in the forum. This presents us with a case of double representation, and compels us to depart from the traditional approach of IOs as black boxes. 14 Instead, opening these black boxes enables us to understand the individual preferences and strategies of member states of both organisations, explaining the opportunities as well as challenges that these present.
The second characteristic that justifies a closer look at this particular case of inter-organisational relations is the fact that it is a relationship between two extremes: a global informal concert on the one hand, and an advanced supranational regional organisation on the other. Especially, concert governance remains underexposed in IOR research so far, which is unjustified, given its high potential for inter-organisational cooperation. The particularities of these two organisations offer an important explanation for the relationship that developed between them, suggesting that proper understanding of the dynamics of the organisations of interest, as well as the interaction between them, is necessary for us to understand international IORs. Indeed, the very fact that the two organisations are so different seems to be the foundation of their long and stable relationship. That said, I also argue that the informality of the summit has accommodated the ambiguity of the international representation of the EU. Thus, the natures of both institutions help to explain why, despite the longevity of the IOR, it has remained a superficial and informal cooperation, that lacks balance because it is primarily shaped by the EU. Moreover, this explanation is complemented by a reasoning of path dependency, which depicts how the two IOs were locked into a relation in a nesting-format with double representation.
I proceed as follows: first, I offer a brief conceptual introduction of inter-organisational cooperation that is to guide the case analysis. Next, I describe and explain the nature of the G8 as an instance of concert diplomacy, as the characteristics of such informal global governance set the parameters of this inter-organisational relation. I then turn to the basic question of how and why the EU obtained a seat at the summit table, after which I continue with an analysis of the development of this IOR since. 15
Understanding inter-organisational cooperation
Inter-organisational relations are links, relationships and modes of interaction between two or more inter-governmental organisations (IGOs), non-governmental organisations (NGOs) or public private partnerships (PPPs). 16 This article exclusively focuses on relations between IGOs, a variety of IOR that has up until recently received limited attention. 17 Inter-organisational relations differ in their degree of formality, depth, longevity and maturity, all of which have an effect on the strength and effectiveness of the relationship. Nesting constitutes a more formalised variety of IO cooperation and occurs when a more specific institution, in terms of region or issue, is a participant in a broader multilateral framework. 18 The supranational characteristics of the EU make it an important example of nesting because it does not only cooperate with other IOs but takes a seat in many of them as well. 19
Although not an entirely new phenomenon, inter-organisational cooperation in international relations has only really taken off since the fall of the Berlin Wall, since when the links between IGOs more than doubled. 20 This development followed the proliferation of IOs, which created a situation of ‘increased actor density, dispersed competences and organisational overlap’. 21 Specialisation and inter-organisational cooperation were ways to deal with the adaptation pressures arising from this situation.
Despite this favourable environment, IORs face an important obstacle. Gouldner 22 notes that organisations have a strong tendency to maintain their autonomy, primarily to avoid losing the freedom to act independently. This so-called ‘stovepiping’ 23 is amplified by self-centred organisational cultures, inter-organisational rivalry and the costs of investing resources in the development and maintenance of an IOR. Adding to this obstacle is the heterarchical nature of the international realm, making cooperation mostly voluntary, as opposed to mandated. 24 Van de Ven thus concludes that ‘an agency prefers not to become involved in an IOR unless it is compelled to do so’. 25
Different suggestions have been made about the factors that help to overcome this fundamental obstacle of stovepiping. From a traditional rationalist point of view, it is suggested that organisations will only sacrifice autonomy when it is in their interest. If both the organisations are committed to the same problem or opportunity, and their resources to tackle these are in fact augmented through inter-organisational cooperation, then a cost–benefit analysis should support the decision to cooperate. 26 According to this explanation, variation is a necessity: ‘(o)nly if an institution can provide, for example, new concepts, expertise or funding which is lacking in another, can the exchange of resources be set in motion’. 27 Moreover, the exchange needs to be equal, providing incentives for both organisations to interact. Resource dependency can also result in defensive cooperation, meaning to prevent encroachment or the loss of resources. 28
This rationalist perspective has long since been complemented by social and cultural explanations for cooperation. These can either reinforce each other or work in opposite directions, which either way shapes the nature of the relationship. Factors such as normative and cultural overlap may mitigate or overcome stovepiping. Aside from organisational affinity, personal relations also matter in this respect, as does individual leadership. 29 Moreover, organisations may be motivated by a desire to enhance their own legitimacy, reputation or visibility through inter-organisational cooperation. 30 A final social factor that has only recently come to academic attention is the (overlapping) preferences of the member states of both IGOs. 31
These rational and social factors are conditioned by several constitutive elements that can either stimulate or impede inter-organisational cooperation by altering preferences. Such elements include increasing issue density and durability, 32 external and internal shocks, such as international conflict or the eurocrisis, and systemic changes, like the end of the Cold War. 33 These developments can alter the urgency of the problem or opportunity that the organisation is committed to, as well as the resources that are available to organisations.
If the rational and/or social circumstances are favourable and IOs decide to engage in inter-organisational cooperation, progress is usually quite slow and incremental. 34 Cooperation generally starts on an ad hoc and informal basis, usually initially only involving single issue cooperation. 35 If previous cooperation proves successful, and the circumstances remain favourable, organisations may decide to continue and/or deepen their interaction. However, once again, we need to move beyond rationalist explanations, and consider the relevance of inter-organisational cooperation as a socialisation process. 36 The following case therefore means to show that the costs and benefits of inter-organisational cooperation, as well as its specific features, are shaped through social interaction both within and between IOs.
The G8 as modern concert diplomacy
The G8 constitutes a particular form of IO that is often overlooked. According to scholars like Hodges 37 and Baker, 38 its informal nature and aims justify the conclusion that it does not fit the prerequisites of an IO. They use a narrow interpretation of the term ‘IO’, focusing disproportionally on tangible and formal characteristics, such as a secretariat or a treaty. However, adopting a more flexible approach on the extent of formal institutionalisation needed in order for cooperation to qualify as an IO enables us to examine the social relations that signify IOs, as opposed to their formal and legal manifestations only. 39 I use this case to examine these dynamic social processes and their effect on inter-organisational cooperation more closely.
Despite its ad hoc nature, the history of the summit dates back to 1975, a fact that in itself already testifies to the summit’s relevance. The first meeting was called by French President Valery Giscard d’Estaing, who invited the leaders of five major Western democracies, the United States, Japan, West-Germany, the United Kingdom and Italy. The decision to assemble at the highest diplomatic level was fuelled by several interrelated crises that transpired in the 1970s, such as the global recession, the energy crisis and a variety of security threats. These challenges drew attention to the declining hegemony of the United States as well as increasing transnational interdependence. At the same time, there was a need for more efficient international cooperation platforms, due to rising frustration with the bureaucratisation and fragmentation of traditional IOs, such as the General Agreement on Tariffs and Trade (GATT), the International Monetary Fund (IMF) and the UN. 40 It was this frustration that would give rise to the principal defining feature of the summit in the decades to come: informality.
From the start, the summit was not meant to be a forum for grand policy solutions for world problems, but a meeting place for global leaders to create broad consensus and compatible approaches to economic challenges. Initiator Giscard d’Estaing wanted the summit to be unstructured, small, select and personal, and wished to limit membership to countries with international weight and influence. Also, he envisaged summits as ad hoc conferences rather than necessarily as annual occurrences. Personal representatives of the state leaders, so-called sherpas, would gather before the actual summit to set out the conditions of the summit. 41 The camaraderie, frankness and solidarity that were aimed for at the state leader level was most significantly accomplished at this sherpa level. 42
In the years that followed, the summit saw little significant change, apart from a broadening of the agenda to include political issues in the 1980s, and the admission of Russia as a full summit member in 1998. 43 The summit has mostly remained devoid of formal institutionalisation, such as a treaty, a secretariat or policy tools that enable it to act in its own right. Nevertheless, the forum has developed other institutions, of an informal and social variety. Examples of such institutions include media outreach, endorsements, global network formation and leadership. 44
The unstructured nature of this IO has led scholars to propose new classifications, such as ‘meta-institution’, 45 ‘quasi-intergovernmental organisation’ or ‘informal group’. 46 In an interesting attempt to revive classic concert theory, Kirton argues that the G8 can be conceived of as ‘a modern equivalent and expression of an international concert, of the sort that great powers in multipolar systems have occasionally formed during the past two centuries’. 47 These classic concerts occurred after major wars were fought to contain a potential hegemon and in situations in which the original international balance of power was disrupted. 48 Similarly, the G8 initiative was a response to threats to the prevailing order, namely, the crises of the early 1970s. Concerts have specific characteristics with regard to purpose, membership and rules of interaction. In terms of purpose, concerts intend to gather and exchange information and views, thereby reducing the transaction costs of international relations. Concerning membership, similar to the classic concerts, the G8 incorporated all, and only, the great powers in the interdependent system, ensuring that there would be a predominance of power in comparison to those who were left out. 49 Finally, the rules of interaction of a concert constitute informal summit diplomacy to ensure ‘repeated games’ and to prevent defection. This characteristic is most clearly fulfilled in the case of the G8. In fact, the G8 might be the best example of such summit diplomacy to date.
As international cooperation generally became increasingly institutionalised during the second half of the twentieth century, concerts became the exception in their function as forums for non-rule-guided international negotiation. 50 This limited institutionalisation has become a central feature and the main added value of concerts. Schwegmann concludes, ‘a concert is an institution that is not meant to be one’. 51 This process is facilitated by its familiarity, its intimate setting, the one-to-one chats and the authority of those who are present. Its great merit is its ability to remind the leaders of the world’s most powerful nations to cooperate internationally, rather than giving in to domestic pressures. 52 Precisely because of its lack of mandate, agenda or bureaucratic infrastructure, the members of the G8 can act very quickly and decisively when in agreement. 53 While informal forums can create the will to act, formal institutions can legitimise this will and realise the collective strength to execute. 54 Recent examples include G8 initiatives to establish the Global Fund to Fight AIDS, Malaria and Tuberculosis (TBC) in 2001 and the Aquila Food Security Initiative in 2009. The G8 has always cooperated closely with other IOs, both new and previously existing ones, in the formulation as well as the implementation of policy decisions. 55 It is therefore a highly appropriate subject for the study of inter-organisational relations.
These advantages to concert diplomacy are mirrored by a number of challenges, as can also be seen in the case of the G8. First, because concert diplomacy is not based upon formal treaties, it can lack legitimacy. Second, its exclusive nature, combined with this restricted legitimacy, can lead to resentment from those states which are left out. Overall, concert diplomacy is detached from the normal democratic process of decision-making. 56
Finally, the G8’s traits offer both opportunities and challenges for inter-organisational cooperation. On the one hand, its unique nature enables it to complement other IOs as opposed to rival them. Its high profile reputation, moreover, makes it an attractive partner for IOs to work with. On the other hand, its exclusive and state-centric nature, as well as the fact that the G8 is in a way no more than the sum of its parts, makes it a difficult and incohesive ‘actor’ to deal with. The next section discusses how these opportunities and challenges played out in practice in the establishment of its relationship with the EU.
The origin of the G8-EU relationship
The G8-EU relationship dates back as far as 1977, which is particularly interesting in light of the fact that EU nesting in IOs was still very rare at that time, and that this relation remains the only established and durable IO relationship that the G8 has formed. This warrants a closer look at the question of how and why inter-organisational cooperation began in the first place. I argue that rational and social-institutionalist perspectives are useful to explain the divided individual preferences of the constituent parts of both organisations, but that European Community (EC) representation was ultimately ruled legally mandatory by the European Council, and facilitated by the informality of concert diplomacy.
The first summit in 1975 was intended to comprise only five participants, but eventually, the French granted the Italian request for an invitation (Italy held the Presidency of the European Council at the time) in an attempt to justify exclusion of the EC. 57 This decision represented a peculiar beginning for the IOR. At the time, it was not at all obvious that the summit would be a continuous event. In light of that fact, the French made the unilateral decision to invite Italy, without any EC pressure or debate. However, when it became clear that the Americans were to continue the initiative, the smaller states began to stir. Aware of their inability to take part in the Group, they aimed for indirect representation through the EC. More specifically, they asked for a mandated nested relation, through independent Community representation, instead of (rather than in addition to) that of the European summit members. 58 Their claim was in line with the existing acquis communautaire and finalité politique, and the ‘single European identity’ aspirations as expressed by the EU leaders in 1972 in Paris. 59 The smaller member states feared that the summit members would take decisions without consulting them or reflecting their interests. Also, they saw the summit as potentially rivalling (or even replacing) ‘the EC as the principal entity to resolve and manage West European and transatlantic issues’, 60 and were concerned that they might become second-class EC members and lose the augmented voice they had gained through Community membership. 61 Joop den Uyl, Dutch prime minister, even claimed that there was ‘no more certain method of breaking up the Community’. 62 The fears of the smaller Community members were partly incited by the informal and ad hoc nature of the summit, which left many of their questions unanswered.
The non-European summit participants were not per se against full Community representation. 63 Instead, the three European instigators of the summit – France, Germany and the United Kingdom (EC-3) – were the most adamant in their opposition to Community inclusion. France took a leading role in the process and held the strongest preferences for Community exclusion from summitry. Giscard d’Estaing, motivated by his initial interpretation of summitry, emphasised the informal nature of summit meetings in which heads of state should be undisturbed by additional parties whose participation would complicate matters and distort the informal and flexible direct contact. This same informality also had Giscard dismiss accusations of a breach with the acquis, claiming that with no formal decisions being taken, the agreed Community rules on unified international representation did not apply. 64
No agreement on representation could be reached before the 1976 summit as preferences were too far apart. France and the Netherlands took the most extreme positions, with Giscard refusing to even consider the idea of any kind of Community representation, and the Dutch parliament going as far as demanding a boycott of the Council meeting and financial sanctions if Community participation was not accepted. 65 Eventually, both the Commission and the European Parliament began to mediate between the bigger and smaller member states. They tabled a compromise proposal that suggested a parallel approach, which was eventually forced upon the EC-3 when an agreement on export credits made at the 1976 Puerto Rico summit was ruled illegitimate by the European Court of Justice because it was made independently by individual EC member states on a matter of Community competence. Given France’s increasingly isolated position, it had no choice but to compromise, which they ultimately did at the European Council meeting in Rome, a few months prior to the 1977 G7 meeting in London. At this meeting, the Council ruled that ‘(t)he President of the Council and the President of the Commission will be invited to take part in those sessions of the Downing Street Summit at which items that are within the competence of the Community are discussed’. 66
Two questions are of analytical relevance here: first, what were the main initial obstacles to inter-organisation cooperation, and second, how were these obstacles overcome? With regard to the first question, assumptions that organisations are reluctant to give up organisational autonomy seem either not to apply in this case or already overcome. This case, particularly because of the ‘no-more-than-the-sum-of its-parts’ nature of the G8, instead confirms suggestions made by, for example, Biermann 67 that when studying IORs, IGOs should not be considered ‘black boxes’ and that an assessment of the individual member state preferences is in order.
The fact that the non-European summit members were not opposed to EC membership suggests that they did not consider EC participation a threat to G7 interests. Rather, they followed a rational logic when they recognised that the Group could benefit in terms of effectiveness because of the fact that the EC held competence in several vital economic areas. Thus, they were committed to the urgency of the problem they wished to address and felt that presence of the EC could add to the Group’s resources. This optimistic perspective was mirrored by the smaller EC member states, who acted in their own interest as well as that of the EC, which was to bolster international legitimacy and reputation through EC membership. This social-institutionalist reasoning was complemented by arguments of resource dependence. Clearly, the smaller member states also had defensive motives, fearing possible encroachment of the summit upon the EC.
Instead, the misgivings by the European summit members constituted the primary obstacle to cooperation. This contravenes cultural and social arguments that would suggest less inhibitions for cooperation due to the overlapping membership between the two fora. Some stovepiping concerns did play a part here, since Giscard deemed the summit ill-suited for inter-organisational cooperation. Borne out of frustration with existing international institutions, he meant for the summit to stay away from these, almost as if to avoid potential contagion. Possibly the biggest obstacle to the IOR, however, was not unwillingness of the organisations to give up autonomy, but rather the bigger member states’ fear of losing their individual autonomy. Especially, France felt that nested representation would entail an unacceptable loss of sovereignty.
While rational and social factors explain the preferences of the different parties, they do not explain the resolution that was ultimately agreed upon. Hence, the original obstacles were in fact not really overcome. In this case, nesting was a spillover effect of EC political integration that had taken off in the early 1970s and simply fulfilled the legal and regulatory requirements that came into play because the summit consisted (among others) of four EC member states. However, this spillover effect was facilitated by at least two other constitutive factors. First, the conceptualisation of the G7 as a modern concert helps to explain how, in a setting that seems so adamantly state-centric, the EU as a regional body has managed to obtain a seat at the table. Without any clear rules excluding the EU, and based upon principles of normative consistency, authority of those who attend and effectiveness, the G7 had no good reason to exclude the EU. In fact, rationally, it made good sense to include the Community. Another factor was the fact that the summit was relatively new at the time, and not (yet) established as an IO, which made the option of EC inclusion much less fundamental than the more recent debates on G8 enlargements. 68
Thus, when we open the black box, and assess the different preferences both within the EC and in the G8, we can conclude that double membership of the EC-3 played an important part. Not only did it cause EC nesting in the G7 to be mandatory, but it was also problematic in the sense that the member states viewed the relationship as potentially rivalling their own sovereignty. Interestingly, the informality of the summit has facilitated these differing preferences. In the next section, I discuss how this relation has evolved under these informal circumstances.
The developing IOR between the EU and the G8
As Van de Ven 69 and Biermann 70 note, inter-organisational relations tend to develop over time. Given the indeterminate design of the relationship between the G8 and the EU, this one was particularly prone to change. However, apart from the proposition that success will usually lead to more cooperation, and that duration often has a beneficial effect on the relationship, little is written on the conditions that shape this development. This section therefore traces this process by describing and explaining the developing IOR between the EU and the G8.
The changing IOR
The compromise of 1977 set the conditions for the IOR. What is most interesting about these conditions is the fact that they represented the divided preferences of the EC as opposed to G8 preferences, and that they remained quite vague. Compromising between the smaller EC states and the EC-3, a pragmatic and parallel approach was chosen that prescribed the participation of both the Commission and the Council president, alongside the four EC member states (EC-4). Moreover, the relationship was limited to the parts of the conference that were within EC competence and not supported by a formal mandate. ‘(T)his solution leaves the question of the final form of the participation of the Community institutions open’, and it was even made explicit that the Council agreement was not to serve as a precedent. 71 Also, the EU delegate was closely monitored by non-G8 member states through coordination efforts in the Council before and after the summits. 72 Even closer monitoring took place by the EC-4, making the EC position rather delicate during this early period. ‘[I]f it was not fully backed by the other states, its credibility would suffer’, making the Commission reluctant to take too forceful or independent a line. 73 All in all, the compromise illustrates how divided preferences and tight control of the constituent parts of IOs can result in a defensive IOR, with limited depth, formality, duration and strength.
On the other hand, the compromise was open-ended and ambiguous in terms of realisation, which enabled it to change over time. Most significant about this change was that the EU delegation increasingly detached itself from the initial member states preferences and started to act more as a full-fledged IO partner in its relation with the G8. This started with the formalisation of the summit representation by the European Council after the London summit of 1977, giving the relation a continuity that might not have been expected beforehand. As is often the case with IORs, the relation between the summit and the EC started out as single issue cooperation, including the Commission only in discussions on trade and North–South relations. However, ‘(i)t was clear that this sub-division of topics was simply a political act of caution in the face of Giscard d’Estaing’s opposition to the Community’s participation’. 74 Commission participation was first extended to all economic issues in 1978, and then to political topics in 1981, regardless of whether the Community possessed competence in a specific policy area. Thereafter, the EU has always been fully involved in every G7/G8 summit, including preparatory meetings and the subordinate minister’s meetings. Although little has changed in formal terms, as the EU remains devoid of a mandate, and still retains a position of so-called ‘shadow membership’ (the EU is a participant, but its membership-status remains unclear), the EU is nonetheless now treated on par with regular G8 members. Up until the summer of 2014, the only remaining difference between the EU and full G8 members was the fact that the EU was unable to host a summit, a noteworthy fact nevertheless in light of the ad hoc nature of each summit. 75
Additionally, the smaller member states have shown a decreasing interest in the EU’s summit performance. Despite considerable debate taking place in the early days, the pre- and post-summit coordination sessions became increasingly focused on the Commission informing the Committee of Permanent Representatives of EU Member States (COREPER), as opposed to COREPER instructing the Commission. 76 This enabled the EU to become a more autonomous summit participant. 77
Despite this emancipation of the EU as a summit participant and a partner for the G8, the double representation of the larger member states continues to shape the inter-organisational relation in an indirect sense. The European Commission, unlike individual EU-G8 members, must always take into account the positions of the other EU-G8 participants, and is therefore still less likely to be proactive or pressing for more radical approaches. 78 In practice, the position of the Commission will never completely contradict that of the other EU members while the appearance of EU block-forming, or caucusing, is avoided. 79 This situation has resulted in the paradoxical outcome of having strong G8-EU collaboration when EU states act in unison, but also an incentive to avoid too much of such unity. In this sense, overlapping memberships of the two organisations still constitutes a challenge to the relationship between them.
In terms of effectiveness, the relationship between the G8 and the EU varies per issue, depending on the extent of EU competence or its international experience on the topic. If the EU is unable to find internal agreement, such as certain issues of international security, or nuclear energy, G8-EU cooperation remains minimal and limited to neutral mediation by the EU at most. 80 On the other hand, when the EU is able to find internal agreement, the relation can prove productive in different ways. The first manifestation of this productivity is when EU participants use the G8 to internationalise EU policy, for instance, the EU’s ambitious climate change policy in the follow-up of the Kyoto Protocol. While not necessarily leading to direct tangible successes on this issue, the summit has presented the EU with an alternative forum for negotiation, maintaining visibility for the topic more generally as well as the EU’s ambitious climate change agenda. 81 A more successful example of this kind of cooperation is the 2012 Deauville Partnership, an EU initiative in origin, but later on embraced by the entire G8, to support the economic and political transitions in the Arab region. 82 This ‘internationalising’ of EU policy is advantageous not only for the EU but also for the G8, which benefited from extensive EU experience in international agreements, enabling more ambitious and concrete plans to be discussed than would otherwise be likely. 83
Another, less seen manifestation of the productivity of the G8-EU relation is the EU’s capacity to provide the organisational infrastructure to realise G8 aims. The delegation of the PHARE project, an economic recovery program for Eastern Europe, by the G8 to the EC in 1989 is the first example of such successful complementarity: the G8 created the will to act, while the EU had the ability to execute. 84 Apart from some instances in the 1990s, such as the 1994 G7 initiative on Global Information Society, which was executed by the EU, 85 examples of such direct complementarity are rare. More common are initiatives that are geared towards EU competences and capacity, such as the 2002 G8 Africa Action Plan that highlighted conflict prevention, directly in line with the European Common Foreign and Security Policy. 86
In conclusion, there was a challenging start to the relation, with a Commission that had to operate with tied hands, under close scrutiny of its member states. However, over the years, the EU developed into a more independent summit participant, enabling effective summit participation and hence ‘mature’ inter-organisational cooperation. The relationship has grown into complementarity in the sense that the EU profits from the G8’s global impact, while the G8 in turn benefits from the EU’s extensive policy-making experience. Still, while the EU approximates G8 membership, its nested status, as a participant alongside its own member states, hampers its ability to take initiative, making the relation less effective than it in fact potentially could be. The relationship remains somewhat unbalanced, in the sense that it is still primarily defined by the EU itself and its member states, and not by the G8. This also results in the EU using the G8 more than is the case the other way around. The next section offers explanation for these developments.
Explanations for change
Thus far, the theoretical exploration of the factors that alter inter-organisational relations has been limited. While it has been suggested that learning and duration tend to stimulate the deepening and widening of inter-organisational cooperation, the circumstances under which this learning and duration take place are scarcely explored. What is particularly interesting about the case of the EU in the G8 is that the original obstacles to cooperation, the objections made by the EC-3, were initially not overcome, but in fact overruled, resulting in a situation of forced socialisation. This means that the obstacles were still in place when the relationship was established, and can be seen reflected in the original characteristics of this relationship. In the decades that followed, the obstacles were gradually, yet not completely, overcome, primarily due to the (erratically) developing capabilities of the EU, which were facilitated by the concert characteristics of the G8. This suggests that nesting in informal fora offers particular opportunities for the advancement of IORs, despite overlapping membership. This analysis moreover confirms that IOR development cannot be solely explained from rationalist or social-institutionalist points of view, but needs to be seen as the result of a socialisation process, in which path dependency plays a substantial part.
Starting from a rationalist perspective, I assert that expanding EU competences have reinforced the mandatory reasons as well as the functional reasons for inter-organisational cooperation. Tying the original agreement to competences meant that the EU automatically gained G8 actor capability upon the transference of a competence to the EC or the EU. But deepening the relationship was also a functional matter, as EU competence, whether exclusive or shared, usually culminated in more EU position coordination, as well as organisational infrastructure. When that happens, there is more to be gained from cooperation for both IOs. However, the logic does not necessarily reverse, which indicates that competences cannot be taken in isolation. Less competence does not inevitably result in less G8-EU cooperation, which is illustrated by the fact that the EU was included in all summit discussions as early as 1981 even though competences at that time did not extend much beyond the single market. Indeed, G8-EU cooperation is strongest on those issues within EU competence, but even when there is no competence and EU states are not in agreement, for example, on energy, the EU can act as a mediator, fulfilling another useful role in the G8. 87
The problem with competences is that over the past 30 years, they have been in constant flux, and also contain several inconsistencies, often justifying the presence of several EU actors at the same table. As a result, a practice of pragmatism has developed in many IOs, 88 as has been the case here. This communauté de vue, or pragmatism, was reinforced by the rules of interaction of the summit. While legal competences have affected the Community’s room for manoeuvre, the limited formalisation of the summit as well as the multidimensional and integrated nature of the discussions diminished the significance of such competences. 89 After all, if the results of the summit are not binding, it matters less whether or not those who participated in the discussion had the legal authority to do so. Moreover, the focus with regard to the content of the summit kept changing throughout the first years, which was also facilitated by the absence of rules and regulations. In this way, an economic summit in which EU presence would seem mandatory turned into a more general political endeavour in which the Commission still played a substantial role, despite the lack of any formal mandate that allowed them to do so. It was simply a matter of practical spillover, from cooperation covering only a very limited number of issues, to all discussions of the summit. As such, the informality of the summit has enabled the inter-organisational relation to mature, despite the incohesive international capabilities of the EU.
Consequently, this functional explanation needs to be complemented by a social-institutionalist perspective. The most significant explanatory factor in this respect has been the determination of the Commission to invest in the relationship to enhance its international legitimacy, reputation and visibility through cooperation with another high profile IO. ‘The flexible and informal nature of summit deliberations […] has presented the Commission with a golden opportunity to independently extend its own agenda’,
90
enabling the Commission to socialise in ‘a high-ranking network of relations from which it had previously been excluded’.
91
For this reason, Commission representatives such as Gaston Thorn and Jacques Delors made extra efforts to justify their presence: [Delors] did it by providing some kind of external neutral kind of expertise before the summit […]. Concretely, he sent papers, discussion papers, just before the summit, presenting more objectively, neutrally, what was the state of the issue. […] In that the [summit host] was saying ‘now we are starting the conversation on this issue, so Jacques, you send us a very interesting paper, maybe you can present us the state of the situation’. He did it, and it was very helpful for the leaders. And as such it was very well accepted. And they were useful for the debate. At the very beginning, the Canadians and the Americans were maybe asking why he was there, but then they were not asking that anymore.
92
This kind of behaviour has been an important contributor to the deepening of the relation in 1978, and again in 1981. It was more practical and useful to include the Community for the whole summit, than to have them come in and leave every other discussion. 93 Again, the informal, integrative and flexible nature of the summit played a part here, offering opportunities for the Commission that it had not in most other IOs. These opportunities were incidentally not offered to the other EU representative, the European Council presidency. Its rotating nature and therefore inconsistent presence challenged its ability to develop an IOR of any substance. 94
The informality of the summit did not only offer opportunities for EU emancipation. The high-powered setting and flexibility have proven a challenging context for the EU delegations, which tend to be somewhat more formal in terms of authority. 95 The distinct organisational cultures of the two IOs have made cooperation difficult at times: ‘They do things so differently [in the G8], it makes us, as the Commission, a little uncomfortable at times’. 96 Also, the EU’s keenness to invest in the relation to enhance legitimacy and visibility is not mirrored by the G8, which as a collective has never drawn much attention to the EU’s presence, not even to address possible criticism regarding its exclusiveness. 97 This explanation therefore confirms the one-sided agency that underlay the developing relationship.
Another important social explanation constitutes the changing preferences of the member states. A threefold change can be distinguished: the EC-3, the non-G8 member states and, although of lesser importance, the non-EU-G8 members. First, the preferences of the EC-3 shifted in favour of substantial and independent EU representation. They soon learned that parallel representation had very limited effect on their own autonomy within the summit due to the informal nature of the summit. In fact, they learned that having another EU member there could strengthen the EU collective, while it also proved to be possible to divert from the EU line if preferred. 98 Another factor explaining this preference transformation was the change of government in France and Germany and the change in personalities that went along with it. Mitterrand and Kohl were more supportive of the Commission in the G8 framework than their predecessors. 99 Despite the generally favourable disposition of the EC-3, however, their dominance in the summit ensures that a united and strong EU position is only pursued when it suits them for that particular topic. 100 Consequently, member states’ preferences remain the primary explanation for the ambiguous position of the EU at the summit.
Second, the preference change by the smaller member states is also explained by G8 informality. The concert nature of the summit, its purpose of guiding other IOs and its rules of exclusive membership and non-binding interaction defused the danger of encroachment. Moreover, most of the issues that made it on to the summit agenda had already been discussed frequently within the EU itself, making explicit position formation beforehand superfluous. 101 Another factor in this preference formation was the practical standstill of European integration in the early 1980s. With European economic policy coordination on the back-burner, none of the members really worried much about Community representation at the G7 anymore. 102
Finally, the preferences of the non-EU summit members had never been particularly strong. The parallel arrangement was respected by other G8 members, which, in line with the characteristics of concert diplomacy, respect each other’s legitimate rights, in this case those of the EC and those of individual member states. As the interview excerpt illustrates, once the Commission had proved its added value to the summit, in terms of, for example, mediation or expertise, they became even more accepted. On the other hand, non-EU summit members are less keen than their EU counterparts to elaborate on pre-existing EU policy or use EU infrastructure for implementation. This endorses the image of G8-EU cooperation primarily on EU terms and hence not being fully effective.
These rational and social factors are reinforced by various constitutive factors that have shaped the resources and preferences of these two IOs. Some of these factors have afforded opportunities for G8-EU cooperation, like the demise of the Soviet Union. This external shock coincided with a fundamental deepening of EU integration in the late 1980s, as well as renewed international faith in IOs and IO cooperation, and finally severe issue density on the G8’s agenda. These changing circumstances not only altered the urgency of the problems that the G7 aimed to address but also the belief in and availability of the resources that the EU might be able to provide to address this problem. On the other hand, other external developments such as the crisis in Yugoslavia revealed internal division and lethargy in the Union, indicating that despite urgency of the problem for both the G7 and the EU, the EU lacked the resources to deal with it. 103 These developments coincided with a downscaling of the summit agenda and ambitions, which stalled further advancement of the G8-EU relation.
The analysis displays an interesting complementarity between two factors: namely, the increasing agency of the EU, combined with the informal structure of the G8. However, this process tracing also indicates that inter-organisational relations are the result of a socialisation process, testifying to a logic of path dependence. Path dependency here is understood as the effect whereby ‘the entrenchment of certain institutional arrangements obstruct an easy reversal of the initial choice’. 104 Path dependence suggests that certain processes are self-reinforcing, meaning that a process, once set in motion, creates forces that make it difficult to undo this process. 105 The benefits of persisting with the chosen alternative increase over time as do the costs of switching to another. 106 Path dependence does not merely suggest repetition or preservation, but instead causality. Hence, an event is repeated, or preserved, or even deepened, because it happened in the past. 107 By this logic, certain (seemingly) small events and decisions – critical junctures – in the past may turn out to be of great consequence later. If this logic is followed, the assumption is that once the decision is made to initiate inter-organisational cooperation, the benefits of this cooperation will increase over time, as will the costs of ending this cooperation. This is especially the case when the set-up costs are high, and the learning effects, as well as expectations culminating from the cooperation, are significant. 108
All of these elements of path dependency can be identified in this particular case. First of all, although the financial and organisational set-up costs of the relation were limited given the ad hoc nature of the summit, the fundamental set-up costs were significant in the sense that the preceding internal EC debate was of such intensity that heavy means had to be used to resolve the matter. After almost 2 years of internal debate, the European Council and the European Court of Justice ultimately forced the issue. Despite the fact that they explicitly declared that their 1977 agreement on the parallel approach was not to set a precedent, the result was quite the contrary due to the power of the legal argument. Once that battle was won, the costs of switching to possible alternatives were already unacceptable, since it meant a direct challenge of the existing Community structure. Also, the decision to nest formalised the agreement in the sense that it was more than merely an ad hoc interaction, as so many other IORs are conceived. Thus, the decision constituted a critical juncture.
The intention to arrange summit representation on the basis of competences can be considered an attempt to rationalise the delegation process. However, when the Commission entered the summit that one time, they were accepted in the club and soon given most of the rights and privileges that accompany such a position. This original restrictive approach proved impractical in the informal context of the summit and was already abandoned in 1981. Once it had attained limited access, full participation was just that much more practical, indicating that deepening of the relationship was a direct result of the earlier decision to participate in the first place. Once the Commission was fully involved, socialisation and learning further increased the benefits of cooperation, for the IOs, as well as their constituent members. The actual practice of the chosen solution defused the fears of the EC-3, as well as those of the smaller member states, making parallel representation in fact also not not in their interest, and hence possible alternatives (i.e. no Community participation, or exclusive Community participation) less feasible than they originally were. The non-EU-G8 summit members were not only outnumbered but also simply not willing to make the effort to change the construction, despite their ambivalent feelings towards it at times. 109 Again, the costs of switching were deemed too high, and the current option was not considered problematic enough.
Path dependency also accounts for the persistence of the double representation, and the challenges that this structure brings. The summit being an initiative by two large European member states (critical juncture again), the presence of the larger EU members in parallel to the EU was never questioned. A quote from a Commission official in 2008 illustrates this: ‘The G8 structure [with all those European representatives] is a mess. It should change, of course. But we are stuck with this system and it will be very hard to change’. 110 Interestingly, even when the opportunity for change presented itself in the new millennium with the public questioning of the relevance of the G8 in its traditional composition, the situation remained as it was. The option of finding a different solution for EU representation, for example, by replacing the European member states with one EU delegation and hence making room for additional G8 members such as the other BRICS, was never viable, due to the impossibility of the European G8 members giving up their seat. 111
Conclusion
This article has made suggestions about the conditions under which inter-organisational relations develop. The case of the G8-EU relation offers two particularities that render closer investigation relevant: first, it is an IOR between two opposites in terms of formalisation and institutionalisation. Second, the case entails double representation of some of its constituent member states.
These particularities have played an important role in the establishment as well as the development of the IOR between the EU and the G8. Consequently, the relationship cannot be simply considered a rational decision resulting from the deliberations between two autonomous organisations. Instead, the cooperation preferences among the G8 members as well as EU member states were divided, initially resulting in no cooperation at all, despite the abundant presence of functional and organisational overlap between the two forums. Ultimately, when the relationship was established in 1977, it was primarily a spillover result from European integration, following the doctrine of parallelism. 112 The early years of the cooperation illustrate how divided constituent preferences can thwart IORs, by limiting the actor capabilities of the organisations involved.
However, the informal setting of the case has enabled this relation to broaden and strengthen over the decades that followed, culminating in a mutually beneficial cooperation in which the EU offers concrete policy proposals and the G8 a global platform for visibility and impact. Still, this testimony of the effectiveness of complementary IORs is nuanced by the fact that the relationship can still be seen as mainly reflecting the terms of the EU and its member states, making a real partnership still problematic in the sense that there seems to be an imbalance between the benefits that these two IOs reap from this cooperation.
Process tracings of IORs have so far remained sparse in the literature, but on the basis of this case and its particularities, we can draw a number of conclusions. First, informal relations are especially prone to change, given their flexibility. However, at the same time, autonomous actor capabilities seem to be important for an IOR to deepen in order for the benefits to be worth the costs of cooperation. To this end, both organisations must not only be motivated to engage in the relation and sufficiently equipped with policy tools but also supported by constituent member states. Even in the long run, however, double representation, especially of large and influential constituent members, remains an impeding factor for cooperation because it limits this autonomy. Second, this case suggests the limitations of rationalist and social-institutionalist explanations, since the development of the relationship cannot be solely attributed to rational deliberations by either the IOs or their member states. A logic of path dependence instead helps us understand how IORs, once started, are locked in due to increasing benefits as a result of learning and socialisation, and inflating costs of switching.
This research presents interesting opportunities for future investigation, particularly on the relation between actor capability and IOR effectiveness, as well as the effect of (divided) constituent preferences. Moreover, it would be interesting to see whether informality is in itself a fertile ground for inter-organisational cooperation, as seems to be suggested by this case.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.