Effects of Contestation Within a Collective Agent in EU Trade Policy-Making

  • Markus Gastinger
Open Access
Part of the Palgrave Studies in European Union Politics book series (PSEUP)


This chapter argues that conceiving the Commission as a unitary actor is a legitimate simplification of empirical reality for most research purposes. Recently, the principal–agent literature has started examining collective-actor features of agents and suggested that internal conflict is as disadvantageous to agents as it is to principals. By contrast, I argue that intra-agent conflict occurs less frequently and with a lower intensity than is typically the case for collective principals. The Commission-as-agent can overcome conflict quickly due to its hierarchical setup and less stringent decision-making procedures. It may even harness conflict through inter-service consultations to draft better initial proposals. On the political level, open conflict among Commissioners can earn it the reputation of credibly defending controversial sectoral interests in the eyes of affected stakeholders. I develop my argument by focusing on five bilateral trade agreements negotiated from 1970–2007. Generalizing it beyond the case of EU trade policy-making, I focus on empirical, theoretical and methodological reasons justifying many scholars’ choice to model agents such as the Commission as unitary actors.

1 Introduction 1

The Commission of the European Union (henceforth simply ‘Commission’) employs 23,330 individuals—from the Commissioners to clerical workers. These were spread across 45 Directorate-Generals (DGs) and services (Commission 2016). 2 Politically, the Commission is represented by 28 Commissioners with varying national and ideological backgrounds from ‘old’ as well as ‘new’ member states speaking at least 24 languages. Despite this enormous diversity, rivaled only by the ancient dwellers of biblical Babylon, ‘the Commission’ has in the scholarly world frequently been portrayed as a unitary actor . This is equally true for the burgeoning literature on the principal–agent model applied to the study of the European Union (EU). But how tenable is the unitary-actor assumption with regard to a ‘collective agent’ such as the Commission? What effect does intra-agent conflict have on the principal–agent relationship? Are there more benign effects than conflict within a ‘collective principal’ (such as the Council) typically has? This chapter draws on the example of EU bilateral trade agreements (BTAs) negotiated between 1970 and 2007 to bring us closer to the answers.

In brief, I argue that conceptualizing the Commission as a unitary actor is, by and large, justifiable. Empirically, the cases presented in this chapter suggest that conflict within the Commission is rare and unlikely to persist for too long. While this finding is based on insight from the EU’s trade policy-making, the factors I identify accounting for the low frequency and intensity of conflict within the Commission equally apply to all other policy areas. Theoretically, collective agents such as the Commission are characterized by a more hierarchical setup and less stringent decision-making procedures than collective principals such as the Council. This generally precludes insurmountable conflict. In fact, the Commission may ‘cultivate’ internal conflict to enhance its discretion. Methodologically, the information necessary to assess conflict is exceptionally difficult to obtain for contemporary episodes of European decision-making. Political conflict among Commissioners with concomitant media attention apart, interviews are the most promising strategy to learn about administrative conflict. Yet it is questionable whether officials would openly acknowledge internal clashes. The principal–agent relationship between the Council and the Commission and the focus on contestation within the collective agent, which form the basis of this chapter, are visualized in Fig. 1.

Fig. 1

Visualization of the principal–agent relationship and contestation within a collective agent

The data for this study are drawn from two sources. Three BTAs fall below the EU’s thirty-year rule and can rely on all information found in the Historical Archives of the European Commission (HAEC) located in Brussels. The two more contemporary BTAs draw on material from specialized news agencies and interviews with Commission officials. Talking only to Commission officials could bias my findings. However, since I am interested in internal Commission conflict, the perspective of outsiders such as Council officials or third-country negotiators promises only limited insight.

2 Theoretical Puzzle: Contestation in Collective Agents

Although constituting a multi-faceted phenomenon, terms like ‘conflict’ or ‘contestation’ are used interchangeably in this chapter. At its conceptual core, conflict can arise in all decision-making contexts in which actors are made up of more than one single individual, which is the smallest possible unit of analysis and always has homogeneous preferences. Where entities comprise two or more actors with discrete preferred outcomes that cannot be attained simultaneously, conflict occurs. 3 When diverse preferences have been aggregated, the entity can speak with a single voice and becomes, for the purpose of analysis, a unitary actor in its own right. Other than this, the study relies on the standard terms and definitions encountered in the principal–agent literature and as developed in the Introduction (Delreux and Adriaensen this volume). The principal–agent model deals with a single dyadic relationship from one actor that conditionally transfers authority (‘principal’) to another actor (‘agent’). This does not imply that factors such as non-principal third parties (e.g., supporting principal control as fire alarms for McCubbins and Schwartz 1984; or increasing agent autonomy through permeability for Hawkins and Jacoby 2006) or the informational configuration of the decision-making context (Kiewiet and McCubbins 1991; Pollack 2003) must be omitted. But they are always linked back to the principal–agent relationship and the struggle between the two whenever their preferences go head to head.

Scholars soon acknowledged the pluralist nature of principals. Weingast and Moran (1983: 770–775), for example, in one of the earliest contributions to the political delegation literature argued that it takes a majority of members in US Congress committees (the principals) to control regulatory agencies (the agents). Within the scope of actions not opposed by a majority of congressmen, the agent is free to select the actions closest to its own preferences. Several individual actors coming together in a single body, functioning as one collective principal, is akin to situations where one agent faces various principals spread across separate bodies as multiple principals (Moe 1984: 768–769; Conceicão-Heldt this volume). Nielson and Tierney (2003) studied how collective principals overcome collective-action problems and how mixed signals from multiple principals propel agency slack. They found that agent autonomy increases with heterogeneity among principals. In EU foreign policies, Council unity is a similarly central factor for explaining Commission discretion (Conceição-Heldt 2011; De Bièvre and Eckhardt 2011; Elgström and Larsén 2010; Niemann and Huigens 2011). Therefore, principal–agent scholars widely share the view that conflict among (collective or multiple) principals results in increased agent leeway.

By contrast, the agent side of the equation has long been left unrefined. The push to specify the principal–agent model in the direction of multiple agents (Helwig this volume) and collective agents is rather recent (cf. Laloux this volume). In Pollack’s (2003: 384) influential book, the Commission-as-agent is still conceived as a unitary rational actor with coherent preferences for deeper integration, which he describes as a useful simplifying assumption ‘despite the regular differences […] among its Commissioners and Directorates-General.’ At the same time, reducing EU institutions to unitary actors has been criticized as ‘extremely questionable’ (Kassim and Menon 2003: 133). For common commercial policy, Kerremans (2004: 367) considered the Commission’s ability to overcome internal diversity essential for its credibility to shape policies. It was then also EU trade negotiations where the unitary-agent assumption was first relaxed. 4 Damro argued that the Commission is better viewed as a collective agent in World Trade Organization negotiations since both DG Agriculture and DG Trade speak for the EU and ‘tensions can arise within the Commission itself’ (Damro 2007: 896). 5 That these tensions can negatively affect Commission influence is not mentioned explicitly but seems a fair extrapolation from his argument. Similarly, Elgström and Larsén (2010) conjecture that Commission heterogeneity has an adverse effect on its ability to shape outcomes. When testing this proposition against the backdrop of Economic Partnership Agreements (EPAs), however, they find the Commission to have been largely unified, leaving their theoretical proposition empirically unverifiable. 6

In summary, there is a clear tendency in the principal–agent literature to equate heterogeneous agents with heterogeneous principals. Indeed, if we model the setting as a simple extension of the chain of delegation (Dür and Elsig 2011: 332; Bergman et al. 2000; Strøm 2000: 267–268; Nielson and Tierney 2003: 249–251; see also Larsén 2007: 862–865), ‘the Commission’ becomes the collective principal that delegates authority to its DGs qua agents. From this perspective, it appears plausible to expect that internal Commission conflict negatively affects its tug of war with the Council. Nevertheless, in this chapter, I follow the traditional conceptualization according to which the Council is the (collective) principal and the Commission its (collective) agent. The Commission is empowered in trade policy-making, first, by member states through the Treaty and, second, by the Council through a specific mandate issued for each negotiation. 7 Let me now turn to five BTAs to assess the level of conflict within the Commission and possible effects.

3 Empirical Analysis: The Negotiation of Five Bilateral Trade Agreements

The five BTAs surveyed in this chapter are the Commercial Cooperation Agreement (CCA) with India of 1973, the Trade Agreement (TA) with China of 1978, the Commercial and Economic Cooperation Agreement (CECA) with India of 1981, the Association Agreement (AA) with Chile of 2002 and the Economic Partnership Agreement with CARIFORUM (C-EPA) of 2007. Why these cases? First, focusing on trade policy seems well suited because the Commission qua collective agent was first identified in this area. Second, the selection represents a variety of cases with countries from different world regions and concluded at different times, thereby mitigating the risk of selection bias. Third, the EP’s role in the pre-Lisbon era was limited (Reichert and Jungblut 2007). Since I focus on the single dyadic relationship between the Commission and the Council, cases with limited involvement of the third main actor in the EU political system mitigate the risk of omitted variable bias. Table 1 provides an overview of all cases and occurrences of intra-agent conflict. 8
Table 1

Cases included in this chapter and occurrences of intra-agent conflict

Bilateral trade agreement

Third party

Intra-agent conflict

Commercial Cooperation Agreement (CCA)



Trade Agreement (TA)



Commercial and Economic Cooperation Agreement (CECA)



Association Agreement (AA)



Economic Partnership Agreement (EPA)



Before introducing each of the agreements in turn, let me briefly summarize my main empirical finding, which is that conflict within the Commission occurs far less frequently than in the Council. While there was substantial variation in member states’ preferences in all five BTAs, the Commission only suffered relatively mild hiccups in forming a common position in the CCP with India and the AA with Chile. If we concede for the moment that in the latter case it was only a deliberate strategy to preemptively heed agricultural concerns (discussed below), we are left with one case: the earliest and the one where inter-service consultations had not yet become deeply rooted in the Commission’s bureaucratic culture. This suggests that conflict in the Commission today occurs rather infrequently.

3.1 CCP India (1970–1973)

Contestation among Commission DGs erupted already before the start of the formal negotiations. In March 1971, a Commission Delegation visited New Delhi to discuss, among other things, future trade relations with Indian government representatives. The Committee of Permanent Representatives (COREPER) charged the Commission to prepare a report on the results of that meeting (Commission 1971a), which was drafted by DG XI (Trade). DG XI planned to submit the report to the College of Commissioners (henceforth simply ‘College’) on May 12 (Commission 1971c; DG XI 1971b). At the meeting of the chefs de cabinet one day earlier, however, the report drew the ire of representatives from various other DGs. They criticized, primarily, that they were excluded from the drafting process (DG XI 1971d). The report subsequently underwent a substantial revision and was adopted on July 28, with the final version having been amended in several respects to make it more palatable to member states (Commission 1971b). This suggests that intra-agent conflict has more benign effects and can improve outcomes by accounting for different viewpoints and—where the interests of principals are not preemptively heeded—preparing the Commission to defend this decision in the Council.

More conflict erupted during the preparation of the draft negotiating directives, which began in November. DG XI again wanted to involve few other DGs, sending its initial draft only to DG I (External Relations), the Legal Service and the Service on Customs Union. In fact, it tried to rush the decision and force adoption of the draft before November 19 (DG XI 1971a, c). In what appears to have been a protest to an arbitrary deadline, DG I circulated its stance on November 22. It considered it better not to propose negotiations to the member states, which probably opposed bilateral negotiations anyway and preferred a regional approach (DG I 1971). DG XI, again, was forced to initiate wider consultations within the Commission. On January 25, 1972, it forwarded an updated version of the draft negotiating directives to DG III (Industry), DG VI (Agriculture) and DG VIII (Development). In a letter to DG I, DG XI pointed out that Commissioner Dahrendorf, who was Commissioner for External Relations and Trade in the Malfatti Commission, explicitly asked for the dossier to be taken forward, effectively overruling DG I’s concerns (DG XI 1972a: 243). Moreover, in 1973, DG I and DG XI were merged under the Ortoli Commission. Conflict among the officials receded noticeably, suggesting that political intervention from the top and institutional restructuring can exert a soothing effect. 9

The draft negotiating directives of DG XI also contained a provision on better utilization of technical and financial aid, which antagonized DG VIII. While the aforementioned conflict could be easily resolved by virtue of both DGs working under the supervision of the same Commissioner, this conflict took more time to be resolved. In a note to Commissioner Dahrendorf, DG XI reported ‘persistent and difficult’ discussions with DG VIII particularly with regard to aid. DG VIII insisted on fully removing this reference from the draft. DG XI conceded the point for the sake of taking the dossier forward, albeit recording in the Commission minutes that this solution was purely ad hoc and did not prejudice future decisions (DG XI 1972b: 272–273). This compromise, however, was overturned at the highest level. Commissioner Dahrendorf raised the issue in the College (DG XI 1972c). As aid does show up in the final draft directives, the other Commissioners must have followed Dahrendorf’s lead. This also highlights the hierarchical element of the Commission enabling it to overturn lower-level decisions. Moreover, the fact that no further intra-Commission conflict occurred during the remainder of the negotiations suggests that contestation is more frequent early in decision-making processes. This is plausible considering that preferences still need to be formed and aggregated.

3.2 TA China (1975–1978)

There was no conflict among Commission DGs in the TA with China. In fact, discussions between the DGs associated with the dossier were extremely short-lived. On September 1, DG I prepared draft negotiating directives (DG I 1977a). On September 9, the Legal Service recommended minor changes (Legal Service 1977), which DG I accepted virtually en bloc (DG I 1977d). On September 22, the final draft was forwarded to Commissioner Haferkamp under an accelerated procedure (DG I 1977c). On September 26, there was disagreement among the chefs de cabinet over China’s accession to the generalized system of preferences. However, the debate was resolved quickly and on the same level (DG I 1977e). The entire process lasted only four weeks, and only the Commission’s Legal Service was formally involved. DG III and DG VI have been involved in the exploratory debates with China informally (DG I 1975), while DG VII (Transport) was involved informally while the directives were debated in the Council (DG I 1977b; DG VII 1977). Informality (adopted probably because of the sensitive nature of the dossier) seems to exert a dampening effect on conflict among Commission DGs.

3.3 CECA India (1978–1981)

By the time of the CECA with India, the Commission seems to have drawn some lessons from the CCA negotiated a decade earlier. DG I was put in charge of the whole dossier. 10 But, crucially, this time various other DGs were involved from very early on in the process. On November 30, an intra-Commission meeting with around ten DGs took place to prepare the exploratory talks with India (DG I 1978). This is an early sign of inter-service consultations aimed to discuss—and improve—the Commission position amid greater levels of coherence and unity among its DGs. This may also explain why no instances of intra-Commission conflict occurred later in the process, for example when the draft negotiating directives were prepared.

3.4 AA Chile (1998–2002)

The AA with Chile is a striking example of conflict within the Commission at the highest political level. The Commission initially scheduled the adoption of its draft negotiating directives for 8 July 1998. Due to the opposition of the Austrian Commissioner for Agriculture Franz Fischler, it was put off for two weeks (Agence Europe 1998a). He voiced ‘fundamental reservations’ in the College that the AA could interfere with the reform of the Common Agricultural Policy. On the other side of the argument, the Spanish Commission Vice-President for External Relations Manuel Marín pushed for a fully fledged Free Trade Agreement that included agriculture (Agence Europe 1998c). On July 22, matters came to a head. After acrimonious discussions in the College that lasted throughout the morning, the issue was resolved only with a formal vote—a rare measure in a body characterized by consensual decisions. Only four of twenty Commissioners opposed a Free Trade Agreement, all from countries with considerable agricultural stakes in the negotiations (Agence Europe 1998b). Nevertheless, Marín easily achieved the simple majority needed for decisions in the College. In this case, the dividing line in the Commission was not along functional lines but largely superseded by national concerns. The decision also highlights the Commission’s ability to aggregate preferences quickly.

If it had not been clear earlier, the episode taught the Commission that agricultural issues would be controversial for the member states. It needed to ensure that agricultural concerns would not torpedo the final AA. In fact, three months before the Commission adopted its draft negotiating directives there was an ‘unplanned and informal’ debate in the Agriculture Council led by France on BTAs. Paris rallied almost unanimous support from agricultural ministers that these should be monitored vigilantly. Fischler followed the debate first-hand, comforting member states that the Commission ‘always bear[s] in mind the possible repercussions on the agricultural sector’ (Agence Europe 1998f). In the week the Commission adopted its draft negotiating directives, the Agriculture Council restated these concerns, again almost unanimously (Agence Europe 1998d; Agence Europe 1998e).

How should we interpret Fischler’s role in the negotiations? He certainly earned a reputation as ‘guardian of agricultural interests’ when the draft negotiating directives were discussed. The fact that Fischler was an outspoken admonisher of the agreement on agricultural grounds during the adoption process did not impede his credibility guarding that flank later in the negotiations. From the Commission’s perspective, he kept track of potential opposition from agricultural interests. In March 2002, member states in the Agriculture Council ‘stressed the urgency of concluding negotiations with Chile’ (Council 2002). The Commission could be sure that no opposition to the AA was waiting in the wings. Commission conflict therefore, arguably, did not negatively affect its ability to conduct the negotiations. In fact, the dispute might have set up the Commission to credibly defend controversial sectoral interests throughout the process.

3.5 C-Epa Cariforum (2002–2007)

The Commission was largely united in the case of C-EPA. Initially, DG Trade conducted the negotiations alone and intended to conclude a ‘pure’ trade agreement. 11 DG Development, which was usually involved in dealing with countries of the African, Caribbean and Pacific (ACP) group of states, was only loosely associated at the beginning. In 2002, DG Trade engaged in inter-service consultations, particularly with DG Development, DG Environment, DG Agriculture, DG Enterprise and DG Taxation. Unlike in earlier periods, no conflict erupted between DG Trade and DG Development at this point. This may be explained by the fact that when DG Trade had been set up only a few years earlier in 1999, many officials previously working in DG Development had been moved to DG Trade. This injected a dose of development expertise into the new service and contributed to the constructive atmosphere among former colleagues (Elgström and Larsén 2010: 213). In 2006, the scope of the C-EPA negotiations was broadened and DG Development became more heavily involved. The cooperation with DG Trade worked smoothly. DG Trade acknowledged that without a solution for development cooperation, a deal was out of reach. DG Trade therefore constantly sought to involve DG Development in all groups, even when negotiations covered pure trade parts, as CARIFORUM insisted on negotiating development issues in parallel (Interview A). DG Development assumed a position as ‘co-leading’ DG without giving rise to major differences (Interview B). This case also suggests that agents restructure their internal administrative setup so as to reduce conflict.

4 Theoretical Argument

In this section, I will tease out likely effects of agent conflict in the framework of the principal–agent model. Generally, it is a truism that any agent that is complex enough to be made up of various units performing diverse functions sends out conflicting messages. For example, since DGs were created for functionally different—often contradictory—reasons, it is hardly surprising that DG Industry and DG Environment normally disagree on the preferred future course of action. 12 Consequently, one should not rush to conclusions and link contradictory statements and intra-agent conflict to unfavorable outcomes for the agent. Instead, one has to carefully consider potential effects and see whether conflict plays out against the agent in the long haul. Moreover, the fact that contestation tends to occur more frequently at the beginning of decision-making processes complicates measuring its effects on agent discretion. Since there are numerous intervening variables before final outcomes are observed, assessing the effect of conflict in isolation becomes a daunting task (cf. Reykers and Beach this volume).

One of the most fundamental intervening variables are agents’ decision-making procedures. The less stringent these procedures are, the less likely it is that conflict has a lasting negative impact. The factors enabling the Commission to take decisions more easily and overcome conflict can be highlighted by a juxtaposition with the Council. First, the College can vote on contentious issues by simple majority ever since the Treaty of Rome. This allows the Commission to take dossiers forward relatively easily, either through an outright vote or in the shadow of it. The Council needs to muster unanimity or, at least, a qualified majority to take decisions. 13  Second, the Commission’s principle of collective responsibility (Hix 2005: 41), which holds that decisions need to be defended by outvoted Commissioners, suggests that it will exhibit strong unitary-actor features after decisions have been taken. The Council does not know any such norm. In fact, states can ferociously defend their stance even after being outvoted to shift the blame for unpopular decisions to Brussels. Third, the fact that the role of the Commission President, originally seen as primus inter pares (first among equals), has been strengthened to become a primus super pares (first above equals; Kurpas et al. 2008: 32) has made it a more efficient preference aggregator. In extreme circumstances, the President can single-handedly define the Commission’s stance. The Council Presidency’s powers, while substantial, fall far short of that mark. Fourth, the Commission exhibits one esprit de corps and is located mainly in Brussels, whereas the Council’s administrative pockets go deep into national territory and are far flung. All of this makes the Commission a very different actor from the Council. It is hard to imagine, for example, that Commission officials could ever find themselves in a Council meeting saying that ‘the Commission’ was unable to forge a common position. If this is true, then the Commission always represents a unitary actor from the Council’s point of view—something that surely does not hold the other way around.

The Commission’s superior ability to aggregate heterogeneous preferences and deal with internal conflict puts the initial analogy between the Commission and the Council in the second section, derived from a chain-of-delegation approach, into question. In general, principals will have more stringent decision-making procedures (including but not limited to voting rules) than agents as otherwise important efficiency gains for overcoming collective-action problems would not be reaped and a strong rationale for delegation falls away. A corollary is that decision-making procedures will be less stringent the further down the delegation chain we go. An inference prima facie confirmed by the EU when focusing on voting procedures: intergovernmental conferences including national parliamentary ratification are more stringent than Council decisions (either by unanimity or by qualified majority) which are more stringent than Commission decisions by simple majority. Even earlier, the delegation from citizens to parliaments and governments through elections is the most demanding decision-making procedure. In the same vein, the unitary-actor assumption should be relatively more justifiable later in the chain.

This does not, however, mean that the Commission is immune to internal contestation. The CCP with India is a clear testament to that effect. Generally, this case suggests that conflict within agents could translate into more discretion by preparing balanced proposals that account for multiple viewpoints. Contestation could thus strengthen the Commission’s position by improving the original proposal in the process of accounting for different interests and—where valid points are not incorporated—by preparing the Commission to defend this decision in other institutional arenas like the Council. The most convincing causal mechanism through which such an effect could come to bear is expert knowledge. As much is suggested by Elgström and Larsén, when writing that the Commission ‘can use the preferences and expertise of the different DGs to push various aspects of the overall Commission position’ (2010: 217). Indeed, Bailer (2014) draws on the principal–agent model for a large-N study and finds that expertise has a significant effect on the Commission’s ability to defend its original proposals (albeit a substantively very small one). Crucially, Bailer operationalizes expertise by resorting to the staff and budget size of the lead DG as proxy variables. Inter-service consultations, which amount to pooling resources across all DGs associated with the dossier, are entirely missing from her argument. By contrast, the Commission seems aware of the advantages of consultations and institutionalizes them (Metcalfe 2000; Nugent and Saurugger 2002). Consequently, intra-Commission contestation on an administrative level carries with it the promise of raising a discretionary dividend in ensuing negotiations with the Council. To what extent this is generalizable to other collective agents is unclear. But where agents have installed institutionalized mechanisms that require deliberation across administrative units, the same relationship between conflict and discretion can be expected to hold.

Another causal mechanism connecting Commission conflict and increased discretion leads through the College. Drawing on the Chilean example, conflict can exert benign effects by giving credibility to individual Commissioners in charge of sensitive dossiers and acting as a guard against involuntary defection whenever ratification by the Council is required at the end. How do we know that the conflict between the Commissioners for Agriculture and Trade over Chile was not, in fact, a ‘stage combat’ to increase the Commission’s agricultural profile? A related argument has been made by Delreux and Kerremans (2010) when identifying paths through which the Commission can affect the Council’s cost-benefit analysis of activating control mechanisms. They argue that the Commission gives national delegations access to multilateral negotiations to create a sense of ‘co-responsibility.’ I extend their argument by specifying how collective-agent features may help the Commission to exploit this dynamic. The more public ‘internal’ conflict becomes, the greater the chances that we witness the first act of a bigger drama. Paradoxically, the more Commission conflict we see, the less conflictual decisions could actually be (reminding principal–agent scholars of ‘observational equivalence,’ see also Delreux and Adriaensen this volume; Kroll this volume; Reykers and Beach this volume).

How then are the administrative and political levels connected? In the CCP with India, we saw that Commissioners take over when the administration strays too far from their preferences. Commissioners can overturn lower-level decisions due to the hierarchical setup. Moreover, all Commissioners have the privilege of raising issues they deem important in the College, irrespective of which or how many DGs are concerned. When this happens, the less stringent decision-making procedures laid out earlier come to bear. Expressed in more general terms, the more hierarchical the structure of collective agents is, the less conflict we should witness. Moreover, since there is no interest for the political level to become bogged down by administrative turf wars, the Commission faces incentives to keep conflict levels at productive levels. The Indian CCP and C-EPA both suggest that the Commission merges services in the hope of oiling its administrative machinery. Therefore, whenever collective agents have been empowered to determine their own administrative structures, the potential for conflict is mitigated. Conflict within collective principals cannot be averted this easily through administrative overhauls. Another strategy for coping with conflict suggested by the TA with China is to involve DGs informally and at more senior levels. Experienced officials will have a greater ability to seek compromises. In sum, there are weighty theoretical considerations that lend confidence to scholars’ choice of treating the Commission as a unitary actor

5 Conclusion

In this chapter, I have argued that conceiving the Commission as a unitary actor is largely appropriate. This is true for applications within the framework of the principal–agent model and also beyond. Empirically, conflict within the Commission occurs infrequently. Theoretically, the Commission is clearly not ‘just another collective principal’ as a simple extension of the chain of delegation would suggest. Whenever conflict does occur, it cannot easily get out of control because of the Commission’s less stringent decision-making procedures and hierarchical setup. In fact, the Commission cultivates conflict through its inter-service consultations in order to produce more balanced proposals and prepare it for discussions with other EU institutions. Whenever conflict occurs in the College at the political level, matters can be relatively easily settled. The concomitant public stage combat can improve the Commission’s position by lending credibility to Commissioners defending controversial sectoral interests. The notion prevalent in the, slowly emerging, principal–agent literature on collective agents is wrong to suggest that conflict is necessarily detrimental for agents.

Finally, there are methodological considerations to adopt the unitary-actor assumption for the Commission. It may simply be infeasible to account for every little detail of preference formation. While conflict within a collective principal made up of nation states such as the Council can often be assessed by virtue of national media outlets tracking their country’s position, even specialized news agencies may or may not have information on conflict within the Commission. Longitudinal research designs and the inclusion of archival material can go some way to addressing this shortcoming and allow more sophisticated applications of the process-tracing methodology (Gastinger 2016; Reykers and Beach this volume). But extracting the necessary information for contemporary decisions may prove impossible. Interviewing Commission officials goes some way to remedying this issue. But officials may be tight-lipped and not speak candidly and in sufficient detail about internal struggles. Therefore, there are sound empirical, theoretical and methodological considerations why the Commission can be modeled as a unitary actor for most research purposes. Crucially, this does not mean that accounting for specific actions by agents becomes irrelevant in the principal–agent model. Quite the contrary, agent strategies are, if anything, under-researched (but see Coremans and Kerremans this volume; Gastinger 2016). But while collective principals should normally not be treated as unitary actors, for agents the same assumption can be a useful shorthand to direct scholarly attention elsewhere. At least in this respect, the principal–agent model has not been rendered obsolescent by contemporary developments.


  1. 1.

    For comments on earlier drafts, thanks are due to Tom Delreux, Johan Adriaensen, Adrienne Héritier, Andreas Dür, Eugénia da Conceição-Heldt, Arne Niemann, Friedrich Plank, Dirk De Bièvre, Mark Thatcher, Hylke Dijkstra, Jens Blom-Hansen, Yannis Karagiannis, Gijs Jan Brandsma, Yf Reykers, Patrick Mello and André Isidro. All remaining errors and inconsistencies are my own.

  2. 2.

    This leaves out thousands of officials working in the European External Action Service (EEAS) and EU Delegations (previously) working for the Commission (e.g., Kostanyan 2014; Dijkstra this volume) as well as specialized agencies (e.g., Dehousse 2008; Thatcher 2011). These are excluded from this analysis to focus on core tasks in a first step to analyze the Commission as collective agent.

  3. 3.

    Actors are not necessarily individuals. Depending on one’s unit of analysis and ontological predispositions, they can also be, for example, political parties in a parliament, (inter)national organizations/bureaucracies (or even sub-units thereof) or states striving for influence in international politics. Either way, actors need to take decisions together by formal or informal accord to change the status quo ante for conflict to become relevant.

  4. 4.

    The Commission was, of course, deconstructed into individual DGs outside the principal–agent model much earlier. For example, it has been described as a ‘multi-organization’ (Cram 1994) riddled by conflict among its DGs (Cini 1996; Nugent 1995). Christiansen (2001) underlined the difficulty of achieving internal coherence both vertically (between Commissioners and their DGs) and horizontally (policy co-ordination). Carbone finds that Commission leadership is ‘seriously damaged by its internal divisions’ (2007: 22), which is a point shared by Dür (2007: 850).

  5. 5.

    Damro actually uses the term multiple agents. Since the various Commission DGs are bound by one and the same organizational structure, this is infelicitous. Damro dealt with the Commission as collective agent following the definition given above.

  6. 6.

    Elsig (2011) studied ‘complex agents’ which means sovereign principals (e.g., national ministers) delegating authority to proximate principals (e.g., ambassadors). Graham (2014: 370) argues that agent fragmentation inhibits the principal’s control mechanisms of oversight, screening and sanctioning because of the agent ‘pull[ing] in different directions.’ Both contributions seem to me to be concerned more with the principal side of the equation. Moreover, while Graham focuses on geographic dispersion of plural agents (global-, regional- and country-level offices), I highlight functional dispersion (e.g., environment, industry, trade) in this chapter. Recently, agent strategies have received increasing attention, such as their ability to exacerbate information asymmetry and form strategic interactions with non-principal third parties (Gastinger 2016).

  7. 7.

    The process behind trade negotiations is as follows. First, the Commission conducts exploratory discussions with the negotiating partner(s). Second, the Commission prepares draft negotiating directives. Third, member states adopt a negotiating mandate. Fourth, negotiations proper are conducted by the Commission assisted by Council bodies such as the Trade Policy Committee.

  8. 8.

    Due to space limitations, the wider context of the BTAs cannot be presented. Please refer to my doctoral thesis (Gastinger 2014), which is available upon request, for contextual information.

  9. 9.

    Or traces of conflict can simply no longer be found in the archival material because it was reduced to intra-DG conflict, which is too specific to be documented even in archival sources.

  10. 10.

    Trade did not constitute a separate DG at the time.

  11. 11.

    Non-trade aspects of EU-CARIFORUM relations were governed by the Cotonou Partnership Agreement.

  12. 12.

    In addition to sectoral DGs, there exist a number of horizontal DGs such as budget, personnel, the legal service and the Secretariat-General, all of which provide cross-cutting services throughout the Commission and thus carry less potential for conflict. Between sectoral and horizontal DGs lie thematic DGs such as DG Internal Market or DG Competition, which combine features of both types.

  13. 13.

    In some procedural matters and to request the Commission to prepare studies or proposals, the Council can take decisions by simple majority as well.


Cited Interviews


Commission official

September 2013



Commission official

September 2013



  1. Agence Europe. (1998a, July 4). Commission expected to adopt draft mandates for trade liberalisation on Wednesday (despite commissioner Fischler’s reservations). Europe Daily Bulletins, No. 7256.Google Scholar
  2. Agence Europe. (1998b, July 23). Commission recommendation to council for approval to negotiate association comprising free trade area, enhanced cooperation, political and security partnership. Europe Daily Bulletins, No. 7268.Google Scholar
  3. Agence Europe. (1998c, July 9). European commission postpones until 22 July its proposals on future trade relations in order to assess agricultural repercussions. Europe Daily Bulletins, No. 7259.Google Scholar
  4. Agence Europe. (1998d, July 20). France calls for a “clear signal” from council over Commission’s announced initiative over Mercosur. Europe Daily Bulletins, No. 7267.Google Scholar
  5. Agence Europe. (1998e, July 23). Majority of agriculture ministers express concerns and reluctance regarding plans for a free trade area with Mercosur. Europe Daily Bulletins, No. 7268.Google Scholar
  6. Agence Europe. (1998f, April 23). Ministers call for caution on agricultural concessions to third countries – Firm stance by French minister. Europe Daily Bulletins, No. 7206.Google Scholar
  7. Bailer, S. (2014). ‘An agent dependent on the EU member states? The determinants of the European Commission’s legislative success in the European Union. Journal of European Integration, 36(1), 37–53.CrossRefGoogle Scholar
  8. Bergman, T., Müller, W., & Strøm, K. (2000). Introduction: Parliamentary democracy and the chain of delegation. European Journal of Political Research, 37(3), 255–260.Google Scholar
  9. Carbone, M. (2007). The European Union and international development: The politics of foreign aid. London: Routledge.Google Scholar
  10. Christiansen, T. (2001). Intra-institutional politics and inter-institutional relations in the EU: Towards coherent governance? Journal of European Public Policy, 8(5), 747–769.CrossRefGoogle Scholar
  11. Cini, M. (1996). The European Commission: Leadership, organisation, and culture in the EU administration. Manchester: Manchester University Press.Google Scholar
  12. Commission. (1971a). Note à l’attention de MM. les membres de la Commission. SEC (71) 1423, HAEC, BAC 3/1978/880, 51.Google Scholar
  13. Commission. (1971b). Projet de communication de la Commission au Conseil concernant les relations entre: I. La Communauté et l’Inde II. La Communauté et l’Iran III. La Communauté et le Pakistan. SEC(71) 2922, HAEC, BAC 48/1984/753, 106–125.Google Scholar
  14. Commission. (1971c). Projet de communication de la Commission au Conseil concernant les relations entre la Communauté et l’Inde. SEC(71) 1747, HAEC, BAC 48/1984/753, 26–36.Google Scholar
  15. Commission. (2016). Distribution of active officials and temporary agents by DG and gender. Statistical Bulletin. Retrieved from (consulted September 2016).
  16. Conceição-Heldt, E. (2011). Variation in EU member states’ preferences and the commission’s discretion in the Doha Round. Journal of European Public Policy, 18(3), 403–419.CrossRefGoogle Scholar
  17. Conceicão-Heldt, E. (2017). Multiple principals’ preferences, types of control mechanisms, and agent’s discretion in trade negotiations. In T. Delreux & J. Adriaensen (Eds.), The principal–agent model and the European Union (pp. 203–226). London: Palgrave MacMillan.Google Scholar
  18. Coremans E., & Kerremans B. (2017). Agents as information asymmetry managers in EU trade policy-making. In T. Delreux & J. Adriaensen (Eds.), The principal–agent model and the European Union (pp. 227–253). London: Palgrave MacMillan.Google Scholar
  19. Council. (2002). Press release: 2419th Council meeting, Agriculture. Doc. 7097/02. Retrieved March 18, from (consulted September 2016).
  20. Cram, L. (1994). The European Commission as a multi-organization: Social policy and IT policy in the EU. Journal of European Public Policy, 1(2), 195–217.Google Scholar
  21. Damro, C. (2007). EU delegation and agency in international trade negotiations: A cautionary comparison. Journal of Common Market Studies, 45(4), 883–903.CrossRefGoogle Scholar
  22. De Bièvre, D., & Eckhardt, J. (2011). Interest groups and EU anti-dumping policy. Journal of European Public Policy, 18(3), 339–360.CrossRefGoogle Scholar
  23. Dehousse, R. (2008). Delegation of powers in the European Union: The need for a multi-principals model. West European Politics, 31(4), 789–805.CrossRefGoogle Scholar
  24. Delreux, T., & Kerremans, B. (2010). How agents weaken their principals’ incentives to control: The case of EU negotiators and EU member states in multilateral negotiations. Journal of European Integration, 32(4), 357–374.CrossRefGoogle Scholar
  25. Delreux, T., & Adriaensen, J. (2017). Introduction. Use and limitations of the principal–agent model in studying the European Union. In T. Delreux & J. Adriaensen (Eds.), The principal–agent model and the European Union (pp. 1–34). London: Palgrave MacMillan.Google Scholar
  26. DG I. (1971). Note à l’attention de Monsieur Hijzen: Projet de décision du Conseil autorisant la Commission à ouvrir des négociations avec l’Inde au sujet d’un accord de coopération commerciale. HAEC, BAC 3/1978/880, 180–182.Google Scholar
  27. DG I. (1975, July 14). Record of meeting with representatives of the Chinese embassy. HAEC, BAC 379/1991/49, 195–198.Google Scholar
  28. DG I. (1977a). Note à l’attention de Messieurs les membres de la Commission. HAEC, BAC 379/1991/51, 243–254.Google Scholar
  29. DG I. (1977b). Note à l’attention de Monsieur CASPARI Directeur Général Adjoint: COREPER du 10 novembre Recommandation de la Commission au Conseil concernant l’ouverture de négociations avec la République Populaire de Chine. HAEC, BAC 379/1991/51, 54–55.Google Scholar
  30. DG I. (1977c). Note à l’attention de Monsieur HAFERKAMP, Vice-Président de la Commission: Chine-Communauté Proposition au Conseil en vue de la détermination des directives d’un accord commercial. I/S-4/77, HAEC, BAC 379/1991/51, 225.Google Scholar
  31. DG I. (1977d). Note à l’attention de Sir Roy DENMAN: Projet de recommandation de la Commission au Conseil concernant l’ouverture de négociations avec la Chine. HAEC, BAC 379/1991/51, 239.Google Scholar
  32. DG I. (1977e). Note de Dossier: Chine. HAEC, BAC 379/1991/51, p. 194.Google Scholar
  33. DG I. (1978). Exploratory discussions with a view to possible negotiations with (1) India and (2) ASEAN. HAEC, BAC 367/1991/76, 207–209.Google Scholar
  34. DG VII. (1977). Note for Sir Roy DENMAN, Director General, DG I: Negotiations with China: Shipping. HAEC, BAC 379/1991/51, 56.Google Scholar
  35. DG XI. (1971a). Note à l’attention de Monsieur le Directeur Général Much [du] Service Juridique: Projet de décision du Conseil autorisant la Commission à ouvrir des négociations avec l’Inde au sujet d’un accord de coopération commerciale. HAEC, BAC 48/1984/770, 237.Google Scholar
  36. DG XI. (1971b). Note à l’attention de Monsieur le secrétaire général: Projet de communication de la Commission au Conseil concernant les relations entre la Communauté et l’Inde. HAEC, BAC 48/1984/753, 37–48.Google Scholar
  37. DG XI. (1971c). Note pour la Direction Générale des Relations Exterieur [et] la Direction Gestion de l’Union Douaniere: Projet de décision du Conseil autorisant la Commission à ouvrir des négociations avec l’Inde au sujet d’un accord de coopération commerciale. HAEC, BAC 3/1978/880, 157–179.Google Scholar
  38. DG XI. (1971d). Vermerk für Herrn Professor Dahrendorf: Beziehungen zu Indien. HAEC, BAC 48/1984/753, 49–51.Google Scholar
  39. DG XI. (1972a). Projet de recommendation de décision du Conseil autorisant la Commission à ouvrir des négociations avec l’Inde au sujet d’un accord de coopération commerciale. HAEC, BAC 48/1984/753, 241–270.Google Scholar
  40. DG XI. (1972b). Vermerk fuer Herrn Professor Dahrendorf: Entwurf einer Empfehlung an den Rat betreffend die Eröffnung von Verhandlungen mit Indien zwecks Abschluss eines Abkommens über handelspolitische Zusammenarbeit. HAEC, BAC 48/1984/753, 272–298.Google Scholar
  41. DG XI. (1972c). Vermerk für Herrn Professor Dahrendorf: Entwurf einer Empfehlung an den Rat betreffend die Eröffnung von Verhandlungen mit Indien zwecks Abschluss eines Abkommens über handelspolitische Zusammenarbeit. HAEC, BAC 48/1984/753, 331.Google Scholar
  42. Dijkstra, H. (2017). Non-exclusive delegation to the European External Action Service. In T. Delreux & J. Adriaensen (Eds.), The principal–agent model and the European Union (pp. 55–81). London: Palgrave MacMillan.Google Scholar
  43. Dür, A. (2007). EU trade policy as protection for exporters: The agreements with Mexico and Chile. Journal of Common Market Studies, 45(4), 833–855.CrossRefGoogle Scholar
  44. Dür, A., & Elsig, M. (2011). Principals, agents, and the European Union’s foreign economic policies. Journal of European Public Policy, 18(3), 323–338.CrossRefGoogle Scholar
  45. Elgström, O., & Larsén, M. (2010). Free to trade? Commission autonomy in the economic partnership agreement negotiations. Journal of European Public Policy, 17(2), 205–223.CrossRefGoogle Scholar
  46. Elsig, M. (2011). Principal–agent theory and the World Trade Organization: Complex agency and “missing delegation”. European Journal of International Relations, 17(3), 495–517.CrossRefGoogle Scholar
  47. Gastinger, M. (2014). Negotiating bilateral trade agreements in the European Union: Commission autonomy and member state control (Unpublished PhD Thesis). Florence: European University Institute.Google Scholar
  48. Gastinger, M. (2016). The tables have turned on the European Commission: The changing nature of the pre-negotiation phase in EU bilateral trade agreements. Journal of European Public Policy, 23(9), 1367–1385.CrossRefGoogle Scholar
  49. Graham, E. (2014). International organizations as collective agents: Fragmentation and the limits of principal control at the World Health Organization. European Journal of International Relations, 20(2), 366–390.CrossRefGoogle Scholar
  50. Hawkins, D., & Jacoby, W. (2006). How agents matter. In D. Hawkins, D. Lake, D. Nielson, & M. Tierney (Eds.), Delegation and agency in international organizations (pp. 199–228). Cambridge: Cambridge University Press.CrossRefGoogle Scholar
  51. Helwig, N. (2017). Agent interaction as a source of discretion for the EU High Representative. In T. Delreux & J. Adriaensen (Eds.), The principal–agent model and the European Union (pp. 105–129). London: Palgrave MacMillan.Google Scholar
  52. Hix, S. (2005). The political system of the European Union (2nd ed.). Basingstoke: Palgrave Macmillan.Google Scholar
  53. Kassim, H., & Menon, A. (2003). The principal–agent approach and the study of the European Union: Promise unfulfilled? Journal of European Public Policy, 10(1), 121–139.CrossRefGoogle Scholar
  54. Kerremans, B. (2004). What went wrong in Cancun? A principal–agent view on the EU’s rationale towards the Doha development round. European Foreign Affairs Review, 9(3), 363–393.Google Scholar
  55. Kiewiet, R., & McCubbins, M. (1991). The logic of delegation: Congressional parties and the appropriations process. Chicago: University of Chicago Press.Google Scholar
  56. Kostanyan, H. (2014). The rationales behind the European External Action Service: The principal–agent model and power delegation. Journal of Contemporary European Research, 10(2), 166–183.Google Scholar
  57. Kroll, D. A. (2017). Manifest and latent control on the Council by the European Council. In T. Delreux & J. Adriaensen (Eds.), The principal–agent model and the European Union (pp. 157–180). London: Palgrave MacMillan.Google Scholar
  58. Kurpas, S., Grøn, C., & Kaczyński, P. (2008). The European Commission after enlargement: Does more add up to less? CEPS Special Report.Google Scholar
  59. Laloux, T. (2017). Designing a collective agent for trilogues in the European Parliament. In T. Delreux & J. Adriaensen (Eds.), The principal–agent model and the European Union (pp. 83–103). London: Palgrave MacMillan.Google Scholar
  60. Larsén, M. (2007). Trade negotiations between the EU and South Africa: A three-level game. Journal of Common Market Studies, 45(4), 857–881.CrossRefGoogle Scholar
  61. Legal Service. (1977). Note à la direction générale des relations extérieurs: Projet de recommandation de la Commission au Conseil concernant l’ouverture de négociations avec la Chine. HAEC, BAC, 379(1991/51), 240–242.Google Scholar
  62. McCubbins, M., & Schwartz, T. (1984). Congressional oversight overlooked: Police patrols versus fire alarms. American Journal of Political Science, 28(1), 165–179.CrossRefGoogle Scholar
  63. Metcalfe, L. (2000). Reforming the Commission: Will organizational efficiency produce effective governance? Journal of Common Market Studies, 38(5), 817–841.CrossRefGoogle Scholar
  64. Moe, T. (1984). The new economics of organization. American Journal of Political Science, 28(4), 739–777.CrossRefGoogle Scholar
  65. Nielson, D., & Tierney, M. (2003). Delegation to international organizations: Agency theory and World Bank environmental reform. International Organization, 57(2), 241–276.CrossRefGoogle Scholar
  66. Niemann, A., & Huigens, J. (2011). The European Union’s role in the G8: A principal–agent perspective. Journal of European Public Policy, 18(3), 420–442.CrossRefGoogle Scholar
  67. Nugent, N. (1995). The leadership capacity of the European Commission. Journal of European Public Policy, 2(4), 603–623.CrossRefGoogle Scholar
  68. Nugent, N., & Saurugger, S. (2002). Organizational structuring: The case of the European Commission and its external policy responsibilities. Journal of European Public Policy, 9(3), 345–364.CrossRefGoogle Scholar
  69. Pollack, M. (2003). The engines of European integration: Delegation, agency, and agenda setting in the EU. Oxford: Oxford University Press.CrossRefGoogle Scholar
  70. Reichert, M., & Jungblut, B. (2007). European Union external trade policy: Multilevel principal–agent relationships. Policy Studies Journal, 35(3), 395–418.CrossRefGoogle Scholar
  71. Reykers, Y., & Beach, D. (2017). Process-tracing as a tool to analyse discretion. In T. Delreux & J. Adriaensen (Eds.), The principal–agent model and the European Union (pp. 255–281). London: Palgrave MacMillan.Google Scholar
  72. Strøm, K. (2000). Delegation and accountability in parliamentary democracies. European Journal of Political Research, 37(3), 261–290.Google Scholar
  73. Thatcher, M. (2011). The creation of European regulatory agencies and its limits: A comparative analysis of European delegation. Journal of European Public Policy, 18(6), 790–809.CrossRefGoogle Scholar
  74. Weingast, B., & Moran, M. (1983). ‘Bureaucratic discretion or congressional control? Regulatory policymaking by the Federal Trade Commission’. Journal of Political Economy, 91(5), 765–800.CrossRefGoogle Scholar

Copyright information

© The Author(s) 2017

Authors and Affiliations

  • Markus Gastinger
    • 1
  1. 1.Technische Universität DresdenDresdenGermany

Personalised recommendations