‘Political Executive’ or ‘Institutionalised Summitry’? Assessing the European Council’s Agenda Setting Capacity
A thesis submitted in partial fulfilment of the requirements for the degree of Master of Science (MSc.) in Public Administration
Daro E. Nakshbande
Faculty of Governance and Global Affairs, Leiden University
Date: 08/01/2020 Supervisor: Dr B.J. Carroll Second reader: Dr D.D. Toshkov
Abstract
Claims of the European Council’s ‘empowerment’ relative to the other European Union (EU) institutions are widespread in the post-Maastricht era. Yet, they rest on remarkably little empirical evidence—indicative of a gap in the public administration literature. This thesis addressed that gap by means of theory building. European Council encroachment on
European Commission competences was selected as a most likely case design of exposing the institution’s increasing prominence in the EU. The research design analysed to what extent the European Council agenda affects the European Commission agenda. To gauge the institutions’ agendas, new datasets of European Council ‘Conclusions’ and European Commission ‘Work Programmes’ were constructed for the time period 1999–2018. Each of these documents was dissected into individual (quasi-)sentences. The (quasi-)sentences were manually encoded by their content into one of the 21 ‘policy areas’ of the EU Policy Agendas Project codebook. Based on the datasets constructed, fractional logit and negative binomial regression models substantiated a modest but highly statistically significant association between the two institutions’ agendas. Public salience was also found to be causally related to the European Commission agenda. By contrast, the analysis found no association with the level of public support for EU membership, GDP growth, or unemployment rate. The results demonstrated that the European Council’s agenda setting capacity vis-à-vis the European Commission is more limited than had been assumed thus far.
Acknowledgements
First and foremost, I would like to express my gratitude to my supervisor Dr Brendan Carroll for his genuine commitment to supporting me in my research endeavour. Without his
constructive, incisive feedback and insightful remarks, I would not have been able to produce this thesis. I would also like to thank my second reader Dr Dimiter Toshkov for his
willingness to be involved in the examination of this thesis. My colleagues and friends at Leiden University and beyond deserve to be acknowledged for their supportiveness and companionship along the way. Last but not least, I would like to thank my mother, father, and sister for their invaluable emotional support throughout the process.
Table of Contents
Abstract ... ii
Acknowledgements ... iii
List of Tables ... v
List of Figures ... vi
List of Abbreviations ... vii
Introduction ... 1
Why Study the European Council’s Agenda Setting Capacity? ... 4
Overview ... 5
Literature Review ... 6
Theoretical Foundations of Agenda Setting ... 6
Application to the European Council and European Commission ... 18
Gaps in the Literature... 22
Theoretical Framework ... 24
European Council and European Commission: (Non-)Hierarchical Agenda Setters?... 24
A New Approach to Inter-Institutional Agenda Setting ... 28
Research Design ... 38 Data Collection ... 38 Operationalisation ... 44 Method ... 52 Results ... 56 Discussion... 75
Interpreting Key Findings ... 75
Theoretical Implications ... 81
Limitations of the Research ... 83
Conclusion ... 86
References ... 91
Appendix A: List of Sources Consulted to Construct the Datasets ... 100
Appendix B: Summary of Main Policy Areas in the EU Policy Agendas Codebook ... 109
List of Tables
Table 1 Intertemporal Pearson R correlation coefficients per main policy area…... 61 Table 2 Fractional logit regression of the inter-institutional transmission
of legislative priorities………. 66
Table 3 Fractional logit regression with lagged covariates………... 70 Table 4 Negative binomial regression of the inter-institutional
transmission of legislative priorities……….... 72
Table 5 Summary of the key findings per hypothesis………... 81
List of Figures
Figure 1 Composition of the European Council Conclusions over time………. 57 Figure 2 Composition of the European Commission Work Programmes
over time……….... 57
Figure 3 Time-series of the inter-institutional convergence of legislative
priorities………... 59
Figure 4 Percentage share of attention for Health……….... 63 Figure 5 Percentage share of attention for Immigration……….. 63 Figure 6 Percentage share of attention for Regional and Urban
Policy and Planning………... 64
Figure 7 Percentage share of attention for Defence………. 64 Figure 8 Frequency of mentions of Agriculture and Fisheries Policy………. 64 Figure 9 Frequency of mentions of International Affairs and
Foreign Aid……… 64
Figure 10 Average marginal effect of European Council attention
per policy area……… 68
Figure 11 Average marginal effect of perceived public salience per
policy area………..……… 68
Figure 12 Marginal effect of European Council mentions on the N of
Commission mentions………... 73
Figure 13 Marginal effect of perceived public salience on the N of
List of Abbreviations BIC Bayesian information criterion CAP Comparative Agendas Project
CFSP Common Foreign and Security Policy
COM European Commission
CSDP Common Security and Defence Policy EMU Economic and Monetary Union
EU European Union
EUCO European Council FAC Foreign Affairs Council GAC General Affairs Council
HR/VP High Representative of the Union for Foreign Affairs and Security Policy
HSG Head of State or Government JHA Justice and Home Affairs TEU Treaty on European Union
TFEU Treaty on the Functioning of the European Union
UK United Kingdom
‘Political Executive’ or ‘Institutionalised Summitry’? Assessing the European Council’s Agenda Setting Capacity
Of all the European Union’s (EU) institutions, the European Council receives by far the most extensive media coverage. European Council summits, at which the heads of state or government (HSGs) of EU member states convene, have “reached a level of visibility that a Commission, Council or Parliament meeting could never expect” (De Schoutheete, 2012, p. 38). Over the course of the past decade in particular, the European Council has acquired unprecedented visibility. It routinely appeared at the forefront of the EU’s crisis responses, including the 2009–2012 Eurozone crisis, the 2015–16 Schengen crisis, and the Brexit negotiations with the United Kingdom (UK) (Dinan, 2017, p. 158). During these crises, the EU-28 primarily used the European Council to forge consensus on contentious policy issues (Wessels, 2016, p. 4). Moreover, whenever European Council meetings did culminate in an agreement, this agreement later proved decisive for the nature of the EU’s policy response (Wessels, 2016, p. 194).
These observations notwithstanding, the European Council’s pre-eminence in EU decision-making is not contingent on the occurrence of these crises. Rather, the stature of the European Council in the post-Maastricht era is the consequence of a structural shift. The institution’s summits have diverted their attention “from long-term institutional questions to short-term challenges and the day-to-day governance” of the EU (Bickerton et al., 2015b, p. 306). This is reflected by the surge in detailed, specific policy interventions by the European Council, as well as by its “intensive involvement. . . in the minutiae” of EU policymaking (Bickerton, Hodson, & Puetter, 2015b, p. 30). Nor is the European Council’s importance limited to times of crisis. To the contrary, the institution has retained its clout after the crises waned (Bickerton, Hodson, & Puetter, 2015a).
At the same time, the European Council appears increasingly self-aware of its increased political ‘weight’. The European Council’s recent defiance of the European Parliament’s Spitzenkandidaten procedure through the appointment of Von der Leyen as Commission President exemplifies this trend (Cloos, 2019, p. 4).
Unsurprisingly, spectators therefore suggest the European Council is the EU’s ‘centre of political gravity’ (Puetter, 2012) or even its ‘de facto government’ (Carammia,
Timmermans, & Princen 2016). New intergovernmentalism is particularly vocal about the European Council’s growing role in the day-to-day governance of the EU (Becker, Bauer, Connolly, & Kassim, 2016, p. 1012). Scholars in this discipline claim that the institution regularly ventures beyond its formal institutional mandate (Foret & Rittelmeyer, 2014, p. 9). As a consequence, the European Council would have tipped the scales of the EU’s
institutional balance towards intergovernmentalism (Hodson & Puetter, 2019, p. 1166). Has the European Council become the “political executive” of the EU (Fabbrini, 2013, p. 1006), or has it remained little more than an “institutionalised summitry” (Wessels, 2016, p. 23)?
Various methods could be used to shed light on this question. Given the scarcity of public administration scholarship on the topic, theory building is a cornerstone of this thesis. It takes the approach of constructing a most likely case of exposing the European Council’s increasing prominence in EU policymaking. The approach focuses on the analysis of the European Council’s inter-institutional relations, or the patterns which govern its working relationship with other EU institutions. This method is particularly suited because claims of European Council ‘empowerment’ are intrinsically relational: asserting that the European Council has ‘gained’ competences implies the ‘loss’ of competences by other EU institutions. Whether the European Council has encroached on the European Commission is of particular interest. Indeed, the institutions’ day-to-day activities converge in one important respect: agenda setting (Becker et al., 2016, p. 1014).
This thesis hypothesises that if the European Council is indeed ‘empowered’ within the EU’s institutional power balance, it would impair the European Commission’s capacity to independently set the EU’s legislative agenda. The focus on agenda setting is justifiable as this competence is shared between the European Council and European Commission. The Treaty on European Union (TEU) stipulates that the European Council must “provide the Union with the necessary impetus for its development” and “define the general political directions and priorities thereof” (Art. 9b TEU). The European Council’s primary task is thus to paint the broad strokes of the EU’s legislative agenda and priorities. Similarly, the
European Commission “shall promote the general interest of the union and take appropriate initiatives to this end” (Art. 9d TEU). Yet, the European Council, unlike the European Commission, can neither suggest nor initiate legislation (Art. 9b TEU). Indeed, the European Commission retains a so-called formal monopoly on the right to initiate directives,
regulations, decisions, and other legislation (Art. 9d TEU). Furthermore, the Treaties stipulate that the European Commission “shall neither seek nor take instructions from any government or other institution” (Art 9d. TEU). This supposed division of labour between the European Council and European Commission in agenda setting is succinctly captured as follows: “Au Conseil européen, l’initiative politique, à la Commission, l’initiative politique et technique” (Eggermont 2012, p. 109).1
In practice, however, the lines between ‘political’ agenda setting in the European Council and ‘technical’ agenda setting in the European Commission are less clear-cut. The European Council frequently invites or calls on the Commission to initiate specific legislation or to speed up ongoing legislative procedures (Höing & Wessels, 2013; Wessels, 2016).
1 “In the European Council, the political initiative, in the Commission, the political and technical initiative”
By doing so, the European Council would be progressively encroaching on the European Commission’s competences. However, there is remarkably little empirical evidence available on the European Council’s capacity to (co-)determine the European Commission’s legislative priorities (Elias, 2019, p. 3). Therefore, the research question is: To what extent does the European Council affect the legislative priorities of the European Commission?
Why Study the European Council’s Agenda Setting Capacity?
The theoretical relevance of this research derives from the gap in the public administration literature on inter-institutional relations in the EU. Several models of the European Council–European Commission working relationship exist, including but not limited to: ‘a tandem driven and governed by the former’ (Werts, 2008, p. 54), ‘competitive cooperation’ (Bocquillon & Dobbels, 2014, p. 34), executive and secretariat (Höing & Wessels, 2013), or ‘the Commission as hidden principal’ and the European Council as agent (Wessels, 2016, p. 97). These existing models share two features: their simplicity and their weak empirical foundations. Currently, the available empirical evidence is limited to
anecdotal evidence, small-N comparative analyses, and case studies (Kreppel & Oztas, 2017, p. 1142). As a consequence, claims of a relative increase in the prominence of the European Council at the expense of the European Commission have been poorly substantiated. Given the centrality of the European Council in EU policymaking, it is imperative to clarify to what extent the European Council can determine the European Commission’s agenda. The answer to questions regarding the power balance between the EU institutions is not merely
institutional ‘tug-of-war’. Indeed, alterations in the “institutional balance of the EU” can— and already have—greatly affect(ed) the trajectory of EU integration (Hodson & Puetter, 2019, p. 1166; Wessels, 2016, p. 189).
By contrast, the societal relevance of this research is more indirect. It derives from the widespread public perception of a democratic deficit in the EU. This thesis analyses the responsiveness of the bureaucratic, technocratic European Commission to the priorities of democratically elected, politically legitimate HSGs in the European Council. Promoting transparency in inter-institutional relations may be a first step toward mitigating public perception of EU governance as obscure.
Overview
The remainder of this thesis is structured as follows. The next chapter will review the literature on the theoretical foundations of agenda setting. Moreover, it explores the literature on the relations between the European Council and European Commission. The subsequent chapter sets out a new theoretical framework for European Council–European Commission relations. Against this backdrop, the following chapter constructs a research design and specifies how the key variables are operationalised. The subsequent chapter reports the results of the analysis and is followed by a chapter dedicated to a critical analysis of the results. Suggestions for further research are provided in the concluding chapter.
Literature Review
This chapter’s overarching aim is to provide an overview of scholarly perceptions of the European Council’s agenda setting capacity vis-à-vis the European Commission. This chapter proceeds as follows. The first section explores the literature on the theoretical foundations of the concept of agenda setting and traces its conceptual and theoretical
innovations over time. The second section builds on the first by showing how contemporary literature on the European Council and European Commission continues to draw extensively from these original theoretical foundations. The literature review is concluded with a
summary of the weaknesses and gaps in the current agenda setting literature.
Theoretical Foundations of Agenda Setting
Throughout the public administration literature, the concept of agenda setting has been invoked to refer to two distinct processes (Kreppel and Oztas, 2017, p. 1120). Broadly speaking, agenda setting pertains either to the formation of the policy agenda in a given institutional venue, or to procedural and institutional capacities that facilitate the actual execution of that policy agenda. In the former understanding, agenda setting aims to account for the composition of policy agendas and policymakers’ allocation of attention (Princen, 2011, p. 929). It is preoccupied with questions of when, why, and how policy issues are constructed, framed, timed, and end up on the agenda of an institutional venue (e.g. Baumgartner & Jones, 2009, pp. 4–5; Princen, 2007, pp. 30–31; Princen, 2009;
Schattschneider, 1960, pp. 66–68). Thus, agenda setting is defined as the process by which “the set of all conceivable subjects or problems to which officials could be paying attention” is condensed into “the set that actually becomes the focus of attention” (Kingdon, 1984, p. 3).
By contrast, the latter understanding of agenda setting refers to the capacities of an actor to present another institution with a “take it or leave it offer” which is difficult if not impossible to amend or reject (Romer & Rosenthal, 1978, p. 2; Tsebelis, 2009, p. 13).
Understood in this sense instead, agenda setting encapsulates the power or capacity of an institutional agent to assert their own priorities over those of another institutional actor. This is done by “manipulating what is being voted on and in what order” and “[biasing] final outcomes in the direction of [their] own preferences” (Fiorina & Shepsle, 1989, p. 28). In this latter strand, the focus is on formal-legal, procedural, and institutional rules and the
consequences of agenda setting for legislative outcomes (Shepsle & Weingast, 1987, p. 101; Pollack, 2003, p. 7). Given the primary aim of this thesis to analyse inter-institutional agenda setting, alternatively viewed as the capacity of the European Council to affect the European Commission agenda, the institutionalist understanding of agenda setting is most relevant and will accordingly be explored for the remainder of this section.2
Pioneering research on agenda setting was preoccupied with institutional analysis of domestic legislatures, most notably the United States (US) House of Representatives and its legislative Committees (Baumgartner & Jones, 2009, pp. 107–11). Precisely because this early literature is crucial to understanding later applications of agenda setting in the EU-context, it is imperative to review its core tenets.
A recurrent theme in this early literature is that agenda setting and principal-agent theory are essentially two sides of the same coin. Both depart from rational choice
institutionalist and game theoretical premises. On the one hand, principal-agent theory explores the (unintended) consequences of an institution (principal) delegating authority and responsibility for specific tasks to another institution (agent) (Sappington, 1991, pp. 47–48). Principal-agent theory explains why delegation to agents with knowledge or expertise may be more efficient for a principal, but only insofar as the principal is able to incentivise or control the agent to perform the task the way they would have preferred, had they done it themselves.
2 An in-depth theoretical discussion of the alternative interpretation of agenda setting lies beyond the scope of
On the other hand, agenda setting is primarily concerned with how institutional agents, operating on behalf of these legislative principals, can exert influence by virtue of their capacity to structure the policy agenda. In other words, the implicit assumption in principal-agent models that principals are “active” and agents are “responsive” is inversed in early agenda setting literature (Fiorina & Shepsle, 1989, p. 28).
The debate in the early literature revolved around the implications of this reversed assumption. Fiorina and Shepsle (1989) distinguish in this theoretical debate between ‘first-generation’ and ‘second-‘first-generation’ agenda setting literature. Whereas articles belonging to the former strand only assume the agenda setter to behave strategically, contributions in the latter strand tend to assume other actors’ agency is also strategic (Fiorina & Shepsle, 1989, pp. 28–29). This is an oversimplification, which overlooks the fact that assumptions about (non)strategic behaviour were not abandoned in a neatly linear fashion. For many decades, literature relying on these assumptions remained just as prominent as those that had
progressed on these assumptions. Some more recent studies in fact retain core assumptions of the foundational literature (e.g., Tsebelis, 2000, pp. 54–55). Similarly, notable foundational arguments in the early literature already questioned the simplistic assumptions of
contemporary agenda setting models (e.g., McKelvey & Niemi, 1978, p. 2).
This thesis will continue using the distinction between first- and second-generation literature, albeit with a different way of categorising the literature. It acknowledges the obscurity of this division and categorises studies according to whether they addressed domestic legislative contexts (first-generation) or instead were geared towards the EU institutional context (second-generation). This approach is not only less arbitrary but also facilitates a more nuanced discussion of the literature ‘within’ the same generation.
First-generation models of agenda setting. A discernible trend in pioneering studies
of agenda setting is the incremental increase in the scope of agenda setting capacity that authors ascribe agents with. Initial contributions were relatively hesitant: they were based on simplistic assumptions and emphasised that agenda setting capacity was highly conditional and context dependent. This is exemplified by McKelvey (1976), who tentatively concludes that the agenda setting capacity of a so-called “chairman” (agent) matters, but only as long as several key assumptions are met. For example, the chairman has full control over the policy agenda, and the principal’s voting preferences are known, not indifferent, and ‘sincere’ as opposed to ‘sophisticated’ (McKelvey, 1976, p. 481). When those conditions are met, an agenda setter could influence the legislative outcome by constructing “an agenda which terminates at his actual ideal point” (McKelvey, 1976, p. 481). In other words, agenda setting capacity persists under conditions of “sincere voting” but diminishes as soon as the principal is assumed to behave strategically. A principal that engages in “sophisticated voting”, ‘quid pro quo’ legislative exchange, collusion, or coordination, constrains the agenda setting agent (McKelvey, 1976, p. 481). Even if these conditions are met, McKelvey argues that agenda setting capacity is—at best—weak and conditional in practice. Given simple majority voting rules, any alternative to the policy proposal tabled by the agenda setter could be introduced and could command a majority. (McKelvey, 1976, p. 481).
Similarly, Romer and Rosenthal (1978) depict agenda setting as conditional and contingent. These authors conceive of agenda setting as the coercive capacity of an agent to issue a credible threat to withhold legislation from their principal to address the status quo. Whether an agenda setter can present a “take-it-or-leave-it” policy proposal depends on how undesirable the status quo is perceived to be (Romer & Rosenthal, 1979, pp. 35–36). Romer and Rosenthal (1978) conclude that agenda setting capacity is greatest when the status quo is highly undesirable.
They assert that in such cases, the agenda setting agent’s threat to withhold legislation would whip reluctant principals to vote in favour of their lesser preferred outcome (Romer & Rosenthal, 1978, p. 35).
Subsequent scholarship made a concerted effort to develop models in which not only the agenda setting agent but also the principal was assumed to behave strategically.
Whilst assumptions such as multiple legislative actors, sophisticated voting, and strategic agency rendered agenda setting models increasingly complex, these models more closely resembled reality. Given the already feeble agenda setting capacity under earlier agenda setting models, one might think such complex models would leave even less scope for an agent to engage in agenda setting. To the contrary, these increasingly sophisticated models pushed the boundaries of the concept of agenda setting capacity. The agenda setting capacity of an institutional agent was no longer the exception and increasingly became the rule.
The most straight-forward argument which exemplifies this incremental development is that of Austen-Smith (1987). This seminal contribution posits that agenda setting agents are aware of the strategic (“sophisticated”) voting behaviour of their principal and reciprocate by means of engaging in “sophisticated” proposing when constructing an agenda (Austen-Smith, 1987, p. 1324). According to this logic, the agent exploits the structure of the
legislative preferences of its principal. By presenting policy proposals sequentially, the agent can guarantee the attainment of its most preferred outcome at all times. This, of course, is assuming that there is only one agenda setting agent who proposes only one policy proposal at a time (Austen-Smith, 1987, pp. 1328–1329).
Shepsle’s (1979) initial contribution to this scholarly debate contradicts that of Austen-Smith (1987). Shepsle denies that an agent’s agenda setting capacity derives entirely from the agent’s strategic exploitation of the principal’s configuration of preferences.
Whilst the likelihood of the agent attaining its preferred legislative outcome certainly “depends on the configuration of preferences”, Shepsle emphasises it “can only rarely be guaranteed by that configuration” (Shepsle, 1979, p. 28). In other words, Shepsle (1979) contends that an agent can have agenda setting capacity independently of factors such as a principal’s (sophisticated) voting patterns and (dishonest) policy preferences (Shepsle, 1979, p. 30). Contrary to Austen-Smith (1987), Shepsle (1979) thus attributes an agent’s agenda setting capacity to institutional and procedural rules governing “jurisdiction and amendment control” (Shepsle, 1979, pp. 28–29). Without these institutional and procedural rules, Shepsle (1979) argues, any stable majority for a single policy proposal would be “inherently
inexplicable” (Shepsle, 1979, p. 27). However, the importance of this contribution to the ongoing debate ought not to be overstated. Shepsle (1979) neither substantively engaged with, nor set out to disprove, the arguments of McKelvey (1976) and Romer and Rosenthal’s (1978) pioneering work.
By contrast, Shepsle and Weingast (1981) did not leave the core tenets of
McKelvey’s (1976) contribution intact. They contradicted McKelvey’s assertion that any legislative proposal could command a majority, which supposedly implied that “anything might happen” (Shepsle, Cox, Aldrich, Alt & Lupia, 2007, p. 2; Shepsle & Weingast, 1981, p. 514).3 Instead, Shepsle and Weingast (1981) claimed that a “convenor” acting as agenda setter constrains “the set of feasible alternatives to which the status quo is vulnerable”
(Shepsle & Weingast, 1981, p. 512). The conclusion reached by Shepsle and Weingast (1981) is that neither voting behaviour nor voting preferences among the legislative principal are decisive. Rather, the structure of “institutional arrangements” would restrain actors from introducing alternative policy proposals, from making amendments, or from choosing among alternatives that ultimately are most significant (Shepsle & Weingast, 1981, p. 515).
Consequently, even if both the agenda setting agent(s) and the legislative principal(s) behave strategically, agenda setting is still possible. Shepsle and Weingast (1981) attribute this to the constraints imposed by the institutional and procedural rules of a given legislature.
This counterintuitive yet persuasive corrective on earlier accounts of agenda setting provided considerable thrust to the first generation of agenda setting literature.
At its most articulate, this literature perceives of agenda setting as ‘gatekeeping’, or an agent’s power to keep policy issues off the legislative agenda of its principal. Denzau and Mackay’s (1983) work is an example of this strand of agenda setting literature. They define gatekeeping as the capacity to “propose no change in the status quo” as a consequence of which the current situation is retained (Denzau & Mackay, 1983, p. 752). An agent with “gatekeeping powers” performs a strong agenda setting role irrespective of the principal’s strategic or sophisticated voting.
Yet, this is only one variant of more sophisticated first-generation agenda setting models. Shepsle and Weingast (1984b) provide an equally sophisticated alternative
contribution. Their incisive account targets the supposition (e.g. in McKelvey, 1976; Romer & Rosenthal, 1978) that agenda setting capacity can be effectively undermined or bypassed by legislative principals if only they wish to do so (Shepsle & Weingast, 1984b, p. 219). These authors posit that a principal’s strategic agency does not provide them with an
advantage per se. Strategically voting against one’s rational preferences could backfire in the future, leading to “opportunism, reneging, and the difficulties of bargaining and policing” (Shepsle & Weingast, 1984b, p. 219). In other words, there always remains a degree of uncertainty on part of the principal regarding its ability to bypass or undermine the agenda-setting agent. As a consequence, the principal’s ability to push through their own preferences, encapsulated by the dictum “where there’s a will, there might be a way”, is much more constrained in practice.
Thus, Shepsle and Weingast (1984b) posit a powerful rebuttal to the supposition that an agenda setting agent has influence only insofar as “the relevant actors would have chosen that outcome without that institution” (McKelvey & Ordeshook, 1984, p. 183, quoted in Shepsle & Weingast, 1984b, p. 207).
This quotation reflects an implicit, incremental, yet significant development in the first-generation agenda setting literature. The rational choice institutionalist assumptions of functional delegation and efficient institutions are gradually relaxed, and agenda setting is treated increasingly seriously.4 Shepsle & Weingast (1984a) explicitly direct institutional analysis research in this direction. They conclude that agenda setting is the “most fruitful way” to study the causal effect of institutional and procedural rules on the capacity of agenda setters to affect the substance of legislative proposals (Shepsle & Weingast, 1984a, p. 71).
To recap, first-generation agenda setting models have incrementally become more sophisticated by relaxing several core assumptions. The empirical successes of the first-generation agenda setting literature were substantial. First-first-generation models gradually accommodated strategic, sophisticated behaviour among both principal(s) and agent(s). This augmented the explanatory power of agenda setting, which was increasingly less seen as an incidental, contingent feature of agents in political legislatures. Yet, the key assumption, namely that agenda setting is inextricably linked to hierarchical, principal-agent relations between two institutions, has been firmly retained. Only in second-generation agenda setting models were the insights of this early literature expanded to contexts other than a domestic political system with a single agenda setter. In the next sub-section, the focus lies on the offspring of first-generation agenda setting models.
4 This development runs parallel with the development of the ‘new institutionalism’ and the notion that
These include adaptations and modifications which aim to make sense of agenda setting in the complexity of the EU’s institutional landscape.
Second-generation models of agenda setting. The second-generation agenda setting
literature retained many of the rational choice institutionalist and principal-agent core
assumptions of the first-generation literature. Consequently, scholars of agenda setting in the EU context have predominantly studied the EU from a rational choice institutionalist and principal-agent analysis perspective. The distinguishing feature of second-generation agenda setting models is their specificity to the EU context. The central question addressed by these contributions is whether supranational EU institutions (the European Parliament, European Commission, and European Court of Justice) are an “independent causal variable” in EU politics and European integration (e.g. Pollack, 1997, p. 99). In this context, the agenda setting capacity of institutions is considered an indicator of the extent to which supranational agents matter in their own right, apart from the underlying preferences of their legislative principal (e.g. Moravcsik, 1999, p. 275; Tsebelis & Kreppel, 1998, p. 42). Due to the institutional complexity of the EU and the overlapping competencies of its institutions and the member states, principal-agent relations are more complex and less clear-cut than among legislative institutions in national polities (Pollack, 2003, pp. 43–45). Theoretical and
conceptual innovations in second-generation agenda setting literature were made specifically to provide meaningful interpretation of the EU’s complex institutional architecture.
One particularly important conceptual innovation, and a recurrent theme in the second-generation literature, is the more precise definition of agenda setting. Pollack coined the distinction between formal and informal agenda setting (Pollack, 1999; Pollack, 2003). Formal agenda setting closely resembles first-generation models of agenda setting.
According to Pollack (1999), it entails the formal and procedural right to determine
legislative procedures and to submit policy proposals to the Council that are easier to adopt than to amend or reject (Pollack, 1999, p. 121). By contrast, informal agenda setting
encapsulates the informal capacities to frame, direct, and influence the policy agenda without resort to “formal powers of initiative or any particular set of voting or amendment rules” (Pollack, 1999, p. 125).
The introduction of this conceptual distinction is consequential. In the EU context, an over-emphasis on formal agenda-setting, like in the first-generation agenda setting literature, would cause more subtle inter-institutional agenda setting dynamics to go unnoticed. Pollack utilises the distinction between formal- and informal agenda setting powers to capture the entire scope of agenda setting relations between the EU institutions. Beyond formal
competences, informal agenda setting captures alternative mechanisms of agenda setting. For example, the European Commission can exert influence by exploiting informational
asymmetries, “policy expertise” and “institutional persistence”. This despite the absence of a formal right to determine the substance of legislation on the Council agenda (Pollack, 1999, p. 102).
The formal-informal distinction also features in Moravcsik (1999), who interprets agenda setting as a mode of “informal supranational entrepreneurship” (Moravcsik, 1999, p. 270). For Moravcsik, the notion of informal agenda setting refers specifically to the capacity of supranational agents to exploit “asymmetrical control over informational and ideational resources unavailable to the principals” (Moravcsik, 1999, p. 272). Thus, Moravcsik contends that agenda setting capacity is not entirely the product of institutional rules and constraints, as argued by first generation literature. More important in the case of the EU are the differential organisational capacities between principal and agent (Moravcsik, 1999).
Despite the fact that the formal–informal agenda setting distinction is by far the most conventional one in the public administration literature on EU agenda setting, several
alternative specifications exist that are worth mentioning here. Tallberg (2003) primarily critiques the formal–informal distinction for its lack of specificity and the conceptual muddle it induces in analytically distinct types of agenda setting activities. Tallberg (2003) contends that his alternative specification addresses these shortcomings by distinguishing between three types of shaping”. Besides conventional agenda-setting, he includes “agenda-structuring” and “agenda exclusion” (Tallberg, 2003, p. 5). The former refers to the agent’s capacity to influence principals’ allocation of attention and prioritisation of certain policy issues over others. The latter pertains to the agent’s capability to exclude policy issues from appearing on the legislative agenda altogether (Tallberg, 2003, p. 5). Tallberg’s (2003) core argument is that attempts to downplay an issue (“the power of non-decision-making”) are as significant as efforts to push an issue higher up the agenda (Bachrach & Baratz, 1963, p. 632, in Tallberg, 2003, p. 11). Whereas Tallberg’s distinction provides a conceptually
sophisticated account of agenda setting within the Council of Ministers, it falls short of an explanation of complex inter-institutional agenda setting relations in the EU-context. Indeed, Tallberg’s account fails to elaborate how either type of agenda setting matters at different phases of the policymaking process and is primarily geared towards intra-institutional rather than inter-institutional agenda setting dynamics.
Tsebelis and Garrett (1996) provide an account which lacks some of Tallberg’s (2003) specificity but embeds agenda setting much more carefully in the intricate relations between supranational agents and member state principals. Tsebelis and Garrett (1996) perceive of agenda setting as a spectrum based on when in the policy process this capacity is exercised. On the one end of the spectrum, at the final stage of policymaking, agenda setting power enables an agent to structure the policy agenda of the principal.
At this point, procedural and institutional rules ban or preclude the principal from making amendments (Tsebelis and Garrett, 1996, p. 346). On the other end of the spectrum, at the very start of the policy process, “veto power” refers to the ability of an agent to maintain the status quo (Tsebelis and Garrett, 1996, p. 352).
Yet, this perspective on agenda setting capacity risks obscuring more than it reveals. Both ends of the spectrum are empirical extremes which do not resonate with more routine patterns of EU policymaking (Crombez, Groseclose, and Krehbiel, 2006, p. 331).
Crombez et al. (2006) effectively demonstrate that none of the EU institutions are in possession of a “gatekeeping right”. Thus, a qualification of the findings of Tsebelis and Garrett (1996) appears necessary. This is provided by Tsebelis (1994) and Tsebelis and Kreppel (1998), who stress that agenda setting power in the EU is inherently conditional (Tsebelis, 1994, p. 131; Tsebellis and Kreppel, 1998, pp. 41–42). The use of the qualifier ‘conditional’ refers to the fact that European Parliament initiatives or European Commission legislative proposals can be amended by the Council of Ministers (Tsebelis and Kreppel, 1998, pp. 41–42). In other words, the agenda setting power of these EU institutions is not absolute but “quasi-dictatorial”, contingent on the capacity to propose “a series of successive votes” (Tsebelis, 1994, p. 131). Moreover, Tsebelis (1994) emphasises the crucial role of impatience among the principal, which in the context of the EU can heighten the perceived cost of refusing legislative proposals in the eyes of EU member states (Tsebelis, 1994, p. 131). Such additional theoretical specifications explain why the second-generation agenda setting literature is considerably better at explaining EU agenda setting dynamics.
To conclude this section, the concept of agenda setting has received considerable attention in the context of EU institutional analysis. Stellar progress has been made in applying insights from classical agenda setting theory to explain agenda setting in the EU.
Of particular significance is the observation that none of these conceptual refinements is useful per se. Rather, the wide range of available alternative specifications of agenda setting is the most tangible progress made relative to the first generation of agenda setting literature. By virtue of the proliferation of agenda setting typologies and definitions, second generation agenda setting models are much more versatile in institutional analysis of the EU. Despite conceptual refinements in the second-generation, the foundational assumptions of agenda setting appear relatively intact until today. Agenda setting has remained an intrinsically rational choice institutionalist, game theoretical, and principal-agent concept.
Application to the European Council and European Commission
This section looks at how scholars have applied theoretical insights on agenda setting to the inter-institutional relations between the European Council and European Commission. Studies on the European Council and European Commission display a tendency to study either institution and to make inferences about the other institution on those grounds. For example, Dinan (2017) scrutinises leadership performance and agenda setting capacity of the European Council’s (permanent) President. Based on the contingencies of “situational context” and “leadership qualities” of the European Council President, Dinan (2017) claims that the European Council comprises “an informal opportunity for the European Commission President to provide political leadership” to that institution (Dinan, 2017, p. 160). In other words, Dinan makes an assertion with regards to the agenda setting capacity of the European Commission based on institutional characteristics of the European Council.
Likewise, Tömmel (2017) provides an assessment of the agenda setting relations between the European Council and European Commission by means of an analysis of the track record of the first permanent European Council President, Van Rompuy (Tömmel, 2017, p. 176).
Tömmel (2017) infers the European Council’s encroachment on the European Commission’s agenda setting capacity based on her analysis of the European Council Presidency and
whether it engages in “transformative leadership” by rallying consensus around Commission proposals, or in “transactional leadership” emphasising inter-state bargaining instead
(Tömmel, 2017, pp. 181–182). Tömmel hereby builds on the principal-agent theoretical foundations of first- and second-generation agenda setting models. With reference to a case study of the leadership performance of Van Rompuy in the European Council negotiations on the Multi-Annual Financial Framework (MFF), Tömmel (2017) concludes that the European Council–European Commission agenda setting relations are dominated by the former at the expense of the latter (Tömmel, 2017, p.181).
Analysis of the European Council Presidency is but one of several strands of
contemporary agenda setting literature on the European Council and European Commission. Alexandrova and Timmermans (2013) make inferences as to the agenda setting capacity of the European Council on the European Commission agenda. They perform a large-N
quantitative analysis of the impact of national executive speeches of the country holding the Presidency on the content of the Conclusions of subsequent European Council summits (Alexandrova & Timmermans, 2013, p. 319). Alexandrova and Timmermans (2013) find that neither permanent nor rotating Presidencies are willing or able to leave “a strong national mark on the European Council agenda” (Alexandrova & Timmermans, 2013, p. 331).
Therefore, Alexandrova and Timmermans (2013) conclude that the agenda setting role which the Presidency fulfils in the European Council deliberations vis-à-vis the European
Commission is not strictly hierarchical and intergovernmental. Agenda setting capacity is “attuned by the norms of neutrality, effectiveness, and consensus building” and constrained by political “urgency . . . domestic political instability, or stalemate” (Alexandrova & Timmermans, 2013, p. 331).
Here, too, can evidence be found of the influence of first- and second-generation agenda setting literature, since Alexandrova and Timmermans (2013) use the extent to which
relations conform to a principal-agent model as indicator of the agenda setting capacity of the European Council on the European Commission.
Carammia et al. (2016) study patterns in the allocation of attention to policy areas in the European Council and find evidence for a shift from a “selective intervention” to a “routine monitoring” model of attention allocation (Carammia et al., 2016, p. 822). In other words, whereas the European Council used to make highly selective, contingent, and ad hoc policy interventions, it would nowadays increasingly scrutinise policy areas across the board. This would indicate the European Council is encroaching on the independent, autonomous agenda setting capacity of the European Commission (Carammia et al., 2016, pp. 822–823). Carammia et al. (2016) thereby draw extensively from the informal variant of agenda setting capacity posited by second-generation agenda setting contributions.
Whereas both Alexandrova and Timmermans (2013) and Carammia et al. (2016) make inferences of relations with the European Commission based on analysis of the
European Council, Kreppel and Oztas (2017) provide an inverse account. Kreppel and Oztas (2017) analyse whether the “success rate” of the European Commission’s legislative acts is higher for those policy areas which it had marked as its legislative priority. Based on whether the substance of “legislative acts” (directives, regulations, and decisions) reflects the
European Commission’s initial preferences, Kreppel and Oztas (2017) make inferences about the European Council’s agenda setting capacity (Kreppel & Oztas, 2017, p. 1121). This analysis contains a distinguishable principal-agent element, as the European Commission’s agenda setting capacity is conceptualised as the discretion it has to pursue its own interests.
In contrast to the inferential method of these contributions, virtually no public administration research has looked simultaneously at both the European Council and the European Commission to scrutinise their agenda setting relationship (Elias, 2019, p. 142). There are only a handful of exceptions to this rule. Interestingly, the few contributions that do have found that principal-agent theory only partially explains inter-institutional agenda setting. For example, Thaler (2016) presents a case study of European Council–European Commission agenda setting relations in the policy area of EU energy policy. Rather than assuming “interinstitutional competition for power”, Thaler (2016) effectively demonstrates that in the case of energy policy, the two institutions mutually reinforce each other’s agenda setting (Thaler, 2016, p. 582). Whereas the European Council is in firm control of the agenda, the European Commission’s preparatory documents provide auxiliary “technical knowledge” and political expertise (Thaler, 2016, p. 579). In spite of a degree of hierarchy and principal-agent relations, Thaler (2016) concludes that the inter-institutional relations are characterised by “concerted action” rather than subservience of the European Commission (Thaler, 2016, pp. 579–580). However, a major limitation of Thaler’s (2016) case study research is the inability to generalise his findings to other policy areas.
By contrast, Alexandrova (2017) employs a large-N quantitative design and finds that European Council and European Commission relations in legislative agenda setting are manifestly not competitive but rather complementary (Alexandrova, 2017, p. 770). The divergent “functional roles” of these institutions leads them to hold unique dispositions to prioritise certain policy over others. Alexandrova (2017) calls this phenomenon “institutional issue proclivity”. Alexandrova (2017) finds that the European Council is preoccupied with policy areas with a low level of integration such as Defence, Foreign Affairs, and
Macroeconomics. This in contrast to the Commission’s focus on common market domains such as Agriculture and External Trade (Alexandrova, 2017, p. 765).
Although this division of labour is ultimately determined by principal-agent relations between the institutions, Alexandrova’s (2017) analytical framework defies the notion that this principal-agent agenda setting relationship is strictly hierarchical.
A final example is the contribution of Bocquillon and Dobbels (2014), who depart from the core assumption that the European Council and European Commission interact in ‘collaborative rather than antagonistic” ways. These authors contend that the institutions are in fact engaged in a continuous, “reciprocal relationship of joint agenda setting” (Bocquillon & Dobbels, 2014, p. 26). Indeed, Bocquillon and Dobbels (2014) assert that the European Commission and European Council are “joint agenda setters” rather than principal and agent (Bocquillon & Dobbels, 2014, p. 27). However, in terms of the provision of political
“impetus” to legislative initiatives, a degree of hierarchy still exists between the European Council and European Commission. Both the “principal-agent” and “joint agenda setting” models are ideal types. Bocquillon and Dobbels (2014) conclude based on three case studies that in reality, European Council–European Commission agenda setting relations consist of a mix of both hierarchy and cooperation (Bocquillon & Dobbels, 2014, p. 34).
Gaps in the Literature
To conclude this chapter, it appears that we might have arrived at a problematic situation. The first section of this chapter demonstrated the intrinsic link between agenda setting and principal-agent theory. The second section demonstrated that these theoretical foundations are still largely intact in contemporary agenda setting analyses of the European Council and European Commission, albeit in modified form to fit the specificities of the EU. Yet, the small subset of this literature which is interested specifically in the former’s inter-institutional agenda setting capacity over the latter, has found that principal-agent theory is at best a partial explanation of the two institutions’ simultaneous engagement in agenda setting.
In short: agenda setting as a concept hinges on hierarchical, unidirectional principal-agent theoretical foundations, which may in fact not hold up to empirical scrutiny in the case of the European Council and European Commission.
Therefore, we may need to entertain the possibility that existent theoretical
frameworks to study agenda setting are simply not fit for purpose. Contemporary scholarship has generally not demonstrated awareness of this paradox in research of the agenda setting relations between these institutions. Thus, there appear to be two interrelated gaps in the public administration literature on agenda setting. First, the scarcity of substantive empirical evidence available to underpin contemporary models and patterns of European Council– European Commission relations in agenda setting. Second, the uncertainty with regards to the suitability of a principal-agent framework to study the two institutions’ interactions. It is against this backdrop that the next chapter seeks to posit a new theoretical framework which is capable of making a contribution toward filling these gaps in the current literature.
Theoretical Framework
The previous chapter concluded that contemporary agenda setting literature is at best unsure as to what extent the European Council and European Commission interact according to principal-agent theory. If the European Council and European Commission do not behave as principal and agent in practice, employing an agenda setting theoretical framework would constitute a mismatch given the intrinsic link between agenda setting and principal-agent theory. An analysis of the agenda setting capacity of the European Council on the European Commission can only be theoretically valid and empirical reliable insofar as it is premised on a theoretical framework which corresponds with empirics. Therefore, this chapter is
structured as follows. The first section compares the de jure and de facto relations between the European Council and European Commission. It analyses whether the European Council and European Commission do indeed act as principal and agent. The second section, based on the revelations of the two institutions’ working relationship, posits a suitable theoretical framework which captures how the two institutions behave in practice rather than in theory. From this theoretical framework, specific, testable hypotheses are subsequently derived.
The European Council and European Commission: (Non-)Hierarchical Agenda Setters?
This section contrasts a strictly de jure reading of the European Council–European Commission relations with an interpretation of their de facto interactions.
Formal relations. At first sight, a strict interpretation of the legal provisions on the
institutions’ mandates appears incompatible with a hierarchical, unidirectional, principal-agent agenda setting relation (Höing & Wessels, 2013, p. 131). To recall, the European Council’s mandate is to “provide the Union with the necessary impetus for its development” and to “define the general political directions and priorities thereof” (Article 15(1) TEU). By contrast, the European Commission’s formal mandate is to “promote the general interest of the Union and take appropriate initiatives to that end” (Article 17 (1) TEU).
As “guardian of the treaties”, the European Commission initiates and proposes legislation to the co-legislators, the Council of Ministers and the European Parliament (Werts, 2008, p. 46). The European Commission has the formal monopoly on legislative initiatives in the EU: “Union legislative acts may only be adopted on the basis of a Commission proposal, except where the treaties provide otherwise” (Article 17(2) TEU).
Several provisions seek to ensure the European Commission’s independence in the exercise of its agenda setting capacity. The European Council is explicitly barred from performing “legislative functions” (Article 15(1) TEU). Furthermore, the Commission is expressly forbidden to engage in any way with “instructions from any other institution, body, office, or entity” (Article 17(3) TEU). From comparing the two institutions’ mandates, it would appear that the European Council is forbidden from having an explicit agenda-setting capacity on the European Commission.
However, a deeper reading of the EU Treaties reveals a more ambivalent picture. In fact, the EU Treaties do not unequivocally prohibit the European Council from affecting if not determining the European Commission agenda. Eggermont (2012) points out that Article 245 of the Treaty on the Functioning of the European Union (TFEU) precludes any
instruction from having a binding effect on the European Commission. This provision notwithstanding, Article 241 TFEU still grants the Council of Ministers, European
Parliament, and European Council the prerogative to “intervene in all issues of Union policy” and to “influence the contents of Commission proposals” (Eggermont, 2012, pp. 105–110). The European Council communicates calls for action through its Conclusions, which
summarise which decisions and political priorities the HSGs have determined at the summit. Höing and Wessels (2013) report that the European Council makes frequent use of this prerogative by using its Conclusions to call upon the Commission to initiate legislation (Höing & Wessels, 2013, p. 134).
Yet, this observation is not evidence of a principal-agent relationship per se. Of greater significance is Werts’ (2008) remark that the letter of the law essentially allows the European Council to go considerably beyond providing ‘political impetus’ or ‘defining general political directions and priorities’. Indeed, Werts (2008) asserts that the European Council Conclusions increasingly tend to “delve into the ‘nitty gritty’, the complicated technical details” (Werts, 2008, p. 52). It is this legal capacity of the European Council to issue increasingly specific demands and calls for action which would “undermine the independent position of the Commission” (Werts, 2008, pp. 46–47).
The principal-agent relationship between European Council and European
Commission is exemplified in the fact that the Conclusions are not only politically binding, but also appear to carry a “a legal status in statu nascendi” because they “aim at the creation of Community law through the action of the Institutions” (Werts, 2008, p. 29). In other words, Conclusions are considerably more than minutes of European Council summits. They comprise unanimous, politically and legally binding pledges of the EU-28’s HSGs to support and abide by legislative initiatives proposed by the European Commission (Werts, 2008, p. 29). Moreover, Conclusions are “subjected to . . . rulings of the Court [of Justice of the EU]”, which has the competency to rule on their “legality” based on Article 269 TFEU (Höing & Wessels, 2013, p. 125), meaning they are far from insignificant. European Council
Conclusions are widely perceived and “treated as binding ‘framework-laws’ to be
implemented by the Commission” (Morgan, 1976, p. 50, in Kleine, 2013, p 64). This would constitute a plausible explanation for why “none of the [EU] institutions has ever refused to follow up a demand of the European Council” (Werts, 2008, p. 27). It also justifies Kleine’s (2013) claim that in spite of the Treaty provisions, the letter of the law leaves sufficient leeway for HSGs to effectively “predetermine the [Commission’s] legislative agenda” (Kleine, 2013, p. 64).
In short: the Treaties are strictly speaking not incompatible with a principal-agent relationship whereby the European Council considerably affects if not (co-)determines the European Commission’s agenda. To the contrary: they seem to facilitate such a hierarchical relationship between the two institutions in their agenda setting capacities.
Informal relations. Yet, the de facto relations between these EU institutions
challenge the picture of hierarchical, unidirectional, principal-agent agenda setting dynamics (Kleine, 2013, p. 68). Two empirical regularities in particular ought to be scrutinised here.
First, the Commission is not merely a passive recipient of the European Council agenda, but actively participates in shaping that agenda, too. Indeed, the European Commission tends to be closely involved in preparations for European Council summits alongside the Council Secretariat and the General Affairs Council (GAC) (Höing & Wessels, 2013, p. 1390). Pursuant Article 15(6b) TEU, the President of the Commission is to be consulted by the permanent President of the European Council in drafting the policy agenda in a spirit of “mutual sincere cooperation” (Article 13(2) TEU). The Commission regularly submits ‘communications’ in advance of the summits to acquire approval of the HSGs to proceed with legislative action in specific policy domains (Eggermont, 2012, p. 79).
Second, the European Commission President is not only subservient to, but also strategically exploits the European Council. Indeed, the Commission President makes strategic use of its seat at the table of the European Council summits as an “effective vehicle to present [the Commission’s] own priorities” (Werts, 2008, p. 53). Werts (2008) suggests that although the European Council’s political activism seemingly undermines the European Commission’s right to initiate legislation, the European Commission has in fact acquired greater political leverage (Werts, 2008, pp. 52–53). Precisely because of the proximity of the European Commission to the European Council, the former can use the latter as an
Once endorsed by the European Council, proposals submitted by the European Commission become more difficult for the Council of Ministers and Parliament to amend or reject (Werts, 2008, p. 53). A similar argument is posited by Nugent and Rhinard (2016), who assert that endorsement of Commission “position or discussion papers” operates as mechanism of “legitimization” (Nugent & Rhinard, 2016, pp. 1200–1201).
James and Copeland (2014) extend this theoretical supposition. They provide a full analytical framework of “reversed” principal-agent relations between the European Council and European Commission. James and Copeland (2014) argue that the European Commission engages in reversed delegation. This would involve returning “responsibility for policy initiation” to the HSG principals in the face of “powerful political constraints” down the line of EU policymaking. (James & Copeland, 2014, p. 524). Precisely by exploiting its proximity to the European Council to encourage this institution to take political initiatives, the
European Commission can mobilise support from HSGs for its preferred initiatives early. Endorsement is then used as leverage in negotiations with the Council of Ministers and the European Parliament down the line (James & Copeland, 2014, p. 524).
A New Approach to Inter-Institutional Agenda Setting
As established in the previous section, the relations between the European Council and European Commission are twofold. On the one hand, the institutions operate on a principal-agent, hierarchical, unidirectional basis in setting the EU’s legislative agenda. In this reading, the European Council as principal sets the agenda to be pursued by the European Commission as agent. On the other hand, the two institutions interact in a non-competitive, non-hierarchical, complementary relationship. This alternative reading suggests that the influence is bidirectional: the European Council’s legislative priorities affect those of the European Commission and vice versa.
The contradictory nature of legal provisions and empirical observations on European Council and European Commission agenda setting holds negative implications for the suitability of existing theoretical frameworks to analyse these agenda setting relations. Indeed, frameworks which presume the relation between the institutions to be either
hierarchical or non-hierarchical cannot account for these relations occurring simultaneously. As a consequence, they risk making inaccurate claims of the agenda setting prowess of the European Council over the European Commission. At the same time, the potential
bidirectional relationship between the two institutions, where either affects the agenda of the other, renders it difficult to establish a causal principal-agent association. Therefore, the challenge is to construct a theoretical framework which assumes both a hierarchical and non-hierarchical agenda setting relation at play, whilst still being able to insulate a hypothesised principal-agent relationship that can be measured empirically. This has not yet been
attempted in the public administration literature.
Disentangling agenda setting relations. In order to further specify such a theoretical
framework, it is necessary to disentangle potentially bidirectional agenda setting relations. This helps clarify how a measurable causal mechanism can be isolated theoretically and empirically. Recall that we are interested in how the European Council affects the European Commission agenda. The first step is to ascertain what a reverse principal-agent causal mechanism would look like. Put more concisely: how could the European Commission affect the agenda of the European Council? Two possible avenues through which the European Commission can exert agenda setting influence need to be scrutinised. First, through the substance of specific policy issues which appear on the European Council agenda. Second, through the prioritisation of some policy issues over others. The first avenue requires us to delve into the details of procedural aspects of European Council summit preparations.
The second avenue necessitates an analysis of the general proceedings and deliberations at European Council summits. These two avenues will be discussed in order.
Though the European Commission is generally perceived to be “fully involved in. . . the preparatory phase” (Höing & Wessels, 2013, p. 139) in the run-up to European Council summits, this involvement is in fact relatively shallow. First and foremost, it is the European Council Presidency which kickstarts summit preparations, supported by the General
Secretariat of the Council of Ministers (Wessels, Valant, & Kunstein, 2015, p. 269).5 Both the draft agenda and draft Conclusions are submitted to the General Affairs Council (GAC) and the Foreign Affairs Council (FAC) for deliberation (Eggermont, 2012, pp. 76–77). The former Council configuration assumes the preparation of Conclusions in all policy domains but Common Foreign and Security Policy (CFSP). In CFSP, this responsibility falls on the High Representative for Foreign Affairs and Security Policy (HR/VP) and the FAC
(Eggermont, 2012, pp. 74–75). Preparatory and exploratory discussions and negotiations by the Permanent Representatives Committee (COREPER) narrow down the range of topics to “only the most contentious issues” upon which the GAC, FAC, and European Council make final decisions (Werts, 2008, p. 64). These preparatory efforts culminate in draft Conclusions which subsequently are thrusted “in circulation” among HSGs shortly before the European Council summit (Werts, 2008, p. 65).
Second, although ‘coordination’ with the Commission President occurs, there is no institutionalised venue for input of the European Commission prior to summits. It is unclear to what extent the European Commission’s ad hoc “suggestions for the composition of the [European Council] agenda” through ‘Communications, White Papers, Green Papers, or Initiatives truly matter (Nugent & Rhinard, 2016, pp. 1201–1203; Werts, 2008, pp. 64–65).
5 Prior to the 2009 Lisbon Treaty, the rotating Presidency of the Council of Ministers also presided over, and
Due to the complexity and disaggregated nature of the preparatory process, the substantive impact of these Communications is not clear-cut and, indeed, disputed. For example, Werts (2008) argues that the European Commission is relatively successful at manoeuvring the exact wording of its policy preferences into the Conclusions. The European Commission would supposedly be able to do so even if the policy issue “did not even get a mention” during the summit (Werts, 2008, p. 53). By contrast, Eggermont (2012) claims that
Commission communications are indeed used, but primarily by COREPER for auxiliary and technical purposes. In this reading, Commission input primarily serves to support politically contentious discussions by the GAC, FAC, and HSGs, while they are supposedly “hardly read” elsewhere and “cannot be overestimated” in their significance (Eggermont, 2012, p. 79). In light of the above it appears that the European Commission influence is relatively shallow when it comes to influencing European Council summit preparations indeed.
It is now necessary to consider the second avenue through which the European Commission could potentially influence European Council summit proceedings. The mere fact that the European Commission President possesses a seat at the European Council table could constitute ‘influence’ in itself. Yet, it has equally been noted that Treaty provisions relegated “secondary status” to the European Commission President in attending summits. Even if the Commission President attends the European Council summits and actively participates in deliberations as full-fledged member (Werts, 2008, p. 52), they remain politically “subordinate” and “absolutely not in a position to block a consensus” among the HSGs (Eggermont, 2012, p. 44). Commissioners’ lack of “legitimacy from popular elections” means they participate as member with “secondary status” (Höing & Wessels, 2013, p. 136). Indeed, Article 235(1) TEU precludes the Commission President from voting (Höing & Wessels, 2013, p. 136).
To delve more specifically into how the European Commission could influence summit proceedings, it is necessary to turn to European Council Conclusions. Whereas the European Council proceeding are confidential, and no minutes of the summits are published, the Conclusions are publicly available. Consequently, the European Council Conclusions are the only available indicator of what the HSGs discussed at a given summit. The European Council Conclusions are not only the sole available proxy, but also the most consistent, reliable, and accurate gauge of the actual agenda of the European Council (Alexandrova, Carammia, Princen, & Timmermans, 2014, pp. 156–157). To reiterate: European Council Conclusions do not reveal the true agenda of the institution word-for-word. However, they do provide an insight in the institution’s allocation of attention to policy areas. In this context, attention is defined as the “occurrence of an issue on the agenda of a political institution” (Elias, 2019, p. 72). According to Alexandrova et al. (2014), the distribution of attention to different policy areas in the Conclusions is a reliable proxy of the policy priorities of the European Council.
The Commission may be able to influence the substance of legislative proposals and their wording. The cases presented by Nugent and Rhinard (2016) exemplify how the European Commission President “[acts] as a purposeful opportunist” during summits by making proposals that serve as “focal points” (Nugent & Rhinard, 2016, pp. 1201–1203; Pollack, 2003, p. 49). By contrast, the European Commission is unable to decide on or to alter the distribution of attention by the European Council to certain policy areas over others. Even if the European Commission influences or decides on the policy, the European Council still allocates priority to some policy over others, by allocating different levels of attention to each issue. As Alexandrova (2017) points out, the “agreement to discuss [an] . . . event in the European Council is separate from the decision on the actual amount of attention the event will receive” (Alexandrova, 2017, p. 520).
