Journal of Contemporary European Research
Volume 12, Issue 4 (2016)
Research Article
In the Shadow of Consensus: Communication within Council Working Groups
Petr Kaniok, Masaryk University
Citation
Kaniok, P. (2016). 'In the Shadow of Consensus: Communication within Council Working Groups', Journal of Contemporary European Research. 12 (4), 881 - 897.
First published at: www.jcer.net
Volume 12, Issue 4 (2016)
Petr Kaniok
Abstract
There are two competing perceptions of the EU Council and its working groups. The first of them argues that the Council works as a battleground for expressing the interests of Member States and other participating actors. A competing view emphasizes the effects of socialization and informal norm shaping behaviour of the actors involved. It thus considers the Council as a forum where consensus prevails. This article analyses how different actors acting in the Council working groups communicate in a formal way. Based upon analysis of non-participatory observation of interventions, it finds that working groups tend to be arenas for real bargaining where the actors enforce their interests. It also finds that even the Council Presidency focuses on interests' promotion and that socialization - which can be found at the COREPER level - does not take place in the working groups.
Keywords
EU Council, Working Group, Presidency, EU Member States
INTRODUCTION
In light of the fact that the Council of the EU is one of the most important EU institutions, it is striking that working groups have been quite neglected as a topic of interest in European Studies. This gap concerns not only the total number of studies and articles primarily devoted to these internal Council bodies, but also the methods and approaches employed in the studies. It is true that the total number of articles, books or chapters has increased substantially in recent years. Existing studies focus primarily on the role played by working groups in the Council political process (Olsen 2011; Hage 2008, 2013), but little is known about the internal life of the Council's working groups. The existing research places particular reliance on data gathered from insiders in the form of interviews (Naurin 2007, 2015) or questionnaires (Naurin 2010). There is, however, no study which attempts to describe working groups using data independent of the actors' own assessments.
This study attempts to fill this gap by analysing interventions within the working groups operating mainly in the area of the internal market. Based upon data gathered during non-participatory observations of more than 20 meetings, the paper aims to uncover whether working groups should be viewed as a battleground for national interests or rather as a forum of consensus where common interests prevail. In doing so, the study focuses on three different factors: the role of key players, the characteristics of these actors, and their affiliations. Moreover, the paper analyses how different players contribute to the overall atmosphere of the working groups.
The main findings of the analysis may be summarized as follows: First, working groups tend to be more competitive than consensus-oriented. Second, actors differ significantly in their behaviour. Member states are the most cooperative actor followed by the European Commission while the Presidency focuses on promoting its own interests. Third, actor affiliation does not play a role, as Brussels-based delegates does not tend to adopt a more cooperative stance than do delegates from the capitals. Also the length of the EU membership does influence actors' behaviour.
The article proceeds as follows: the first section introduces the position of the Council working groups and their role in the Council's decision-making system. The second part provides an overview
882
Volume 12, Issue 4 (2016)
Petr Kaniok
on existing research, followed by theoretical framework and hypotheses. The third section is devoted to a description of the data used, as well as the process of gathering data. It also offers an explanation of the methods used in the analysis. Then the paper focuses on the analysis and results in the context of possible further research.
THE WORLD OF COUNCIL: MINISTERS, COREPER AND WORKING GROUPS
The Council itself consists of three basic levels: working groups, preparatory bodies such as the Committee of Permanent Representatives (COREPER) and the ministerial level. While the ministerial level is ordinarily treated separately from the technical and semi-political dimensions, current research sometimes treats working groups and preparatory bodies as a single entity or similar entities1. In this analysis, however, I take into account only the working groups, leaving out consideration of bodies such as COREPER, the Antici Group or the Mertens Group. Also omitted are ad hoc and consultative committees.
Working groups represent the most basic element of the Council's work. Different authors give different estimates of their numbers - usually, there are between 170 - 200 working groups2. Fouilleux et al. (2007: 98) maintain that working groups should be embedded in the institutional structure of the Council, consist of attaches from the Member States' permanent representation and national experts, deal with several pieces of legislation at a time, exist for a number of years and prepare COREPER and ministerial level meetings3. The function of a working group may be described as that of a body which enables the negotiation of Member States' positions. Nevertheless, the Member States are not the only parties involved. Important roles are assigned to the Presidency, to the Commission, and to the Council Secretariat. Legislative work consists of the deliberating proposals for the EU legislature, non-legislative activities include preparing Council conclusions. Essentially, each working group is assigned with preparing a particular file for the Council decision. This means the working party should reach a consensus on the text which will enable its adoption at the COREPER level and subsequently its formal approval by the ministers at the Council level.
Although working groups vary in many respects (see Fouilleux et al. 2007), their usual activity may be characterized in terms of several shared features. Each group is composed of one or more representatives from each Member State, members of the Council General Secretariat, members of the Commission staff and the chair. The group is tasked with going through the legislative or non-legislative documents in order to find a compromise which maximally suits the parties involved. This is usually done article by article. While Member States and the Commission primarily express their interests, the Presidency is supposed to listen and try to find a compromise solution. The Council Secretariat is present specifically in order to explain legal difficulties and possibilities. However, the Council Secretariat may go beyond its traditional technocratic role and play an important political role (Beach 2007). Different types of delegates attend. Member States are represented by the staff of their permanent representations in Brussels. These attaches cover one or more working groups simultaneously. They may be accompanied by national experts from the capitals. Sometimes a meeting may be attended only by the expert or only by the attache. The Commission is represented by the head of the unit responsible for particular legislation, along with one or two other officials. The team from the Presidency consists of the chair and one or two assisting persons.
This article examines formal oral communication within the working groups. As communication are understood oral formal expressions that are presented during meetings by those who attend them -so called interventions. Interventions, generally speaking, represent the most direct route by which a working group's actor can influence its business. In intervening, Member States are theoretically restricted by the Council's Rules of Procedure, which say that Member States should not intervene unless they are proposing a change to the item under discussion (Council Decision 2009/937/EU,
883
Volume 12, Issue 4 (2016)
Petr Kaniok
annex 5). In practice, however, the content of interventions does not always follow this rule. Participants are allowed to speak about whatever they wish. Interventions are not the only manner by which a particular issue can be influenced or communicated. Actors may, for example, also send written comments and may negotiate bilaterally or multilaterally on a purely informal basis4. Such forms of communication are however omitted as data for their research can be hardly approached.
LITERATURE REVIEW
Little previous research has been directed at Council working groups. This may be seen as surprising, since research into the overall role of committees in the EL) decision-making process is quite broad and well-developed (see Pedler and Schafer 1996; Christiansen and Larsson 2007; Blom-Hansen 2011; Heritier et al. 2013). All these committees exercise varied roles within the EU, since they are part of institutions emphasizing different interests. Thus it is very difficult to treat committees as a compact entity, even though such approaches exist (Quaglia et al. 2008).
Council working groups are seen from two broad perspectives. The rationalist perspective considers them as formally important, because they serve as communication channels for expressing national interests. Members of working groups are bound by national instructions based upon preferences formulated in their home capitals. These preferences reflect the interests of economic, social and political actors from each member state and the outcome of bargaining that may need to take place at this level in order to establish a single national stance to be maintained in European-level negotiations (Beyers and Diericks 1998; Moravcsik 1998). By contrast, for the neo-institutionalists, working groups play a more active role. They are seen as arenas within which preferences are bargained for and where the very rules governing such negotiations are defined. In short, the members of Council working groups go beyond the function of merely negotiating among preexisting interests. Instead they contribute to redefining European public issues, the rules and norms that structure negotiation and sometimes even the very identities of the actors involved (Lewis 1998; Lewis 2005; Aus 2008).
Existing research may be divided into three basic groups. The first consists of work focusing on the role of working groups in the decision-making process, as well as their influence and capacity. Common wisdom indicates that working groups prepare and decide the majority of Council outcomes (Hayes-Renshaw and Wallace 1997; van Schendelen 1996). Most of these conclusions have been based either upon pure estimation or upon information which comes from insiders. Current research repeatedly challenges such figures. Hage (2008) found that working groups were responsible for less than 40 per cent of decisions, Olsen (2011: 159) notes that an even smaller amount of decisions, only 33 per cent, are made by working groups.
The second line of research targets the issue of communication. Two works of Beyers and Diericks (1997, 1998) may be seen as pioneering in creating systematic in-depth analyses. The first piece aimed at exploring links between communication involving national delegates and several discretionary factors, revealing that discretion matters (Beyers and Dierickx 1997). The second study analysed the form of communication which takes place within working groups. It showed that informal communication is intense in working groups populated by Brussels-based attendants. Surprisingly, this communication is led by non-state actors. The more influential actors were revealed to be those coming from large member states, and communication patterns followed a North-South division (Beyers and Dierickx 1998). The presence of this conflict line was later confirmed by Naurin (2007).
The third branch of research consists of studies looking into the loyalty of delegates. Especially for research using data gathered at the COREPER level, the findings suggest that delegates acting in
884
Volume 12, Issue 4 (2016)
Petr Kaniok
these groups have shared loyalties, both to the group and to their respective states (see Egeberg 1999; Beyers, 2005). This is supported by several studies conducted by Lewis (1998, 2003, 2005) which state that members of COREPER develop process and relationship interests, as well as a sense of collective responsibility. Lewis claims that COREPER is driven not only by the logic of consequences, but also by the logic of appropriateness (Lewis 2005: 942). Trondal and Veggeland (2003) confirm the shared loyalty thesis even with the delegates of Member States' in European Commission committees. Moving to the level of working groups, Naurin (2010) discovered that there are prevailing patterns of discussion within working groups. Naurin (2010: 50) claims that the main intention behind giving explanations is to try to convince others more often than it is to clarify one's position in order to facilitate a compromise in working groups. In his most recent study, Naurin (2015) challenges the prevailing conceptualization of the Council as an arena where intergovernmental negotiations can be characterized as consensual, claiming that particularly the 'Big 3' (France, Germany and the United Kingdom) are unwilling to make generous concessions.
The prevailing message from existing research seems to be that socialization takes place in lower levels of the Council (particularly at the COREPER level) and actors behave in a manner which is far from being driven only by their self-interests. However, all research relies on data gathered from insiders5 or on the detailed study of the negotiation of one or more pieces of legislation. This approach is quite understandable, since the Council is not one of the most transparent institutions in the EL). By relying on information from insiders, however, there is a risk that the data may be biased. Firstly, insiders may overestimate their own roles, mix the formal versus informal levels of negotiation, or may simply provide only that information which they find comfortable to discuss. Also problematic is the fact that insiders are often asked to evaluate not just themselves, but other delegates and their positions as well, or to adopt a general stance resulting in a 'mean stance' for the particular Member State. Moreover, the majority of research dealing with the internal behaviour in preparatory bodies relies upon data from COREPER. However, COREPER is a very specific entity which differs in terms of attendance, substantial knowledge of files received from working groups. It is therefore problematic to merge these two discrepant levels and automatically assume that working groups share the same features as does COREPER. This paper thus contributes to our knowledge of working groups by exploring relevant data acquired directly at the working group level in a manner which is independent of the actors involved.
This study seeks to make a contribution to the research on the communication within working groups. It focuses on two key issues - firstly, it studies what kind of communication - in form of interventions - takes place in working group meetings. Is this communication more cooperative in nature? Or does it rather incline to a pattern in which the fulfilment of one's own interests is paramount? Following existing research on COREPER and other Council preparatory bodies, I expect that communication within working groups tends to be cooperative rather than uncooperative. I am also interested in whether there are differences in building this consensual communication according to the type of actor involved. Secondly, how do various actors influence the internal communication of working group meetings? Do the interventions which construct this communication show significant differences depending upon the types of actors involved?
Motivated by these two main questions, the following four hypotheses are tested in this paper:
Hf. The general pattern of communication within the working groups will be cooperative rather than competitive.
H2: At the individual level, Brussels-based delegates will be significantly more cooperative in communication than delegates coming in from the capitals or than those in mixed delegations.
885
Volume 12, Issue 4 (2016)
Petr Kaniok
H3: Of the collective actors involved, the Presidency and the Council Secretariat will be significantly more cooperative in communication than delegates representing the Member States and the European Commission.
H4: The longer a particular collective actor is part of a working group's structure, the more cooperative in communication it tends to be.
Generally speaking, all hypotheses are rooted in the socialization argument. This concept is very broad as it may be applied both to social constructivism as well as to rational choice theory (Quaglia et al. 2008: 157). While for the former it deals particularly with the internationalization of norms, for the later it means especially strategic role play. Such conceptualized socialization means that actors adjust their strategies to the legal, informational, and organizational opportunities and constraints provided by committees and multiple principals, and their behaviour varies accordingly. In both cases - for the social constructivist as well as the rationalist - socialization has an effect on behaviour although the mechanisms differ (Checkel 2005; Trondal 2007).
When developing these expectations into more specific assumptions, it can be argued that the Council is far from being a purely intergovernmental arena which serves for the expression and defence of national interests (Aus 2008) and as such there is a strong preference for mutual cooperation among parties rather than for contestation (Hi). Proceeding to the individual level, the affiliations of delegates make a difference in their behaviour (Fouilleux et al. 2007). Delegates who are permanently deployed in Brussels share a sense of responsibility and thus act in more a cooperative way than their fellows from the capitals. This differentiation is important because Brussels-based diplomats tend to behave and negotiate in different ways than do national experts6. Succinctly put, the former are set to adopt a more cooperative style in negotiating than the latter. (H2).
The socialization argument is also valid from the collective actors' perspective. There is a broadly accepted assumption that the Council Presidency (H3) acts as an impartial, neutral actor which gives up the pursuit of its own interests (Tallberg 2006; Tallberg, 2008: 187; Bunse 2008: 39). Such claims are also connected with social constructivism or sociological institutionalism as they deal with the logic of appropriateness. Following this concept, the Presidency is constrained by expectations from other Member States or by shared norms of impartiality. Last but not least, socialization takes into account time as a factor which enables various actors to accept internal norms and rules. One could thus expect that the longer a collective actor takes part in working groups, the more it accepts and follows their internal norms of consensus and cooperation (H4).
When evaluating the hypotheses, I control for three factors which may also influence actors' communication behaviour. Firstly, the size of the actor matters. Possession of more resources can affect willingness of such states and institutions to cooperate or act independently (Naurin 2015). Secondly, salience plays a role in actors' behaviour and their willingness to adopt a compromise on a particular issue. There is evidence that legislative bodies in the EL) use their procedural powers more forcefully when facing important issues (Selck 2003). For example, politically salient proposals are more likely to be decided in the first reading stage (Rasmussen 2007). Whether a decision is made at the administrative or the ministerial level in the Council also depends on the political salience of an issue (Hage 2007). Schneider et al. (2010: 92) claims that higher salience leads to a higher willingness to make concessions to reach an agreement at all. Thus one may expect that cooperation in the working groups will be higher when dealing with legislative proposals than when preparing non-legislative documents. Thirdly, the language used when intervening can also importantly influence the degree of cooperation. English is the modern lingua franca in the Council, with a substantial majority of both formal negotiations and informal communications among delegates carried out using it (Egeberg et al. 2003: 27-30; van Els 2005). Also, in formal negotiations delegates rarely use either French or German. If they do not use their mother tongue, they are using English. As Egeberg
886
Volume 12, Issue 4 (2016)
Petr Kaniok
et al. claim (2003: 28), in the late 1990s, 90 per cent of non-native English speakers representing their countries in the Council were able to communicate to some extent in English, and more than 80 per cent spoke English well or very well. Therefore, using English may be seen as a factor which supports cooperation in the working group as it saves time and gives a substantial majority of delegates' equal conditions in the negotiation process.
DATA AND METHOD
The data employed in this study comes from the non-participatory observation of more than 20 meetings of various Council working groups dealing with policies related particularly to the Single Market. An overview of meetings is presented in Table 1. Working groups were selected for observation based upon the willingness of relevant attaches to enable non-participatory attendance. The working groups presented in this paper thus do not comprise a comprehensive picture of all working groups across all policy sectors. Nevertheless, this sample offers a unique perspective on the internal life of Council working groups. The observation took place from the beginning of October 2013 to the end of November 2013. Council working groups usually work all-day, with a 90-minute lunch break. Sometimes groups may meet for a half-day only, either in the morning or in the afternoon.
Table 1. Overview of working groups
Day	Working group name	Character	Agenda
9.10.	Gl - Competitiveness and Growth	Half day	Legislative
10.10.	101 - Social Questions	Half day	Legislative
15. 10.	Gl - Competitiveness and Growth	Full day	Non-legislative
17.10.	H5 -Telecommunications and Information Society	Full day	Legislative
18.10.	101 - Social Questions	Full day	Legislative
22.10.	Gl - Competitiveness and Growth	Half day	Non-legislative
23.10.	G7-Technical Harmonisation	Full day	Legislative
24.10.	G7-Technical Harmonisation	Full day	Non-legislative
28.10.	G23 - Consumer Protection and Information	Full day	Legislative
29.10.	Gl - Competitiveness and Growth	Full day	Legislative
31.10.	G12 - Competition	Full day	Legislative
5. 11.	H5 -Telecommunications and Information Society	Full day	Legislative
6. 11.	Gl - Competitiveness and Growth (High level group)	Full day	Non-legislative
7. 11.	Gl - Competitiveness and Growth	Full day	Non-legislative
11. 11.	Gl - Competitiveness and Growth	Full day	Legislative
12.11	H5 -Telecommunications and Information Society	Full day	Legislative
18.11.	Gl - Competitiveness and Growth	Full day	Legislative
19.11.	H5 -Telecommunications and Information Society	Full day	Legislative
21.11.	H5 -Telecommunications and Information Society	Half day	Non-legislative Legislative
25.11.	A16 - Friends of the Presidency Group (Integrated Maritime Policy)	Full day	Legislative
26.11.	H03 - Aviation Working Party	Full day	Legislative
887
Volume 12, Issue 4 (2016)
Petr Kaniok
Data gathering was carried out in two phases. The first consisted of two weeks of observation of the working groups7 to identify which kinds of interventions are present and how they might be defined. Then a coding scheme describing all variables was constructed. There were five basic variables that were followed during the subsequent observation process - the actor, intervention, language used, type of delegate and the character of the agenda.
The first variable, 'actor', consists of four values which are used to identify each type of actor who attended in order to express their views during the working group meetings. The first of these is the Presidency; the second actor type is the Member State; the third is the European Commission; and the fourth is the Council Secretariat. The second variable captures the content of interventions expressed by the various actors during the meetings. Each type of rhetorical act was defined in terms of its content and assigned coding values. This variable contains six values; these are described in Table 2.
Table 2. Coding of interventions
Label	Description
Position	The actor explicitly communicates only her/his own substantial position/opinion/request without referring to the other parties.
Procedure	The speaker's intervention concerns a procedural, insubstantial matter.
Support	The intervention communicates support for another party's position without explicitly expressing the actor's own position.
Position and procedure	The actor's own substantial position is mixed with procedural requests/remarks or issues. The intervention clearly contains both these parts.
Position and support	The actor explicitly communicates its own position but frames it in the context of other actors, expressing its support for their position and opinion. The intervention clearly contains both these parts.
Support and procedure	The speaker's intervention concerns a procedural, insubstantial matter but at the same time, the actor also explicitly praises another party's position or approach. The intervention clearly contains both these parts.
This division of interventions into six categories is based upon the three basic messages that delegates communicate when taking the floor. The first of these is a clear expression of their own position or interest ('Position'). In doing so, the delegate simply states what he or she wants - e.g., how the particular Member State wishes to rewrite a specific sentence or document, or which changes it finds acceptable. Such intervention contains only a demand and is not accompanied by any complimentary remark or statement supporting another actor. In the context of cooperation within the working party, a simple expression of a state's position is thus considered to be a factor which decreases the level of cooperation. It neither explicitly contributes to the existing coalition, nor does it show appreciation for the activity of any other actor. As it usually expresses new demands, it rather complicates the process of negotiation.
The second common type of message is a procedural intervention ('Procedure'). In making a procedural intervention, a delegate may wish to clarify further proceedings, for example. Interventions made by the Presidency in yielding the floor to other delegates are also classified as procedural interventions. As such, procedural interventions are treated as neutral in their contribution towards the atmosphere in the working groups. Procedural interventions were included into the dataset even though they do not necessarily bear any message directly related to the content of negotiation. On the other hand, they shape the internal communication atmosphere within a working group. For example, referring to a particular procedural rule has an effect on how smoothly a working party goes through its agenda.
888
Volume 12, Issue 4 (2016)
Petr Kaniok
The third basic message is an expression of support for another party's opinion or stance, or praise for the work of another actor ('Support'). The former does not exclude an actor's own preference from the statement but it always indicates his/her willingness to cooperate or his/her awareness of existing positions. An actor's position is, in such a case, present only implicitly. Support for another actor's position is thus treated as factor which increases cooperation within the group. The remaining three categories are based upon combining the abovementioned three kinds of simple interventions. First, an actor may combine its own preference with an emphasis on a procedural issue ('Position and Procedure') - the latter part of such an intervention may serve as an argument supporting the actor's demand or it may merely explain the procedural motivation underlying it. Secondly, an explicit expression of one's own position may go hand in hand with an expression of support for another Member state ('Position and Support'). Finally, support for another's point can be combined with a procedural remark ('Support and Procedure').
As a result, binary dependent variable 'Communication' was created. Value 1 ('Cooperative') merges all interventions which contain support for another actor - either being the only message of intervention or being accompanied by procedural remark or by speaker's own position. Value 0 ('Uncooperative') on the contrary unites interventions bearing speaker's own position, either as the only content of the intervention or being accompanied by commenting on procedural issues. Purely procedural interventions were not included into the exploratory analysis as the can be classified as neutral. However, they are presented and commented in the descriptive part of analysis in order to illustrate which kind of actors express them and how important they are in the overall communication within the working groups.
The logic behind the dependent variable ('Communication) is based upon experience expressed by practitioners8 and the Council's internal norms. For the first, practitioners say that not only what is said during the meetings but how it is expressed is highly significant. Disagreement or dissatisfaction with, for example, changes made by the Presidency may be expressed in various ways which substantially affect both the overall atmosphere of the meeting and the perception of the speaker. Demands which are articulated in the context of other actors' positions are considered as more acceptable and more 'user friendly' than the mere expression of the speaker's interests. Even the Council's internal norms prefer certain values such as efficiency, consensus or cooperation among Member States. For example, the Council's Rules of Procedure considers a full round table as proscribed in principle and encourages delegations to express their demands collectively. This concerns particularly like-minded delegations which should hold consultations prior the meetings and then present their common positions. The Council's Rules of Procedure also expect that concrete proposals for amendments should be sent in written form9.
Regarding the independent variables, the first of them captures actor intervening ('Actor'). There are two distinctions. The first of them differentiates between collective actors as a whole, dividing them into Presidency, Commission, Member State, and Council Secretariat. If a Member State intervened, the particular Member State was coded. Lists of participants were used to construct a variable labelled 'Representative'. This captures whether a Member State is represented strictly by a Brussels-based diplomat, or by a national expert coming in from the capital, or a combination of the two.
Variable 'Length of EU membership' is expressed as the number of years an actor had been a member of the European Union. When it comes to non-state actors, they are treated as having existed since the beginning of the European integration process. As the Presidency combines an institution and a member state holding the office, the value for the Presidency is computed as a sum of the length of member state membership and the length of the Presidency divided by two10.
Concerning the control variables, the first of them (variable 'Size') is based upon the Panke's catalogue of small EU Member States, which is in turn based upon voting power in the Council (Panke 2010: 15-18). In this variable, the Commission is treated as a large actor. Even though it does
889
Volume 12, Issue 4 (2016)
Petr Kaniok
not vote in the Council, its overall power in negotiations and communication is enormous. The Commission sets the agenda, has its own interests in negotiations and thus has the power to influence the negotiation process even at the working group level. The Council Secretariat is treated as a small actor because it assists the working groups' work and can hardly enforce its own interest. When it comes to the Presidency, it is in line with literature classified as a small actor as well as it should follow the norms of neutrality and impartiality. The second control variable 'Language used' captures the language used during the meetings. The basic distinction is between English and other languages. Thirdly, meeting agendas accessible in the room or on the Council website prior to the meeting were used to construct the variable 'Item', which divides the agenda between legislative and non-legislative issues.
The second phase of data gathering consisted of collecting interventions during the meetings. The predefined intervention categories given in Table 2 were used to note each intervention by an actor in terms of its content, language and - if applicable - the Member State speaking. This was possible due to the fact that interventions articulated during meetings are usually quite brief in terms of time, visibility and audibility. The researcher was present throughout 21 meetings from beginning to end, noting the interventions with the help of the prepared table. All in all, during the meetings, the research gathered 5021 interventions (including procedural ones). In the analysis that follows, a binary logistic regression is used in order to evaluate how independent variables contribute to the communication patterns within the working groups. Prior this explanatory part, detailed descriptive insight into dataset is provided in order to demonstrate differences in formal communication among various actors.
ANALYSIS
The first step of the analysis introduces a descriptive overview of the data, as summarized in Figure 1, Figure 2 and Table 3. While Figure 1 presents all interventions including the procedural ones, Figure 2 excludes them and offers an overview of substantive communication within the groups. Then, Table 3 summarizes the data used for the exploratory analysis in terms of distribution of the dependent variable ('Communication').
As Figure 1 indicates, if all interventions are taken into account, the dominant speaker taking the floor during the working groups' meetings is the Presidency. However, its role is mainly procedural as the obvious majority of its interventions are of a purely organizational nature. Thus, the Presidency can be described as a 'dealer' distributing the floor among other actors and ensuring that the process goes smoothly. This finding perfectly fits with the role of the Presidency as described in the literature. According to it, the Presidency should focus on the role of the chair who wants to find a common interest, leaving their own goals behind. On the contrary, the Member States and the European Commission intervene in rather substantive terms. Both actors focus on expressing their positions, but at the same time they also express support for the other parties' interests. Even this picture corresponds with what we know about the Council in general. The Council's General Secretariat presence is, in terms of interventions, almost invisible which is not surprising as well.
However, if purely procedural interventions are removed from the sample, a completely different picture emerges. As Figures 2 shows, if procedural comments are deleted from the dataset, Member States become dominant actors. Their focus on interest promotion and coalition building is not changed, which is the case for the Commission as well. This is not surprising - the Council and its components are designed exactly for interest articulation and aggregation. However, what changes compared to the complete dataset is the role of the Presidency. As Figure 2 illustrates, even the Presidency has its own interests which it tries to promote. Without being hidden in the 'procedural fog', this dimension of the Presidency becomes quite clear. Figure 2 also reveals that the Presidency
890
Volume 12, Issue 4 (2016)
Petr Kaniok
prefers combining its interest promotion with procedural interventions. It seems that the Presidency uses this approach in order to soften its demands, make them more acceptable and to be in line with its expected role of neutral chair and honest broker.
Figure 1. Interventions in the working groups
2500 n
Presidency Commission Member States General Secretariat
■ Position ■ Position and Procedure ■ Procedure
■ Support and Position    ■ Support and Procedure ■ Support
Figure 2. Interventions in the working groups without procedural items 1200 n
Presidency Commission Member States General Secretariat
■Position ■ Position and Procedure ■ Support and Position ■ Support and Procedure "Support
Finally, Table 3 reports the distribution of the dependent variable according to the actors. The very simple analysis summarized in Table 4 does not support Hi, which anticipates cooperative, rather than competitive, communication in the working groups. This means that participants tend to push their own interests by intervening, rather than taking into account the positions of other actors. The
891
Volume 12, Issue 4 (2016)
Petr Kaniok
working groups are thus slightly closer to functioning as arenas of intergovernmental negotiation than to socialized forums where common interests prevail. This means that real negotiations take place there and this formal level bargaining is important. It can hardly be claimed that negotiations can be found only behind the scenes and that meetings of working groups merely present the outcomes of these unseen processes.
Table 3 .An overview of communication within working groups according actors
	Cooperative	Uncooperative	Total
Actor			
Presidency	54	311	365
Member States	963	1086	2049
Commission	58	183	241
General Secretariat	0	20	20
Total	1075	1600	2675
In the second part of the analysis, a binary logistic regression was used to investigate what types of independent variables influence the dependent variable and to what extent. In this analysis, all interventions expressed by the Council Secretariat were removed from the dataset. As Table 3 suggests, there is no variation in the Council Secretariat's interventions which makes it problematic for statistical analysis. The reduced dataset for exploratory analysis thus contains 2655 interventions. Its results are summarized in Table 5 reporting B (and its SE) and Exp(B) coefficients as well as Wald coefficients.
Table 5. Results of binary regression on communication
	B(SE)	Wald	Exp(B)
Independent variables			
Actor: Commission	.75 (.24)***	10.10	2.12
Actor: Member State	1.78 (.16)***	117.95	5.95
Representative: Brussels based	.11 (.01)	1.46	1.12
Length of EU Membership	.00 (.00)	.45	1.00
Control variable			
Item: Non legislative	.58 (.13)***	20.22	1.79
Size: Big actor	-.20 (.11)*	3.33	.82
Language: Non-English	.02 (.01)	.06	1.02
			
Constant	-1.97		
Nagelkerke R .10 . *** p < 0.01, ** p < 0.05, * p < 0.01. Standard errors in parentheses.
Table 5 shows fairly well that in general terms the model does not explain many of the communication patterns within the working groups. The value of Nagelkerke R2 coefficient is quite low, even the difference between values of -LL for initial model (3583.98) and model for the regression (3375.22) is quite small. However, the goal of the analysis was not to explain amount of variation in the formal communication, but to test theoretically developed hypotheses.
892
Volume 12, Issue 4 (2016)
Petr Kaniok
When it comes to their evaluation, the analysis addresses a few interesting findings - in terms of existing research. Regarding the actors involved, the Presidency - if 'stripped from procedural clothes' - is not as cooperative as might be expected. Quite on the contrary. Both the Commission and the Member States tend, in their substantive interventions, to be more cooperative than the country in the helm. These findings are statistically highly significant and contribute the most to the model's explanatory power. This can be interpreted as that the actor characteristic is the most decisive regarding the communication atmosphere within the working groups. Even though the General Secretariat's interventions had to be omitted from the analysis - which weakens the conclusion - H2 cannot be supported.
Also the third hypothesis, expecting that the Brussels-based delegates would be more cooperative than their capital-based fellows, cannot be confirmed. Although the analysis shows that the Brussels-based delegates increase the degree of cooperation in working group formal communication, this contribution to the cooperative atmosphere within the groups is not statistically significant.
Last but not least, the data does not support hypothesis H4 either. That means that regarding the degree of cooperation in the formal communication, there is no difference in the length of EL) membership. Socialization at the collective level thus does not play that much of a role. One cannot thus expect that the longer a particular actor is involved in the process of the European integration, the more cooperative it is.
When it comes to the control variables, type of item and size of actor play a statistically significant role. Legislative items substantially decrease the degree of cooperation in the formal communication. The same can be said regarding the size. Big actors in the working groups such as influential Member States or the Commission tend to be less cooperative than the smaller ones. Both these findings are not surprising - the legislation is generally seen as more important than non-legislative points and also big actors tend to pursue their interests more actively than their smaller counterparts. On the other hand, language does not play a role. There is no significant difference between English and non-English speakers. When English is used, the level of cooperation in communication increases. However, this contribution is not statistically significant.
CONCLUSION
The working groups of the EU Council are not among the most-described players in the EU decisionmaking system. Due to their role and position within the Council, access to data which describes their functioning is limited. The existing research suffers from two major shortcomings. First, it places particular reliance on the information provided by insiders and the ex-post evaluation of their activity. Second, the majority of studies use COREPER data. There is, therefore, a remarkable deficit in our understanding of how working groups fulfil their roles and how the parties involved behave. This study fills this gap by analysing formal oral communication within working groups using the non-participatory observation of interventions. Based upon existing research and particularly on the socialization argument, the study expected that the communication pattern in the groups would be cooperative rather than competitive. With regards to particular actors and their contribution to the degree of cooperation, the study anticipated that the role of the Presidency would be important, as well as the affiliations of delegates, and that length of EU membership would play a role.
The findings of this analysis should be - in general - seen as challenging existing research. There is no shared consensus among scholars as to which pattern of behaviour prevails in the Council. The analysis of interventions supports those who depict the Council and its components as an intergovernmental arena. In this respect, for example, the study supports the findings of Naurin (2010), who sees the working groups involved more in argument rather than deliberation. This
893
Volume 12, Issue 4 (2016)
Petr Kaniok
finding is hardly surprising. Working groups do form the basic level at which Member States may express their interests and enforce them. The analysis however shows that even the Presidency is quite active in this regard. It is also shown that as its comparative advantage, huge amount of procedural interventions can be seen, which enable the Presidency to hide its demands and goals.
Additionally, the variety of working groups seems to limit the possibility of creating any 'spirit of common interest'. In this sense, working groups differ substantially from COREPER where a limited group of people meets twice a week.
To sum up, this study suggests that working groups seem to form a quite unique level of the Council structure where actors behave in a substantially different way than in structures which aggregate interests. As they construct the first possibility for expressing substantive demands, actors involved in them use working groups particularly for interest articulation and for coalition building. In terms of substantive interventions, actors differ in their behaviour substantially, particularly if the Presidency is compared with Member states or the Commission.
Why does the behaviour of the Member states and the Presidency differ? There may be several explanations for this. For the first, Member states know that if they want to succeed in interest promotion, they have to find partners and form suitable coalitions. As a Member state is always a Member state, it has to do so constantly. The same applies to the Commission. On the other hand, the Presidency is in a different role. For the first, it is a unique opportunity to promote something that is, for a country in the helm, important. As the Presidency has substantive procedural power, it may hide such promotion in the 'procedural fog' - who would notice that the Presidency wants something if such a demand is wrapped in the typhoon of procedural interventions? Additionally, each Presidency has some substantive agenda. Even though it should be neutral and impartial, it has to promote its priorities unless it should be considered as a failure. One can hardly imagine a Presidency which would totally resign on the promotion of policy goals and focus its power only on consensus building and Council administrative management. And if any Presidency wants to promote anything, it has to start to do so already at the working groups' level.
Regarding the effect of socialization where existing research suggests that Brussels-based participants should be more cooperative than their fellows coming from the capitals; the study does not confirm this expectation. It seems that working groups attendants differ from those attending the COREPER - and here it should be noted that the "socialization hypothesis" is based upon research analysing COREPER activities. In comparison to the COREPER meetings which take place regularly twice a week and usually last for almost half of a day, working groups meet less frequently. As they are the first opportunity for expressing what Member states want, their participants use them in this way. The same perhaps applies for the importance of the length of EL) membership where one could expect that socialization takes place as well.
There are of course limitations of this analysis. First, this study took into account only a limited number of working groups, particularly those related to the Single Market. It would therefore be valuable for future research to include working groups acting in areas that are more intergovernmental. However, with the increased level of intergovernmentalism, hardly any different results could be obtained. Second, the study focused only on formal oral communication, leaving aside for example written inputs or informal processes. Additionally, the study builds upon research that dealt particularly with the COREPER level of the Council. Such a point of departure obviously biases initial expectations as COREPER is in many aspects very different from the working groups. This study highlighted the need for such differentiation and can therefore be seen as an important contribution to our understanding of how the EU Council and its components work.
***
894
Volume 12, Issue 4 (2016)
Petr Kaniok
CORRESPONDENCE ADDRESS
Petr Kaniok, Department of International Relations and European Studies, Masaryk University, Jostova 10, 602 00 Brno, Czech Republic [kaniok@fss.muni.cz]
ENDNOTES
1 For instance, Fouilleux et al. (2005) begin their paper on the role of working groups by referring to Lewis (1998) focusing on research on COREPER.
2 The Council Secretariat regularly publishes a list of working groups. In the last such overview from July 2013, there were 158 'preparatory bodies' altogether, 125 of which were chaired by the Presidency and 33 of which were chaired by a permanent chairman (Council Secretariat 2013). In the period between July 2000 and December 2005, the number of groups varied from 254 to 289 (Hage 2013: 22).
3 Fouilleux et al. additionally say a working group should 'have a change of presidency every 6 months' (2007: 98). However, especially after the Lisbon Treaty created a permanent chair of the Council of External Relations, a substantial number of groups have a fixed chair. Moreover, some working groups are chaired by an elected chair and some are chaired by the Council Secretariat (Council Secretariat 2013:16).
4 A typical example of this kind of negotiation is a 'like-minded group' (Elgstrom 2000: 465).
5 There are some exceptions, one being a study by Cross (2011), which analyzed the conditions and circumstances under which Member States in the Council are ready to intervene. Relying on footnotes noted in official records kept by the Council Secretariat, Cross identified significant differences among Member States in the number of interventions.
6 This difference is precisely captured by quote "When national experts are present, I never let them have the microphone. If I let the experts take the microphone, they would just say what we want from the negotiation and the meeting would be over. Instead our job is to persuade (Fouilleux et al. 2007:104).
7 These meetings are not counted into the number of attended working groups.
8 Interview with attache 9. 10. 2013, Interview with attache 17. 10. 2013, Interview with attache 18. 10. 2013, Interview with attache 23.10. 2013.
9 See, in particular, Annex V of Council Rules of Procedure.
10 In this case it means a value of 35=61 (the Presidency)+9(Lithuania)/2.
REFERENCES
Aus, J. P. (2008). The Mechanism of Consensus: Coming to Agreement on Community Asylum Policy. In: D. Naurin and H. Wallace (Ed.) Unveiling the Council of the European Union. Games Government Play in Brussels. Houndmills/New York: Palgrave Macmillan, pp. 99-118.
Beach, D. (2007). Oiling the Wheels of Compromise: The Council Secretariat in the 1996-97 and 2003-4 IGCs. In: D. Beach and C. Mazzucelli (Ed.) Leadership in the Big Bangs of European Integration. Houndmills/New York: Palgrave Macmillan, pp. 76-93.
Beyers, J. and Dierickx, G. (1997). 'Nationality and European Negotiations: The Working Group of the Council of Ministers'. European Journal of International Relations, 3(4), pp. 435-471.
Beyers, J. and Dierickx, G. (1999). 'The Working Groups of the Council of the European Union: Supranational or Intergovernmental Negotiations? ' Journal of Common Market Studies, 36(3), pp. 289-317.
Beyers, J. (2005). 'Multiple Embeddedness and Socialization in Europe: The Case of Council Officials'. International Organization, 59(4), pp. 899-936.
Blom-Hansen, J. (2011). The EU Comitology System in Theory and Practise. Keeping an Eye on the Commission? Basingstoke: Palgrave Macmillan.
Bunse, S. (2009). Small States and EU Governance. Leadership through the Council Presidency. Houndmills/New York: Pallgrave Macmillan.
895
Volume 12, Issue 4 (2016)
Petr Kaniok
Council Secretariat (2013). List of Council preparatory bodies.
Cross, J. P. (2012). 'Interventions and negotiation in the Council of Ministers of the European Union'. European Union Politics, 13(1), pp. 47-69.
Egeberg, M. (1999). 'Transcending intergovernmentalism? Identity and role perceptions of national officials in EU decisionmaking'. Journal of European Public Policy, 6(3), pp. 456-474.
Egeberg, M., Schaefer, G. and Trondal, J. (2003). 'The many faces of EU committee governance'. West European Politics, 26(3), pp. 19-40.
Elgström, O. (2000). 'Norm negotiations. The construction of new norms regarding gender and development in EU foreign aid policy'. Journal of European Public Policy, 7(3), pp. 457-476.
Foilleux, E., de Maillard, J. and Smith, A. (2005). 'Technical or political? The working groups of the EU Council of Ministers'. Journal of European Public Policy, 12(4), pp. 609-623.
Foilleux, E., de Maillard, J. and Smith, A. (2007). Council working groups: space for sectorized European policy deliberation. In: Christiansen, T. and T. Larsson (Ed.) The Role of Committees in the Policy Process of the European Union. Legislation, Implementation and Deliberation. Cheltenham/Norhampton: Edwar Elgar, pp. 96-120.
Häge, F. M. (2007). 'Committee decision-making in the Council of the European Union'. European Union Politics, 8(3), pp. 299-328.
Häge, F. M. (2008). Decision-Making in the Council of the European Union. The Role of Committees. Rotterdam: Optima Grafische Communicatie.
Häge, F. M. (2013). Bureaucrats as Law-makers. Committee decision-making in the EU Council of Ministers. New York/Oxon: Routledge.
Heritier, A., Moury, C, Bischoff, C. and Bergström, F. (2013). Changing Rule of Delegation. A Contest for Power in Comitology. Oxford: Oxford University Press.
Checkel, J. (2005). 'International institutions and socialization in Europe: introduction and framework'. International Organization, 59(4), pp. 801-26.
Christiansen, T., Larsson, T. (Ed.) (2007). The Role of Committees in the Policy Process of the European Union. Legislation, Implementation and Deliberation. Cheltenham/Norhampton: Edward Elgar.
Lewis, J. (1998). 'Is the "hard-bargaining" image of the Council misleading? The COREPER and the local elections directive'. Journal of Common Market Studies, 36(4), pp. 479-504.
Lewis, J. (2003). 'Institutional environment and everyday EU decision making. Rationalist or constructivist? ' Comparative Political Studies, 36(1-2), pp. 97-124.
Lewis, J. (2005). 'The Janus Face of Brussels: Socialization and Everyday Decision Making in the European Union'. International Organization, 59(4), pp. 937-971.
Moravcsik, A. (1998). The Choice for Europe: Social Purpose and State Power from Messina to Maastricht. Ithaca: Cornell University Press.
Naurin, D. (2007). 'Network Capital and Cooperation Patterns in the Working Groups of the Council of the EU. EUI Working Papers RSCAS 2007/14.
Naurin, D. (2010). 'Most Common When Least Important: Deliberation in the European Union Council of Ministers'. British Journal of Political Science, 40(1), pp. 31-50.
Naurin, D. (2015). 'Generosity in intergovernmental negotiations: The impact of state power, pooling and socialization in the Council of the European Union'. European Journal of Political Research, 54(4), pp. 726-744.
Olsen, I. (2011). Council working groups. Advisors or de facto decision makers? Aarhus: Politica.
896
Volume 12, Issue 4 (2016)
Petr Kaniok
Panke, D. (2010). Small States in the European Union: Coping with Structural Disadvantages. Farnham/Burlington: Ashgate.
Pedler R. H. and Schaefer, G. F. (1996). Shaping European Law and Policy. The Role of Committees in the Political Process. Maastricht: European Institute of Public Administration.
Quaglia, L, Fabrizio De Francesco, F. and Radaelli, C. M. (2008). 'Committee Governance and Socialization in the European Union'. Journal of European Public Policy, 15(1), pp. 155-166.
Rasmussen A. (2007). 'Early conclusion in the co-decision legislative procedure. EUI Working Papers 2007/31.
Schneider, G., Finke, D. and Bailer, S. (2010). 'Bargaining power in the European Union: An evaluation of competing game-theoretic models'. Political Studies, 58(1), pp. 85-103.
Selck, T. (2003). 'Evaluating the predictive power of a procedural model for the European Union legislative process'. Journal of Legislative Studies. 9(3), pp. 140-152.
Tallberg, J. (2006). Leadership and Negotiation in the European Union. Cambridge: Cambridge University Press.
Tallberg, J. (2008). The Power of the Chair: Formal Leadership by the Council Presidency. In: D. Naurin and H. Wallance (Ed.) Unveiling the Council of the European Union. Games Government Play in Brussels. Houndmills/New York: Palgrave Macmillan, pp. 187-203.
Trondal, J. (2007). 'The public administration turn in integration research'. Journal of European Public Policy, 14(6), pp. 960-72.
Trondal, J. and Veggeland, F. (2003). 'Access, voice and loyalty: the representation of domestic civil servants in EU committees'. Journal of European Public Policy, 10(1), pp. 59-77.
Van Els, T. (2005) 'Multilingualism in the European Union'. International Journal of Applied Linguistic, 15(3), pp. 263-281.
Van Schendelen, M. P. M. C. (1996) 'The Council Decides: Does the Council decide? ' Journal of Common Market Studies, 34(4), pp. 531-548.
897