Serving one client only? Assessing the openness and inclusiveness of European Parliament's intergroups

ABSTRACT
 Openness and inclusiveness are central mechanisms for civil society participation in EU governance. Focusing on European Parliament’s intergroups, this article explores unofficial participatory venues and their receptiveness towards civil society. Based on a novel data set with 135 intergroup meetings and 435 civil society speakers across four intergroups in the 8th European Parliament (2014-2019), this study delivers the first empirical account of intergroup participants, and thereby provides original insights into the types and functions of intergroups as participatory mechanisms. Employing openness and inclusiveness to assess the density, diversity and insider status of civil society, this explorative study maps intergroups on a continuum of participatory vehicles that serve both pluralist interests and individual clients. It reveals an insider status for different types of interests across the four intergroups and points to the special role of EU associations in intergroups. By opening the black box of intergroups, this study explores an understudied phenomenon in EP-civil society research and sheds light on civil society practices taking place beyond the official structures of the Parliament. It adds to the literature on civil society access in EU governance and contributes to the wider political and academic debates on civil society’s contribution to the EU’s democratic legitimacy.


Introduction
Prompted by the publication of the European Commission's (EC) White Paper on European Governance (2001), civil society's contribution to the European Union's (EU) democratic legitimacy has been extensively discussed in various contexts, e.g.within the framework of participatory (e.g.Bouza Gracia, 2015;Kohler-Koch & Quittkat, 2013), associative (Sanchez Salgado, 2014) and deliberative democracy (e.g.Eriksen & Fossum, 2000;Liebert & Trenz, 2011;Steffek et al., 2008).The EU itself favoured participatory democracy, as a supplement to representative democracy, to counter the EU's legitimacy deficit.In this context, civil society is primarily portrayed as enhancing participatory-oriented input legitimacy or performance-oriented output legitimacy (Liebert & Trenz, 2011).
However, the value added of civil society participation to the EU's democratic legitimacy is contentious (Greenwood, 2007a).In the context of (informal) governance and policy networks voices have been critical as to whether civil society is 'friend or foe of input-based legitimacy' (Greenwood, 2007a, p. 355;Warleigh, 2003).EU networks need to display pluralistic and inclusive features to serve as friends of input legitimacy (Skogstad, 2003).Likewise, participatory approaches emphasis that participatory mechanisms need to fulfil certain conditions, namely inclusiveness, openness, and equal representation, to enhance input legitimacy (Kohler-Koch & Finke, 2007, p. 213;Papadopoulos & Warin, 2007).
Inspired by this theoretical work, this article explores European Parliament's (EP) intergroups and their openness and inclusiveness towards civil society.In doing so, the study explores a phenomenon that thus far received only a little attention in EP-civil society research (Landorff, 2019a;Ringe et al., 2013).Until now, studies focused on the official structures of the Parliament, for example, the parliamentary committees (Marshall, 2010;Rasmussen, 2015), political groups (De Bruycker, 2016;Marshall, 2015) and hearings (Coen & Katsaitis, 2019).Additionally, studies looked at social movement campaigns in the EP and the Parliament's support for participatory tools, such as the European Citizens' Initiative (Crespy & Parks, 2019).As a genuine co-legislator, the Parliament is nowadays a highly important target for all types of interests, ranging from business to diffuse interests (Ahrens & Woodward, 2021;Dionigi, 2017;Greenwood & Roederer-Rynning, 2020).
Intergroups feature sporadically in these accounts, although they have been part of Members' practices since the first direct election of the Parliament in 1979.Over time, they expanded continuously in scope and numbers, mounting to 28 officially recognised intergroups in the 8th EP (European Parliament, n.d.). 1 They gather Members of the European Parliament (MEPs) across political groups and parliamentary committees, representatives of other EU institutions and civil society in their meetings.Unlike political groups and parliamentary committees, intergroups are not official bodies of the EP.Hence, they operate in an institutional twilight zone.
Ever since they came into existence, intergroups have been under the suspicion of being too close to lobby groups (Corbett et al., 2016).Given the fact that legislative texts are discussed in intergroups (Landorff, 2019a), it is more important than ever to analyse who is represented in intergroups, and who can influence opinion formation in intergroups?Therefore, this article explores intergroups as participatory mechanisms by asking: Who is invited to speak at intergroup meetings?Are some interests more often invited to intergroup meetings, and hence more included than others?
Creating an original data set with 135 intergroup meetings and 435 civil society speakers across four intergroups in the 8th EP (2014)(2015)(2016)(2017)(2018)(2019), this study delivers the first systematic, empirical account of intergroup participants.In this way, it develops a typology of intergroup participants that provides new insights into intergroups and civil society practices taking place beyond the official structures of the Parliament.Employing Coen and Katsaitis (2019) innovative operationalisation of openness and inclusiveness, this study contributes to the literature on EU interest representation, particularly to the aspect of civil society access in EU governance.Overall, this research adds to the wider political and academic debates on civil society's contribution to the EU's democratic legitimacy.
The article is structured as follows: the next section reviews the literature on EP intergroups.Subsequently, the theoretical backbone of the article is outlined, and the two criteria of openness and inclusiveness are introduced.Thereafter, the data set and the operationalisation of openness and inclusiveness are presented.Next, the analysis examines the openness and inclusiveness of four intergroups, namely 'Welfare and Conservation of Animals', 2 'Seas, Rivers, Islands and Coastal Areas', 3 'Social Economy' and 'Biodiversity, Hunting, Countryside'.The results of the study are discussed in the final section of the article.

Intergroups in the European Parliament
The nature of intergroups is not easy to grasp.They differ substantially in their membership numbers, attracting up to 136 MEPs. 4 They vary in their meeting arrangements, organisational features and thematic foci.Some intergroups meet monthly (e.g.'Animal Welfare', 'Urban Areas') while others organise three meetings annually (e.g.'Social Economy').Some intergroups cooperate with an EU-level association as their external secretariat. 5Some intergroups pursue general goals (e.g.'Extreme Poverty and Human Rights'), while others represent very specific issues (e.g.'Rural, Mountainous and Remote Areas').Some intergroups work in line with the Parliament's legislative agenda (e.g.'Trade Union', 6 'Urban Areas' 7 ), while others work predominately on non-legislative issues (e.g.'Biodiversity, Hunting and Countryside' 8 ).
Intergroups have been presented as 'one of the main avenues for citizen interest groups lacking access to other institutional structures' (Cullen, 1999, as cited in Greenwood, 2017, p. 44; see Mény, 2009, for a similar argument).They have been studied regarding their potential as lobby groups.However, scholars are divided on intergroups as lobby venues.Bouwen (2004, p. 485) describes intergroups as 'attractive but unreliable arenas for […] legislative lobbying' as opposed to parliamentary committees, while Nedergaard and Jensen (2014, p. 11) conclude that 'many intergroups are affected by lobbyism', although to different degrees.
Recent works apply social network theory to conceptualise intergroups as 'relatively strong policy networks' that consist of formal and informal connections (Nedergaard & Jensen, 2014, p. 3).In this context, intergroups have been attributed with informational, coordinative, and signalling functions (Nedergaard & Jensen, 2014;Ringe et al., 2013).They serve as venues to gain knowledge, exchange information, and to coordinate and signal political views within and beyond the Parliament.Landorff's (2019a) sociological account of EP intergroups supports these findings.Additionally, she highlights that MEPs and civil society alike utilise intergroups to foster opinion formation and policy development in the Parliament.Moreover, she points to the role of EU associations acting as the secretariat of an intergroup (Landorff, 2019b).For instance, in the intergroup 'Animal Welfare', the Eurogroup for Animals is responsible for the organisation of meetings and the invitation of guests.It coordinates follow-up actions of the intergroup, helps to set up meeting agendas and provides updates on the intergroup's website (Landorff, 2019b; Intergroup on the Welfare and Conservation of Animals, 2019, p. 3).Hence, the European association exercises a substantial role in the functioning of the intergroup.
Departing from these latest works, the article explores whether participation in intergroups takes place in a circle of insiders or fosters wider public participation.Being interested in the overall openness and inclusiveness of intergroups, this study also examines how inclusive EU associations are to their members by providing them with direct participation opportunities in the intergroup.

Openness and inclusiveness of participatory mechanisms
This article approaches the legitimising and democratising potential of civil society from the perspective of participatory governance.Following the Commission's (2001) approach to Governance, participation is considered as a 'necessary supplement of representative democracy' (Kohler-Koch, 2013, p. 1).In this framework, civil society is comprised of 'trade unions and employers' organisations, nongovernmental organisations; professional associations; charities; grass-roots organisations; organisations that involve citizens in local and municipal life with a particular contribution from churches and religious communities' (European Commission, 2001, p. 11).
According to this definition, civil society's primary function is representation. 9Civil society adds to the EU's democratic legitimacy by providing voice to the concerns of people (input legitimacy), and by delivering services that meet the needs of citizens (output legitimacy) (Kohler-Koch & Quittkat, 2009, p. 14).Consequently, political pluralism is decisive for efficient and legitimate EU policymaking (Kohler-Koch & Quittkat, 2009, p. 15).Accordingly, participatory input-legitimacy is associated with 'pluralist interest representation and intermediation' (Liebert & Trenz, 2011, p. 8).Pluralist interest representation is characterised 'by no bias in favour of some particular interests […]' (Bunea, 2014, p. 11).
Departing from a principled conception of participation that links the democratic value of participation to standards, such as fairness, equality and inclusiveness (Kohler-Koch & Finke, 2007, p. 207), this study argues that participatory mechanisms need to fulfil certain conditions to improve the EU's democratic legitimacy.Openness, inclusiveness and equal representation constitute important yardsticks for judging the democratic quality of participatory processes (Kohler-Koch & Finke, 2007, p. 213).Given that openness and participation are cornerstones of EU governance, their usage has been extensively discussed in the literature.This article is inspired by the following two research lines: First, openness and inclusiveness have been discussed in the overall context of EU interest intermediation, though with a focus on the Commission, addressing issues of representativeness, unequal participation and access of civil society organisations (CSOs) in EU policymaking, and thereby pointing to the role of elites in consultative and deliberative fora (e.g.Kohler-Koch, 2010, 2013;Liebert & Trenz, 2009;Smismans, 2003;Steffek et al., 2008).Since previous research highlights the role of EU-level organisations in intergroups (Landorff, 2019b), the aspect of representativeness is relevant to the present study.
In terms of representativeness, studies focus on collective organisations and how the inclusion of European associations in the Commission's consultation regime 'would make the European integration process more inclusive' (Smismans, 2003, p. 491).In this context, scholars addressed the question of how receptive European umbrella organisations are in representing their members at the supranational level (Greenwood, 2007b;Kohler-Koch & Quittkat, 2013;Kröger, 2019;Smismans, 2003).
Second, openness and inclusiveness have been studied in combination with the concept of throughput legitimacy to assess the procedural standards of EU governance (Schmidt, 2013;Schmidt & Wood, 2019).This body of research offers a neat definition of openness and inclusiveness that this article adopts.
Accordingly, openness is defined as 'the extent of opportunities for non-state organisations and individuals to become involved in EU consultations and civil society umbrella groups' (Schmidt & Wood, 2019, p. 733).Inclusiveness is understood as the 'breadth of non-state organisations and individuals involved in EU consultation and civil society public and private interest groups' (Schmidt & Wood, 2019, p. 733).Hence, the question of legitimacy is explored through the balance of access among different civil society groups (Schmidt, 2013).
In this line of research, Coen and Katsaitis (2019) offer an excellent example of how to operationalise openness and inclusiveness.Employing throughput legitimacy in the study of EP committee hearings, they develop two measurements for openness and inclusiveness that allow an empirical assessment of how receptive hearings are to interest groups.In this way, they show how parliamentary hearings do not serve the purpose of increasing 'citizens' participation via public interest groups' but rather serve epistemic and coordinative functions (Coen & Katsaitis, 2019, p. 766).Inspired by their work, this article applies openness and inclusiveness to assess how accessible intergroups are towards civil society interests.

Data selection and operationalisation
This article pursues an explorative approach to the study of intergroups.Being interested in the general accessibility of intergroups to civil society, and the receptiveness of European associations to their members' interests, four intergroups with an external secretariat have been selected for the analysis (Table 1).
These intergroups cover a breadth of topics in terms of the policy areas that they represent 10 and, associated therewith, a wide spectrum of concerns, ranging from public to private interests. 11Moreover, these four intergroups provided the most comprehensive data sets. 12Their meeting agendas were publicly available on their websites or distributed via a mailing list. 13 As a result of its explorative approach, the study looks at four intergroups that vary in their meeting frequencies during the 8th EP.Further details are provided in the analysis.These differences need to be considered in terms of the opportunities that intergroups offer for civil society to become involved in the consultation.
Overall, 135 agendas across the four intergroups have been analysed.This resulted in a sample of 797 invited speakers during the 8th EP (2014)(2015)(2016)(2017)(2018)(2019).These 797 speakers have Regarding the operationalisation of openness and inclusiveness, the article draws on Coen and Katsaitis (2019).First, openness is assessed as absolute openness, that is, how many speakers representing civil society (incl.think-tanks, research and academic institutions), have participated in intergroup meetings during a specific period.For instance, a total of 88 civil society speakers attended the intergroup 'Animal Welfare' during the 8th EP (Appendix A).
Secondly, openness is assessed as relative openness, that is, how open intergroups are to different types of organisations (e.g.professional consultancies, in-house lobbyists, NGOs, etc.).Relative openness measures different types of organisations as a percentage to the absolute number of speakers (Coen & Katsaitis, 2019, p. 2).Inclusiveness is operationalised as outsider inclusiveness, that is, whether some organisations and speakers have received repeated access to the intergroup.Repeated access may point to some sort of insider/élite status of the respective interest.To calculate outsider inclusiveness, the total number of groups per type of organisation is divided by the total number of speakers per type of organisation and multiplied by 100 (Coen & Katsaitis, 2019, p. 3).

Inclusive Outsider =
Total number of Groups per Type of Organisation Total number of Speaker per Type of Organisation × 100.(2) Thereby, an outsider score of 100% indicates that each speaker in this section represented one distinct organisation.A score of 80% indicates that 80% of the participants represented one distinct organisation, while 20% of the speakers received repeated access.Hence, the lower the outsider score, the more speakers of one and the same organisation gained repeated access.

Openness towards civil society
Departing from the definition of openness as 'the extent of opportunities' for individuals or organised groups to have access to policy makers (Schmidt & Wood, 2019, p. 733), it is notable that the Parliament provides intergroups with official meeting slots on Thursday mornings during the monthly plenary sessions in Strasbourg (European Parliament, 1999Parliament, / 2012)).
However, the four intergroups under consideration vary substantially in their meeting frequencies during the 8th legislative term: the intergroup 'Animal Welfare' met monthly, 15 while the intergroup 'Biodiversity, Hunting and Countryside' met roughly every other month. 16The intergroup 'Social Economy' met three to five times a year. 17 The intergroup 'SEARICA' met twice a month, monthly or every other month. 18With these meeting patterns the four intergroups cover the spectrum of meeting arrangements identified in the overall population of intergroups in the 7th EP (Landorff, 2019a).Thus, the four intergroups offered different opportunities for civil society to access policy makers.The numbers on their absolute openness towards civil society reflect the different practices (see Table 2).

Absolute openness towards civil society
Initially, it shows that civil society represents over 50% of all invited (coded) speakers in the intergroups 'Animal Welfare' (60%), 'SEARICA' (58%) and 'Biodiversity, Hunting and Countryside' (67%), while it accounts for less than 50% of invited speakers in the 'Social Economy' intergroup.This constitutes an important insight regarding the initial purpose of intergroups to promote contact between MEPs and civil society (European Parliament, 2014Parliament, -2019, p. 29), p. 29).In the 'Social Economy' intergroup, representatives of the European Commission and MEPs are nearly as much sought as speakers as civil society actors (see Appendix A).
Moreover, the four intergroups differ in their absolute openness towards civil society.The intergroup 'SEARICA' invited 237 civil society speakers to 44 meetings, while 88 civil society representatives were invited to speak at 56 meetings of the 'Animal Welfare' intergroup.Sixty-six civil society speakers attended 23 meetings of the 'Biodiversity, Hunting and Countryside' intergroup, and 44 civil society representatives spoke at 12 meetings of the intergroup 'Social Economy'.These differences cannot be attributed exclusively to different meeting frequencies.The 'SEARICA' intergroup conducted less meetings than the 'Animal Welfare' intergroup.Though, different meeting formats appear to facilitate higher civil society participation.The intergroup 'Animal Welfare' typically conducted one hour, one-and-a half hour meetings with two to three invited speakers.The 'Social Economy' intergroup conducted its meetings as two-hour public hearings and/or conferences with up to 18 speakers. 19 The intergroups 'Biodiversity, Hunting and Countryside' and 'SEARICA' mainly ran their meetings as (half-day) conferences with up to 16 speakers. 20Thus, the conference format allowed for higher participation of guests in the first place.Moving from absolute openness to relative openness, Table 3 shows how open the four intergroups have been towards different types of organisations as a percentage of the absolute number of speakers.

Relative openness towards civil society
It shows that the intergroups 'SEARICA', 'Social Economy' and 'Biodiversity, Hunting and Countryside' were more open to a diverse mix of interests, while the 'Animal Welfare' intergroup mainly provided access to one type of interest, namely non-governmental organisations (NGOs).
Furthermore, the numbers reveal that the 'SEARICA' intergroup granted nearly equal access to 'In-house lobbyists et al.' (20%) and 'Non-governmental organisations et al.' (15%), while 'Think tanks et al.' (32%) and 'Organisations representing local, regional and municipal authorities et al.' (29%) were almost on par with each other.In this context, it is conspicuous that speakers were announced on intergroup agendas as representing the 'scientific perspective', 'civil society perspective', 'legislative perspective' and 'industry perspective'. 21This points to an ambition of the intergroup to grant access to a diversity of organisations.The intergroups 'Biodiversity, Hunting and Countryside' and 'Social Economy' invited speakers from across the same types of different interests.However, they did not cover as many diverse types of organisations as the 'SEARICA' intergroup.
Additionally, Table 3 highlights that the four intergroups differ regarding the type of organisation that they primarily invite.Constituting over 80% of all invited civil society speakers, 'Non-governmental organisations et al.' were predominantly granted access to the 'Animal Welfare' intergroup.'Think tanks et al.' took the lead in the 'SEARICA' intergroup with 32%.By contrast, 'In-house lobbyists et al.' represented the largest individual group of organisations being invited to the 'Social Economy' intergroup (48%).In the 'Biodiversity, Hunting and Countryside' intergroup, 'Non-governmental organisations et al.' took the lead with 48%.Though, they were not as dominant as in the 'Animal Welfare' intergroup.
Regarding the openness of intergroups towards civil society, the first part of the analysis revealed the following patterns: Initially, intergroups are receptive to civil society, though they differ in absolute and relative terms.The 'SEARICA' intergroup stood out as being particularly open to civil society in absolute terms.By contrast, civil society speakers in the 'Social Economy' intergroup accounted for less than 50% of all speakers.This may indicate a different purpose of the intergroup.To sustain this claim, further systematic research is necessary.
In terms of relative openness, the intergroups 'SEARICA', 'Social Economy' and 'Biodiversity, Hunting and Countryside' were more open to a diverse mix of interests, ranging from private to public to research interests, while the 'Animal Welfare' intergroup was mainly, and nearly exclusively, open to public interest organisations.
In terms of their primary participants, that is, the type of interest with the largest percentage in relative terms, the four intergroups differ again.The 'Animal Welfare' and 'Biodiversity, Hunting and Countryside' intergroups attracted mainly NGOs, while the 'Social Economy' intergroup attracted predominantly businesses, trade and business associations.In the 'SEARICA' intergroup research and academic institutions (mainly universities) constituted the largest individual interest.Returning to the idea of intergroups functioning as a venue for opinion formation, these results indicate that the debates in the respective intergroup are shaped by these primary speakers, though not exclusively.

Inclusiveness towards civil society
Outsider inclusiveness measures whether some types of organisations gained repeated access to the intergroup and therefore, enjoyed an insider status that would compromise input legitimacy.Linking inclusiveness to the representativeness of EU-level associations, particular attention is paid to the European umbrella organisations that act as an intergroup secretariat.

Welfare and Conservation of Animals & Eurogroup for Animals
For the intergroup 'Animal Welfare', an outsider score of 100% for the sections of 'Professional consultancies et al. ' and 'In-house lobbyists et al.' indicates that each speaker in this section represented a distinct organisation (Table 4).
Thus, the 'Animal Welfare' intergroup was open to invite speakers from different kinds of private interests (e.g.'Professional consultancies et al. ' and 'In-house lobbyists et al.').None of the speakers in these sections enjoyed repeated access.However, the low absolute number of speakers in these sections indicates that these types of interests were not the main target of the intergroup.As revealed in the previous section, the intergroup was particularly open to non-governmental organisations.
Here, an outsider score of 45.95% for 'Non-governmental organisations et al.' shows that 54% of the speakers in that section represented their distinct organisation more than once.Hence, the 'Animal Welfare' intergroup was less open to invite speakers from different NGOs.It is notable that the Eurogroup for Animals and its members fall under the category of NGOs.Hence, the type of interest with the lowest score for outsider inclusiveness resembles the type of organisation that hosts the intergroup secretariat.In the section of 'Non-governmental organisations et al.', the Eurogroup and its members accounted for 50 out of 74 speakers.The Eurogroup for Animals and its members 22  were present at nearly 70% of all intergroup meetings (39 out of 56).They accounted for 50 out of 88 civil society interventions, and thus for 57% of all civil society speakers across the various sections.Members of the Eurogroup spoke at 23 out of 56 intergroup meetings, and thereby delivered 28 interventions.The Eurogroup for Animals spoke at 22 meetings and delivered 22 presentations.These numbers indicate that the Eurogroup maintained regular contact with its members in the intergroup and provided them with an opportunity to present their interests directly to EU policy makers.Except for non-governmental organisations, the intergroup largely displayed inclusive features.However, the outsider score of 48% for 'Organisations representing local, regional and municipal authorities, et al.' is noticable.This means that 52% of the speakers represented one distinct organisation.In other words, the intergroup was less inclined to invite speakers from different organisations representing local, regional and municipal authorities, et al.As observed in the 'Animal Welfare' intergroup, the section with the lowest outsider score hosted the secretariat of the intergroup and its member.The Conference of Peripheral Maritime Regions (CPMR) and its members represented more than 78% of all speakers (54 out of 69) in the section of 'Organisations representing local, regional and municipal authorities, et al.'.Thus, the intergroup secretariat enjoyed an insider status in terms of representing this particular interest.In absolute terms, however, the CPMR and its members accounted for roughly one-fifth of all civil society speakers (54 out of 237).
Concerning the aspect of representativeness, the CPMR and their members were present at 25 meetings (out of 44) and delivered 54 interventions.Twenty-nine out of 54 CPMR interventions were made by a CPMR member.Thus, the CPMR gave precedence to its members and provided them with a direct opportunity to participate in EP consultation.

Social Economy & Social Economy Europe (SEE)
For the intergroup 'Social Economy', Table 4 reveals an outsider score above 50% for all sections with speakers.The lowest outsider scores are displayed for 'In-house lobbyists et al.' and 'Non-governmental organisations et al.'.Thus, the intergroup was less open to invite speakers from different in-house lobbyists and NGOs.This is notable as these two types of organisations represented the types of interests that were mainly granted access to the intergroup.Moreover, Social Economy Europe (SEE) has been coded in the section of 'In-house lobbyists et al.', while some members of SEE have been categorised as 'Non-governmental organisations et al.'.As detected in the intergroups 'Animal Welfare' and 'SEARICA', the type of interest that resembles the intergroup secretariat displays the lowest outsider score.
Social Economy Europe and its members accounted for 50% of all civil society speakers (22 out of 44).They were present at 10 out of 12 meetings and provided 22 interventions. 23Thereof, eight presentations were made by Social Economy Europe itself, while 13 presentations were given by a member of SEE. 24Hence, the intergroup constituted a participatory venue for the members of Social Economy Europe to voice their domestic input to EU policy makers.

Biodiversity, Hunting and Countryside & Federation of Associations for Hunting and Conservation of the EU (FACE) & European Landowners' Organisation (ELO)
For the intergroup 'Biodiversity, Hunting and Countryside', outsider scores well above 50% are documented for all sections with speakers (Table 4).25% of 'In-house lobbyists et al.' speakers enjoyed repeated access, while 19% of 'Non-governmental organisations et al.' and 13% of 'Think tanks et al.' speakers presented their organisation more than once.These numbers indicate that the intergroup was open to invite speakers from different private, public and research organisations.In the case of 'Professional consultancies et al.' (100%) and 'Organisations representing local, regional and municipal authorities et al.' (100%), each speaker represented one distinct organisation.However, the low absolute numbers for these types of organisations indicate that these types of interests were not the main audience of the intergroup.
FACE and ELO representatives spoke at two-thirds of all analysed meetings (15 out of 23).Overall, they provided one-third of all civil society interventions (21 out of 66).FACE and ELO gave priority to their members in providing input to intergroup discussions.Out of the 21 interventions, 10 presentations were made by a FACE member or partner, while five presentations were made by an ELO member.In terms of its intergroup activities, the European association has been receptive to its members.
Regarding the inclusiveness of intergroups towards civil society, the second part of the analysis revealed the following patterns: Overall, the four intergroups displayed inclusive features as outsider scores above 50% were calculated for most of the different types of interests.
However, the 'Animal Welfare' and 'Social Economy' intergroups were less inclined to invite different types of NGOs.The 'SEARICA' intergroup was less open to invite different types of organisations representing local, regional and municipal interests, while the 'Biodiversity, Hunting and Countryside' intergroup was less open to invite different types of 'In-house lobbyists et al.' to its meetings.For the three intergroups 'Animal Welfare', 'SEARICA' and 'Social Economy' the numbers indicate that the intergroup secretariat enjoyed an insider status, that is, it was granted repeated access to represent its particular interest.Finally, all five European associations provided their members with direct opportunities to present their cause.Numbers indicate that these European associations have been receptive to their members in the context of the intergroup.

Discussion
The analysis revealed patterns of openness and inclusiveness across the four intergroups that reflect the multifaceted nature of intergroups.Differences in absolute and relative openness can be attributed to different meeting formats and frequencies but also to different functions of intergroups.
The strong orientation towards research and academic institutions in the intergroup 'SEARICA' indicates that invitees are primarily expected to provide scientific expertise.This could be interpreted as an attempt to depoliticise opinion formation in the intergroup (for a similar argument on EP public hearings, see Coen & Katsaitis, 2019).A strong focus on academic expertise would also point to an epistemic character of the intergroup, that is, to the intergroup being a knowledge-based expert network (Haas, 1992).Though, the analysis also showed that organisations representing regional interests ranked shortly after research and academic institutions and have been granted with an insider status (e.g. the CPMR and its members).Given that the intergroup deals with maritime affairs, this finding points to a coordinative-informational function of the intergroup, that is, as a participatory venue it allows for an exchange of views between potential stakeholders and policy makers.
The primary orientation towards business interests in the 'Social Economy' intergroup points to a similar interpretation.The intergroup predominately serves as a participatory mechanism for representatives of the social economy (cooperatives, mutuals and social enterprises).A strong orientation towards public interests (e.g.non-governmental organisations) in the intergroup 'Animal Welfare' may be interpreted as enhancing public participation in EP consultation.However, the analysis showed that the 'Animal Welfare' intergroup almost exclusively serves animal welfare organisations, and therefore one client only.
Coen and Katsaitis' (2019) operationalisation of openness constitutes a handy tool to assess the receptiveness of intergroups towards certain interests.However, it does not account for the fact that speaking at intergroup meetings requires an invitation.Here, intergroups resemble official parliamentary organs, such as EP public hearings.Invitations to intergroup meetings are issued by the intergroup chair, bureau and/or the secretariat, which highlights again the important role of EU associations in intergroups.
Finally, the usage of the Commission's pluralistic definition of civil society needs to be addressed.This all-encompassing interpretation paints a favourable picture of intergroups as a participatory mechanism for civil society.Though, civil society is a contested concept.Including business and professional interests under the umbrella of civil society is highly controversial (Kohler-Koch & Quittkat, 2013).This analysis showed that intergroups are no longer 'allies' for citizen interest groups only.Instead, they are also participatory venues for business and professional interests (e.g.intergroups 'Social Economy', 'Biodiversity, Hunting, Countryside), territorial interests (intergroup 'SEARICA') and academics (intergroup 'SEARICA').Moreover, it is predominantly organised interest as opposed to individual citizens that participate in intergroups.

Conclusion
Inspired by participatory approaches in EU civil society research, this explorative study opened the black box of EP intergroups to examine the accessibility of these unofficial participatory venues to civil society.Driven by the questions of 'Who is invited to speak at intergroup meetings?' and 'Are some interests more often invited to intergroup meetings, and hence more included than others?', the article identified and mapped participants across four intergroups in the 8th EP, and thereby detected the following patterns: Concerning their openness and inclusiveness towards civil society, the four intergroups can be depicted on the continuum between two opposites, represented by the intergroups 'Animal Welfare' and 'SEARICA'.The former represents a type of intergroup that is mainly open to one type of interest and grants this type of interest an insider status.The latter represents a type of intergroup that grants access to a diversity of interests and acts inclusively towards all types of interests, except for one.The intergroups 'Social Economy' and 'Biodiversity, Hunting and Countryside' settle in the middle.As a result, opinion formation is shaped by a mix of different interests (public, private and scientific) in the intergroups 'SEARICA', 'Biodiversity, Hunting and Countryside' and 'Social Economy', while political views in the 'Animal Welfare' intergroup are mainly shaped by NGOs/public interest.
Linked to this observation, the study revealed different types of primary participants of intergroups, ranging from business/private interests ('Social Economy') and NGOs/public interest ('Animal Welfare', 'Biodiversity, Hunting and Countryside') to research organisations ('SEARICA').To a certain degree these prime speakers can be linked to the thematic focus of the intergroup, though not exclusively.Finally, the study showed that European umbrellas enjoyed an insider status in three out of the four intergroups.Hence, they were granted repeated access to represent their particular interest.
Overall, this research shows that intergroups have the potential to contribute to pluralist interest representation (see 'SEARICA' intergroup).However, they may also serve as a participatory vehicle for one client only (see 'Animal Welfare' intergroup).Consequently, this explorative study shall serve as a trailblazer for future research on EP intergroups.The following suggestions may serve as an inspiration.
This study focused on the numerical presence of civil society in intergroups.The data does not allow to draw conclusions on the diversity of opinions or the plurality of ideas that are discussed in intergroups.Future studies may analyse intergroups from the perspective of deliberative democracy.In combination with the concept of throughput legitimacy (Schmidt, 2013), studies could address the question of whether the involvement of (scientific) experts in intergroup deliberations leads to better throughput legitimacy by enhancing the quality of the debates.
Further, this study showed that EU associations provide their members with direct participation opportunities in intergroups.Admittedly, they did so to different degrees depending on the frequency of their meetings.It remains unclear how satisfied domestic constituencies are with their representation at supranational level.Conceptualising representativeness as 'consisting of regular contacts' and satisfaction with representation, studies may examine the role of European associations in intergroups in a more comprehensive way (see Kröger, 2019, p. 770).
This study revealed different types of participants across intergroups.Drawing on the debate on élite versus chameleon pluralism in the EU interest group literature (Coen & Katsaitis, 2013;Eising, 2007), future research may explore the causes for the variance of (organised) civil society presence across intergroups more thoroughly.
Finally, this study indicated that civil society may not be the main target of intergroups (see 'Social Economy' intergroup).Future studies may investigate the motivation of European Commission representatives to participate in intergroups.

Notes
1.For an overview, see European Parliament (n.d.) Relative = Number of Speakers per Type of Organisation Total Number of Speakers × 100.(1)

Table 1 .
Intergroups, secretariats, meetings and members (2014-2019).beencoded in line with the sections of the Joint Transparency Register (JTR).14ThesectionsreflecttheCommission's (2001)all-inclusive definition of civil society.Throughout the coding process, the sections have been modified to accommodate as many organisations as possible.Additional categories have been set-up to account for representatives of EU/European institutions, Member States, intergovernmental organisations and individuals.Intercoder reliability has been ensured through at least two independent coders, that is, at least two different researchers agreed on how an organisation/ speaker shall be coded.Like this, 435 out of 797 speakers have been identified as civil society.For a detailed overview of the categories and speakers, see Appendix A.

Table 4 .
Outsider inclusiveness of intergroups in %(2014)(2015)(2016)(2017)(2018)(2019).Except for 'Organisations representing local, regional and municipal authorities, et al.', all sections display an outsider score above 50% (Table 4).With the highest outsider score for 'Professional consultancies et al.' (100%) and 'In-house lobbyists et al.' (83%), the intergroup was particularly open to invite speakers from different private interests.By contrast, an outsider score of 69% for NGOs and 75% for 'Think tanks et al.' indicates that the intergroup was less open to invite speakers representing different public interests and research organisations.