Deliberative layering: Explaining diverse interest mobilization across the European Parliament's Policy Cycle

Correspondence Alexander Katsaitis, ARENA-Centre for European Studies, University of Oslo, Postboks 1143, Blindern 0318, Oslo, Norway. Email: alexander.katsaitis@arena.uio.no Abstract Drawing from work on deliberation and information-access, this paper conceptually frames why and when different types of interests mobilize across the parliamentary policy cycle. We posit that each policy stage holds its own deliberative purpose and logic, leading to a variation in the type and volume of information demanded. The legitimacy of the expertise interest groups provide is affected by their organizational characteristics. To ensure the smooth flow of the policy process, members of parliament encourage groups that legitimately hold relevant information to mobilize at each policy stage, while lobbyists choose to mobilize when their expertise allows them to better influence policy-makers' debates. We test our argument in the context of the European Parliament, following a unique survey of the 8th legislature (2014–2019). The responses lend support to our model. In a policy process that contains various stages of deliberation, different organizations hold an informationexpertise key that gives them access at different stages. Significantly, less studied groups, such as think tanks and consultancies, mobilize well ahead of others in the cycle's initial phases; while lobbyists representing public constituencies dominate in the final stages. The paper contributes to broader theoretical discussions on pluralism, bias, and deliberation in policy-making.


| INTRODUCTION
To protect its democratic credentials against accusations of business bias, the European Parliament encourages the mobilization of a diverse interest population to inform its policy-making. As a result, over the past 15 years, the lobbyists working with the institution have shifted from primarily corporate to more general societal interests (Coen & Katsaitis, 2019a;Lehmann, 2009). In this paper, we aim to explain when and why different interest groups mobilize across the EP's policy cycle.
Drawing from deliberative theory and information exchange models, we argue that each policy stage has its own deliberative purpose and logic, influenced by the number of policy-makers it involves and by its degree of openness towards different constituencies (Eriksen & Fossum, 2000). These factors lead to a concomitant variation in the type and volume of information demanded. On the supply side, the legitimacy of the information interest groups provide is affected by their organizational characteristics, notably the number of their principals and their lobbying objectives (Zürn, Binder, & Ecker-Erhardt, 2012). MEPs encourage groups that offer legitimate and relevant information to engage at each policy stage, while lobbyists mainly choose to mobilize when they expect that their expertise will allow them better to influence policy-makers' debates. We assess our model through a survey conducted with MEPs during the 8th legislature, asking how often different groups contact their office at different policy stages. The results reveal variation across the cycle in the mobilization of different interest group categories. This paper contributes to discussions of political pluralism and interest group mobilization in parliaments (Brack & Costa, 2018;Chaqués-Bonafont & Muñoz Márquez, 2016;Coen & Katsaitis, 2019b).
Theoretically, it develops a conceptual frame for analysing interest group mobilization across a variety of lobbying actors and during the whole policy cycle of the EP. It raises questions about unequal access, highlights hitherto under-researched third party groups that may have an impact on policy outcomes, and reaffirms the need to track lobbying footprints at the cycle's earlier stages. Empirically, the results provide an assessment of the lobbying activities faced by MEPs across the policy cycle. Finally, the paper has normative implications regarding the role of deliberative processes in policy-making by representative institutions (Fishkin & Mansbridge, 2017).

| THE POLICY CYCLE AND DELIBERATION
The EP faces a diverse population of groups providing information (inputs) to policy-makers that demand it, in exchange for inside information, influence over the final output, and insider status (Coen & Katsaitis, 2015, 2019a. Assuming that the EP wishes to maintain its democratic legitimacy, MEPs can be expected to demand information from interest groups to engage in debates that produce legitimate legislative outputs, a form of deliberation where policy-makers receive and process information to make policy-choices.
Because the EU's policy-making circuit is complex, institutional involvement and interest representatives' participation change as the process evolves. Each policy stage has its own feedback loops, but ultimately it must produce a useful component of the package and then move the legislative proposal towards the final output. It is rather unlikely that the entire population of lobbyists mobilize homogenously and that policy-makers' information demands remain stable across the cycle's steps.
With limited resources, both policy-makers and interest groups aim to maximize their utility by meeting when their demand and supply preferences correspond (Hall & Deardorff, 2006). At each policy stage, MEPs will primarily interact with the groups that legitimately supply relevant information, interest groups will prefer to mobilize when their information is in greater demand, and they expect to exert more influence. More generally, groups that are given better access at each policy stage will on average mobilize more than others.

| Supply and organizational structure
The legitimacy of interest groups' information is impacted by their organizational structure, operationalised by two criteria: the group's incentives and the number of its principals. The group's organizational incentives are a qualitative criterion that can be located on a continuum between two extreme profiles (Zürn et al., 2012). At one end, a group may only seek private benefits such as financial profit or increased market share. On the other, a group may only have public-spirited objectives, such as improving general welfare or advocating moral arguments. For example, banks are likelier to be invited to provide information on banking regulation, whereas civil society groups are more likely to be asked to provide information on human rights.
Secondly, the type of information a group can provide is affected by the number of principals it has. On one extreme, a group can represent a single private principal such as a single physical or legal entity (e.g., a natural person or an individual company). On the other extreme, a group can represent numerous principals such as a global constituency. Groups that have fewer principals represent interests set out by their clients (e.g., consultancies, law firms), their board (companies), or a finite number of members (associations, trade unions). Such groups have limited direct contact with the public domain, and are less dependent on public approval to reach their objectives. This predisposes them to operate within elite circles, and makes it less likely that their information will change over time due to socio-political factors. Groups with fewer principals may also be capable to function as a crystallising core within advocacy coalitions, connecting different groups under broader messages while maintaining versatility.
Interest groups that have as their main objective to benefit broader categories of principals, such as a local electorate in a region or municipality, or members of NGOs or religious groups serving complex public goals operate more closely with the public domain and need a stable or growing membership and/or public approval to retain their relevance (Katsaitis, 2015;Warleigh, 2001). The information they provide is reflecting public opinion and its fluctuations. The two variables are of course correlated: organizations with fewer principals are more often linked to private objectives, multiple principals are more often associated with organizations defending public-spirited objectives.

| Demand and deliberation
Demand for information at each policy stage is linked to the deliberative logic in place (Eriksen & Fossum, 2000). Considering the extensive literature discussing various criteria (see e.g., Curato, Dryzek, Ercan, Hendriks, & Niemeyer, 2017) we select inclusiveness and openness as the main variables that can help us distinguish the types of deliberation prevalent at different policy stages. To operationalise these two variables, we propose two measures that help us to describe the deliberation logic of each policy-making stage.
1. Policy-maker Inclusiveness, refers to the extent each policy step draws policy-makers from the EP to discuss and process information before moving on to the next step (the number of MEPs involved at each stage). As more MEPs participate in the legislative process, the overall demand for information increases, leading to more mobilization of interest groups. Except for trilogues, policymaker inclusiveness increases during the procedure, with more MEPs included at each step. Trilogues are informal meetings with representatives from the EP, the Council and the Commission that have been gradually institutionalised since wider use is made of the co-decision procedure.  In a parliament that represents different national and political interests this small group of MEPs has an incentive to keep the process limited to a few interest representatives that can provide information linked to normative and general issues, which will frame the discussion's main topics and constrain options for the broader set of actors at later stages. Therefore, the groups that are likelier to be asked to supply information are network architects representing a limited number of principals and committed to their objectives.

The Commission's Formal Proposal: Pragmatic Deliberation.
Once the Commission tables its formal proposal to the EP, the debate opens up to the competent committee in charge of the legislative file. Usually under pressure to produce a timely output, the committee members engage in a pragmatic deliberation (see Habermas, 1996). Policy-making now involves the most pertinent constituencies that have the capacity to understand the proposal's technical language and its likely political impact, and to propose effective amendments.
The committee's influential interlocutors, especially the rapporteur and the shadow rapporteurs, seek information from network architects who continue to provide information on central points of contestation or agreement with respect to broader coalitions within the committee. Simultaneously, the committee's MEPs engage in 'cheap-talk' (Farrel & Rabin, 1996) demanding input from organizations within their constituencies to discuss potential responses, courses of action and their political costs. Overall, this stage gives a competitive advantage to organizations that act as network builders, professional organizations with the capacity to represent private interests, and organizations representing local or thematically limited public constituencies.
The committee amendment and vote stage is a critical component in the EP's policy cycle (Marshall, 2010). Because parliamentarians face a significant workload, they outsource responsibility over each proposal to their most relevant colleagues in the responsible committee. If a legislative proposal passes this stage it is very likely to pass the plenary, making the committee's proposed draft resolution also the EP's final position. To maintain its legitimacy as a crucial institutional sub-unit, the committee takes time to consider and avoid potential political opposition that could arise against its proposal in the plenary. Put differently, the proposal that reaches the plenary must take into account the balance of political preferences to ensure its viability at the final stage. Simultaneously, to avoid electoral costs, the committee's MEPs act and wish to be perceived as their constituencies' active representatives.
Therefore, the committee must engage in a broader discussion that identifies compromises between specific constituencies' preferences and those of the broader electorate. At this stage, the MEPs are open to multiple interest constituencies. Every interest group category may be able to provide useful information and many groups are invited to engage with the committee. 4. Filtering the Debate: Trilogues.
Informal contacts and negotiations between the three institutions may occur at any stage of the legislative procedure but proper trilogues usually start after the responsible committee has adopted a negotiating mandate. Their purpose is to facilitate the debate within the EP's main political groups, to tackle disagreements between the Council and the EP, and to reduce the time it takes to produce the final legislative output (Reh, Heritier, Bressanelli, & Koop, 2013); as such trilogues allow limited access to external actors. This leads recurrently to critical discussions among the political leadership of the Parliament with regard to transparency and democratic legitimacy.
Our expectation is that member state governments have an insider track during this stage and mobilize to use it effectively. The MEPs involved in trilogues are likely to invite member states to provide information to better understand and discuss the dossier's progress before the formal proposal is accepted for the plenary debate.
Due to the lack of inclusiveness of trilogue negotiations, aggregate mobilization is weak at this stage.

Plenary Amendments and Plenary Vote: the Public
Deliberation.
Once the proposal passes the committee vote or, in many cases, the trilogue, it is submitted to the Parliament for the plenary vote, a process that leads to stronger politicization of policy-making (Grant, 2005). At this point most MEPs engage in the deliberation, considering how the electorate is likely to react to a vote in favour or against the proposal. This process is steered by political groups' leadership and entails a peak of the aggregate mobilization rate. The plenary serves as a forum where MEPs act as and represent ordinary citizens, striving to understand, assess and decide in accordance with broader political objectives and preferences. As informational input should represent significant electoral constituencies, civil society associations, regional authorities, religious organizations and member state governments have a competitive advantage.
We should note that plenary amendments are sometimes less contested because they are submitted by the rapporteur and the committee as a whole. In such cases, the objective is not to change core principles of the proposal but to adapt technical details with a view to final compromises. While we expect substantial aggregate mobilization rates in both phases, we expect that lobbying groups with a good technical understanding of the proposal mobilize more strongly during the plenary phase. To summarise, we outline our model in Table 1.

| SURVEYING MEPS
To assess our framework, we conducted an anonymised survey during

| Aggregate mobilization
The results reproduced in Figure 1 show that as policy cycle stages become increasingly inclusive (except for trilogues), interest groups contact MEPs more often. During the plenary vote, for instance, nearly half of the respondents were contacted 'Very Frequently', more than at any other moment. However, at the policy cycle's earlier

| Variation across the policy cycle
To better understand which are the interest groups involved at each stage, we assess below the results given in response to the question 'Do some types of interest groups contact you more often during different phases of the policy-making cycle?'. The results given in  invited to participate have mostly specific private objectives and few principals. But these lobbyists have the opportunity to shape the debate early on, potentially influencing the legislative proposal more than any other group at later stages. We note that during the committee amendments and votes all interest groups, across the board, increase their activities, confirming other studies assessing lobbying at the committee level (Rasmussen, 2015). As the committee´s task is to prepare the draft proposal for a resolution to be tabled to the plenary committee members need broad expertise to increase the chance for a strong vote in favour in the plenary.
Considering member state governments´relatively constrained lobbying during committee work one could surmise that their quasi- between the most notable lobbying groups, business and civil society: both mobilize significantly but at different stages. However, mobilization does not necessarily mean influence. Moreover, the assumption that different groups mobilize against each other does not fit the policy-making or lobbying logic employed in Brussels. Therefore, to grasp whether mobilizing at different moments entails losses in influence we need a better understanding of the coalitions in place and of access bias at specific stages. Think tanks capacity to mobilize at the policy cycle's earlier stages suggests that they have a better position to influence the discussion than business or civil society. But without an understanding of think tanks' broader coalition partners, such observations do not have much explanatory or comparative value.
However, the results do show that some groups are likelier to be invited at some points of the cycle than others.

| Variation and influence across groups
In the following section we proceed with two empirical questions following from the previous chapter: do some groups mobilize more than others? And, which groups are seen as more influential? In other words, regardless of the mobilization intensity during certain policy stages, how do MEPs perceive the overall contribution of specific interest groups to the deliberation on the whole? And how do MEPs assess these groups´impact on the final outcome?
To answer the first question our survey asked the MEPs 'How That is to say, simply because these groups mobilize more does not necessarily mean that they are perceived as equally influential. To assess that dimension, we asked MEPs 'How influential do you believe different types of interest groups are in the European Parliament?', providing them with the ten interest group choices and five ordinal options from´Not at all Influential´to´Extremely Influential´.
According to their responses (see Figure 4), MEPs perceive all organizations, including NGOs, as influential to some degree (only religious groups and law firms are an exception). This is in line with the work of scholars arguing that civil society has attained a prominent role in the EU's policy-making. From a methodological perspective, the results highlight issues of inferring influence from interest groups mobilization rather than output-based approaches. It also reflects the subjectivity of MEPs´impressions, emphasizing the impact of the biggest groups (e.g., business associations). Member states, NGOs, and companies are perceived as the most influential lobbying organisations but we have to remind ourselves that the real influence of private interests or member states is difficult to pin down in the multidimensional lobbying game.

| CONCLUSIONS
In this article, we concentrated on the differences in interest group mobilization across the EP's policy-making cycle. Our research question was to assess whether the wide spectrum of lobbyists working In conclusion, we found a layered structure where different types of deliberation are involved in the EP's policy-making. Public interest mobilization in the EP neither serves as an inclusive participatory tool for citizens, nor as a professionalized setting that serves solely a business bias. MEPs rather act as political entrepreneurs, selecting constituencies relevant to their deliberative logic in order to move the legislative proposal further along the policy conveyor belt. From this perspective, the results suggest that less visible groups such as consultancies or law firms are more easily recalled by MEPs because of the targeted activity they conduct at the cycle's less crowded phases.
The paper contributes to the theoretical analysis of pluralism, bias, and deliberation in the EU's policy cycle. We have seen that we cannot assume that across the policy-making cycle informational demands are homogenous or just divided up between the largest groups, that is, business and civil society. In a complex policy-making process that is made up of various logics of deliberation, different organizations seek to have appropriate access at different stages. We submit some exploratory hypotheses and data as a contribution to the literature on intra-institutional deliberation and the external factors influencing its outcomes. Our model of a multi-layered policy cycle analysis is based on data obtained from a survey of MEPs and their reactions to a diverse interest group population's activities. Hence, we provide a map of lobbying strategies as experienced by the MEPs, and show that some understudied interest group categories have substantial access to the EP legislative train at important stations. From a normative angle, this raises some questions regarding the EP's democratic legitimacy and the involvement of public interests at latter policy stages. At the same time, we wish to highlight that citizens' hold their own distinct procedures into Brussels' policy-making. Future research should aim to combine democratic theory and empirical research on interest group mobilization to advance this important field of public policy.