The two faces of conflict: how internal and external conflict affect interest group influence

ABSTRACT

Conflict is one of the most fundamental concepts in the interest group literature. The more conflict there is, the less likely it is that interest groups can impact policymaking. In this paper we systematically explore whether some types of conflict affect interest-group influence more than others. More precise whether conflict among organizations that are perceived by policymakers as homogenous (‘internal conflict’) is more detrimental for the influence interest groups have on policymaking than conflict among groups that are not perceived by policymakers as homogenous (‘external conflict’). Empirically we focus on four cases of EU policymaking with varying levels of internal and external interest group conflict. Our case studies highlight that agreement among similar types of lobby organizations is a necessary condition for these groups to influence policymaking, while external conflict still provides much opportunities for interest groups to influence policymaking.

KEYWORDS:

• Interest groups
• influence
• conflict
• European Union

Introduction

Why are some interest organizations more successful in influencing public policy than others? Over the past decades scholars have sought to answer this question, but a simple answer remains absent (Lowery, 2013). On the one hand, this is due to the difficulty in defining and operationalizing influence (Dür, 2008). On the other hand, ambiguity on interest-group influence is also a reflection of the ‘messiness of politics’, where the specific context is critical for who wins and who loses (Lowery, 2013, p. 3). In line with this view, a great number of scholars have sought to identify issue characteristics, which condition the success of interest groups’ chances. Examples include the position towards the status quo (cf. Baumgartner et al., 2009; Dür et al., 2015), the salience of issues (cf. Junk & Rasmussen, 2019; Junk, 2019; De Bruycker & Beyers, 2019), the position in the policy cycle in which the lobbying takes place (cf. Bevan & Rasmussen, 2020), and the level of public support (cf. Giger & Klüver, 2016; Rasmussen et al., 2018; De Bruycker, 2020).

Of all the issue characteristics identified to affect the success of interest groups, among the most critical is the level of conflict (Schattschneider, 1960 ; Baumgartner et al., 2009; Dür et al., 2015; Mahoney, 2007; Dür et al., 2015; Rasmussen, 2015; Junk & Rasmussen, 2019; Chalmers, 2020; but see Klüver, 2011). Conflict refers to a clash of interests between interest groups, mostly operationalized as the division of policy position surrounding a policy issue, and is widely seen as limiting the chances for lobbying success (Dür et al., 2015). While much attention has been devoted to the relationship between conflict and influence, many questions remain unanswered. In this paper we set out to add two innovations. As a first contribution to the literature, we explore how different types of conflict within lobby communities affect their influence on policymaking. We are especially interested in how ‘unexpected’ conflict in lobby communities, such as among a set of seemingly coherent business groups or among a group of NGOs, impacts the influence these organizations have on policymaking. To fully understand these dynamics, we juxtapose cases where we found much conflict among similar type of organizations (which we call ‘internal conflict’), with cases where there was more ‘natural’ conflict between groups which are already perceived by policymakers as standing on opposite sides of most conflicts; such as conflict between a group of business organization on one side and a group of NGOs on the other side. We call this type of conflict ‘external conflict’. By contrasting these two types of conflict, we can see how important unity among similar type of groups may be for gaining influence in policymaking.

Our second contribution results from our research design. The current literature relating conflict to influence (like interest-group studies more broadly) is generally of a quantitative nature (but see Dür & Mateo, 2014; Rasmussen, 2015). The clear benefits of such quantitative approaches are the ability to control for many alternative explanations, and the generalizability of the relationships found. Yet, a downside is that it is less clear why conflict exactly matters for the success of interest groups, as quantitative analyses are less equipped to address underlying causes of the relationship. To overcome this limitation, we rely on four case studies of interest-group conflict at the European Union (EU) level. Our objective is to generate plausible causal mechanisms, which tie various types of conflict between and within interest-group communities, to policy influence. As a theory-building exercise, it is our explicit aim to generate tentative hypotheses that can be further tested in future, quantitative, studies.

Theory: the role of conflict for interest-group influence

Interest groups are seen by policymakers as important intermediaries between constituents and politics. Lack of time and expertise to follow up on each and every piece of information, however, prevents policymakers from being able to value every single argument provided by interest groups. This should lead them to rely on cues and heuristics. While heuristics are broadly understood as important drivers of decision-making (Lau & Redlawsk, 2001; Miler, 2009; Gigerenzer & Gaissmaier, 2011), the logic of heuristics has hardly been used to explain why (some) interest groups are more influential than others (but see La Pira, 2008; Miler, 2009). This is surprising as it seems likely that heuristics affect how lobbying is perceived by policymakers. For example, if many NGOs mobilize on a certain issue, this likely indicates to policymakers that the constituents of these types of organizations care about this issue. Even though it may be a (very) biased representation of citizen priorities in society, policymakers could still be triggered to consider it is a relevant issue due to the mass mobilization with which they are confronted.

Another potential heuristic policymakers may rely on is the level of conflict in interest-group community. The degree of conflict is defined as the disunity of policy positions within the interest-group community. When conflict is high, resulting in many conflicting positions among involved interest groups, the literature suggests that the opportunities for (some type of) interest groups to influence policymaking are compromised (Schattschneider, 1960; Baumgartner et al., 2009; Dür et al., 2015; Mahoney, 2007; Dür et al., 2015; Rasmussen, 2015; Junk & Rasmussen, 2019; but see Klüver, 2011). Baumgartner and Jones note in this regard that ‘nobody likes protracted conflict and continual competition. Much preferable to a system of constant conflict is one where each side retreats into a given area where its influence is uncontested’ (1993, p. 6). The reasoning is that conflict signals to policymakers that there is much division among constituents. Whatever choice policymakers thus make, they will not have full support in society or the business community. In such circumstances, it becomes easier for policymakers to ignore (at least parts of) the lobby demands and follow their own path. In contrast, if conflict is low, this signals broad support for the positions conveyed by the interest groups. This makes it much harder for policymakers to ignore these demands. High conflict should therefore lead to less potential influence of interest groups, while low conflict tends to lead to more opportunities for influence (Mahoney, 2007, p. 48; Dür et al., 2015, p. 968).

Conflict, however, does not affect the chances of all types of organizations equally. As Schattschneider (1960) already alluded to long ago, it should be the weak who want to expand the scope of conflict. More specifically, this implies that NGOs have more to gain when conflict increases, while business organizations prefer to keep conflict contained. Empirical work seems to confirm this assertion. For instance, Dür et al. (2015, p. 970) find that at high levels of conflict business organizations are far less likely to be successful in their lobbying efforts than NGOs, while at low levels of conflict there is no difference in the chances of NGOs and business groups. This confirms the view that conflict is mainly a problem for business groups, which they therefore seek to avoid. NGOs on the other hand have a strong incentive to expand the scope conflict, especially when they are up against business groups.

Another important addition to this debate relates to the level of unity that lobby coalitions convey to policymakers. Nelson and Yackee (2012) show that it is very important for lobby coalitions to speak with one voice – thereby also containing the level and scope of conflict in a policymaking process – to maximize the influence of groups. In this view, conflict is not related to the overall conflict among interest group camps, but refers to the level of conflict (or unity as the flipside) within a set of actors which stand on the same side of the larger political conflict. Junk and Rasmussen (2019, p. 501) highlight how the unity with which such ‘lobby camps’ promote dominant frames has a strong effect on the lobby success of the interest groups. Rasmussen (2015) has similar findings. In an extensive discussion of four lobby cases, she finds that one of the most important factors to explain business success is when ‘European business federations are internally united’ (2015, p. 380).

In this paper we build on these studies and explore further how unity and conflict affect interest-group influence on policy outcome. More precisely, we aim to see whether it matters if unity and conflict emerge in sets of actors that may be perceived by policymakers as homogenous or heterogeneous. We discuss the difference between these types of conflict in the next section.

Weighing the effect of internal and external conflict for influence

Our theoretical curiosity focuses on conflict among sets of organizations which are expectedly perceived by policymakers to represent similar types of interests, but are either in open conflict regarding the specific political demands or convey a united message. More specifically, we conceptualize two types of conflict. External conflict refers to conflict between a set of NGOs and a set of business groups. Internal conflict, instead, refers to conflict among business groups or among a set of NGOs. While we recognize that business groups and NGOs also form coalitions among each other and sometimes share a common position (Junk, 2019), this is by far a minority in the EU (our case). Beyers and De Bruycker (2018, p. 972) highlight that only 9 per cent of the NGOs active around a random set of 70 legislative proposals in the EU participated in such heterogeneous coalitions, and only 6 per cent of the business groups participated in such coalitions. As such we can assume that conflict between different types of interest groups (such as between business groups and NGOs) should be perceived by policymakers as more typical, and conflict among such similar types of group types as atypical.

Our aim is to explore how internal and external conflict could affect the influence interest groups have on policymaking. There are certainly reasons to expect that it matters whether conflict occurs within sets of more homogenous group types or not. That is, conflict among groups which are perceived by policymakers to share a common concern, such as conflict among NGOs, could (i) confuse policymakers, incentivizing them to postpone decisions as they see too little support for the demands of, or receive contradictory information from, interest groups; or (ii) provide an opportunity for policymakers to surpass the demands of interest groups and follow their own agenda. In both cases, internal conflict should limit interest-group influence. In contrast, external conflict, i.e., conflict between groups that are already perceived by policymakers to be on the opposite side of a conflict – such as a lobby camp of NGOs on the one hand and a camp of business groups on the other – offers a clear and rather typical ideological debate to policymakers. In these cases, it becomes more difficult for policymakers to ignore the demands of the lobby community and they are incentivized to either pick a side or find a compromise.

To explore the plausibility of our ideas, this paper focuses on four cases of EU legislative lobbying, each representing different configurations of internal and external conflict (see Table 1). In the next sections we discuss the four configurations in more detail and how they might affect the influence chances of interest groups. As these ideas are designed to produce testable hypotheses, we also consider alternative explanations in our analysis and extensively discuss these in the conclusion as potential avenues for future research.

Table 1. Set of configurations among internal and external conflict and its impact on organizational chances for influence.

	Internal conflict
	Low conflict	High conflict
External conflict
Low conflict	I. Much influence on policy process by interest groups	II. Limited influence on policy process by interest groups
High conflict	III. Medium influence on policy process by interest groups	IV. Limited influence on policy process by interest groups

The first option (low internal conflict – low external conflict) indicates a situation where there is a community of similar types of interest groups, which internally agree on a position. This refers to a situation where policymakers are confronted with interest groups all sharing a similar position, and so will see broad support among constituents, thereby making it difficult for them to ignore the interest-group demands. Moreover, if most interest groups seek a similar solution, the chances of having an impact on policy outcomes should be relatively high as there is hardly any competition for influence among the interest-group community (Mahoney, 2007; Rasmussen, 2015; Dür et al., 2015). In such instances it seems reasonable to assume the impact of lobbyists to be the highest of all four quadrants.

The second option (low internal conflict – high external conflict) reflects issues whereby NGOs and business actors stand on clear opposite sides of the conflict. While conflict may be high on this issue, the NGOs and the business communities have internally unified positions. Obviously, this situation should decrease the chances of individual lobbyists having an impact on policymaking compared to the first option, as not every position can be reflected in political decisions (Rasmussen, 2015). Yet, we still expect interest groups to have a substantial impact on policy process, as this type of conflict is interpreted by policymakers as ‘normal’ or ‘regular’. They can make a decision based on economic values (the business side) or more societal values (the NGO side), or simply find a compromise. The key here is that policymakers have an incentive to push the process further as the position of key interest-group camps is, at least, clear and united. As either one camp wins and the other loses, or there is a weighted compromise linking the two internally united camps, the overall influence in this conflict configuration should be medium, and we explore whether this is true and why.

Option three (low external conflict – high internal conflict) reflects situations where business organizations have internal conflict, and NGOs are mostly absent, or where NGOs have political conflict, and business organizations are not involved. While often overlooked in the literature (see Baumgartner & Leech, 2001), this is a very common situation. In many cases, conflict occurs not between business groups and NGOs, but among business groups or among NGOs themselves. In this situation, policymakers cannot simply make an ideological decision based on clear-cut political cleavages, and therefore feel they cannot trust the signals that interest groups provide to them. Rather than listen to interest groups, policymakers have an incentive to postpone the issue or push policies further based on their own judgement. The potential for interest groups to influence policymakers in this situation should therefore be low, despite having limited opposition from other types of organizations.

The fourth and final option (high internal and high external conflict) is the least attractive for interest groups and policymakers. This does not only signal conflict among cleavage lines, but also within perceived homogenous networks of interest groups. Policymakers will interpret such circumstances as a signal of a lack of support across the board, both among economic and societal constituents. It seems likely that policymakers confronted with both high internal and external conflict to, depending on their own preferences, either postpone any decisions or focus on their own judgement without too much interference of interest groups. As a result, we expect that the impact of lobbyists is very low for these types of cases. We explore whether this is true and, if so, why this is the case.

Data and methodology section

To understand how conflict may affect influence we compare four cases of interest representation in EU policymaking. The cases were selected on the basis of their high/low internal and external conflict dimensions – i.e., the independent variable of our study – to see how that possibly impacts interest groups’ ability to influence policy outcome as the dependent variable. Each case represents one of the quadrants of interest-group conflict presented above (see Table 2).

Table 2 . Cases per level of internal and external conflict.

	Internal conflict
	Low conflict	High conflict
External conflict
Low conflict	I. Pension funds Directive	II. Child abuse Directive
High conflict	III. Industrial Emissions Directive	IV. Data protection regulation

The case showing very low internal and very low external conflict is the Institutions for Occupational Retirement Provision (IORP) II Directive. Instead of expected conflict between (at least) pension funds, employers’ organizations and trade unions, all these major actors formed a coalition to oppose the proposed legislation. Conflict among interest groups in the Industrial Emissions Directive (IED) can be regarded as ‘normal’ with considerable conflict between NGOs and industry but low (if any) conflict within these ‘camps’. The Directive on Combating Child Sexual Abuse forms the case study that shows high (and for policymakers unexpected) internal conflict among NGOs: child protection NGOs and data protection NGOs clashed severely on the Directive's proposed approach to combat child abuse. With business organizations being mostly absent in the policymaking process of the Child Abuse Directive, this case shows low external conflict. Finally, the General Data Protection Regulation (GDPR) represents a case of high internal and high external conflict: the spectrum of interests and positions was extremely scattered, to the extent that one can neither speak of the position of ‘business’ nor that of ‘the NGOs’.

We rely on process tracing as our research method to analyze in which configuration interest groups have been more influential, and use qualitative, semi-structured interviews with key stakeholders as our primary source of data. This in-depth qualitative methodology allows for an examination of the successive stages of legislation and mutual interactions between and among policymakers and interest-group actors in each case (cf. Gerring, 2007, pp. 172–173; Beach & Pedersen, 2013, p. 1). More specifically, this paper utilizes theory-building process tracing, as it seeks to further explore what underlying mechanisms connect the × (interest-group conflict) to the Y (influence) (Beach & Pedersen, 2013, p. 16). By way of gauging (self-)perceived influence, respondents were asked whether they believe to have been successful in a particular policymaking process (cf. Dür, 2008; see Appendix for elaboration on the methodology).

We conducted 61 interviews in the four different case studies, with a minimum of twelve interviews per case. In order to obtain a representative and balanced picture of each of the lobbying processes, the interviews per case were selected in such a way that all the different ‘sides’ were represented by at least one interviewee (see Appendix 2 for selection of actors and the list of respondents).

Results

We present the four cases in sequence. In each case we first describe the overall level of conflict between and among lobby camps. Then we analyze how this affected the influence of interest groups, focusing on the perspective of the policymakers involved.

The IORP II directive (low internal/low external conflict)

In an attempt to harmonize and tighten regulation of pension funds’ investment behaviour, the European Commission proposed a revision of the Institutions for Occupational Retirement Provision (IORP) Directive in 2011. Among the main elements of the IORP II proposal were a standardized format for pension funds to report on their financial activity, the introduction of capital requirements for pension funds, and an enhanced role for the European Insurance and Occupational Pensions Authority (EIOPA) as the European supervisor reviewing pension funds’ investment behaviour. Interest groups, ranging from pension funds associations to trade unions and employers’ organizations, strongly opposed the IORP II proposal and were in the end able to fundamentally influence the IORP II Directive. Opposition was based on the argument that the measures for regulating pension funds were not flexible enough to respect the diversity of pension fund systems across the EU Member States, and that harmonizing supervision at the European level would tie pension funds’ hands in an incorrect – and even harmful – way.

The IORP II Directive represents a case of low internal conflict and (remarkably) low external conflict between interest groups. Besides strong internal coherence within the pension funds sector, a broader coalition referred to as the Group of Nine (G9) was set up explicitly for this particular Directive. This G9 coalition covered practically the entire range of relevant stakeholders in the economic community – indeed, involving interest groups that typically oppose one another.¹ The G9 agreed on a number of overarching common denominators, which allowed them to make firm statements on the main issues in the Directive. ‘We up-scaled the lobbying issue from pension fund level and technical details to the […] issue that was unifying all of these actors, regarding the investment environment’ (interview Matti Leppälä, PensionsEurope). Adding to the low external conflict character of this case, it is noteworthy that on top of the strong coalition lobbying against capital requirements and the role of EIOPA, there were hardly any active ‘dissidents’ to the G9’s view in the policymaking process. Although insurers had a different position than the G9, they were not per se in favour of the proposal either, and thus did not obstruct the G9 lobby. And as consumer organizations were inactive, there was practically no external conflict in this lobbying process.

Important in this case is that policymakers had a very clear agenda, and were rather unified on the policy solutions they proposed for monitoring and regulating investments of pension funds. Yet the extraordinary coalition of lobbying groups was powerful enough to block that unified position of policymakers. Interviews with both policymakers and members of the G9 suggest that it was in particular the unorthodox composition of the coalition, voicing a consistent and unified position throughout the process, that gave it significant political clout from the onset. ‘It was not social partners alone, it was the pension industry, and it was asset managers and private equity managers, which was quite an extraordinary alliance. All these came together in one group’, explains Claudia Menne of the European Trade Union Confederation (ETUC). Since the G9 covered nearly the entire range of stakeholders, the Commission and EIOPA had to take the arguments seriously. ‘They were saying ‘oh this bloody G9!’ Even though we were an informal group […] they could not ignore us’ (interview Matti Leppälä, PensionsEurope.).

Matti Leppälä, chair of PensionsEurope and initiator of the G9, explains how it was the primary strategy to have and maintain this unorthodox alliance of interests (cf. Nelson & Yackee, 2012). Aware that the different interests would eventually clash on the details of the proposal, the common target in the coalition was deliberately kept broad:

We were able to put it on that level and that language that we agreed. There would be many things where we wouldn't agree, but we were careful not to do that. […] Then we had a basic cooperation, and many wanted to join. (Interview Matti Leppälä)

Interviews with policymakers suggest that they were in the end not necessarily convinced by the substantial arguments that were made against the policy (e.g., interview Sandra Hack, EIOPA). Anna Kadar of DG FISMA commented in an interview: ‘of course we have higher ambitions than this. We want a standardized model for the whole of Europe which is not going to be the case now’. Although the DG was convinced that it would be good to regulate pension funds by means of capital requirements, they had to drop it because of the heavy opposition. ‘The Directive was agreed without any technical provisions and to this day [the interest-group stakeholders] are very grateful and very happy; otherwise they would have been very concerned’.

Since it were clearly not the arguments that made the difference, it seems that the political clout that came with the sheer composition of the G9 accounts for its political influence to a great extent. With low internal and extraordinarily low external conflict, interest-group influence in this case was so big as to block the core of the proposed legislation altogether, against the intentions of the policymakers at the European level.

The industrial emissions directive (low internal/high external conflict)

As a revision of the 1996 Integrated Pollution Prevention and Control Directive, the 2010 Industrial Emission Directive (IED) aims to prevent or minimize the emission of pollutants by large industrial installations. As the old Directive introduced a number of innovative governance procedures and arrangements, the main issues of the IED proposal concerned clarification, improvement and overcoming shortcomings of those governance arrangements. The Directive uses context-dependent and revisable Best Available Techniques (BATs) as the standard upon which emission levels per sector are determined, and on which basis national authorities issue permits for individual installations. BATs are defined per industrial sector (e.g., steel, chemicals, paper and pulp) and are the outcome of a cooperative process in which Member States, industry, and environmental NGOs are closely involved. The main bones of contention in the policy process leading to the IED were the legal status of the BATs in issuing permits, and the room for derogation from the BAT-based emission levels in specific cases of permitting.

With regard to the two faces of conflict, the IED case represents the rather typical configuration of low internal conflict and high external conflict. Environmental NGOs, led by the European Environmental Bureau (EEB), lobbied for binding BAT-based permits and narrow room for derogation. A varied coalition of industries, united in the IPPC Alliance of Energy Intensive Industries, called for BATs that guide but not define emission levels, and stressed that there should be significant room for derogation to allow healthy industrial operations and avoid bankruptcy due to policy (see position paper: www.eurofer.org). Policymakers, seeking to give the Directive ‘teeth’, preferred as much clarity as possible, meaning binding BAT standards, and little room for derogation for the sake of coherence and functioning of the legislation at all (interview Serge Roudier, EIPPC Bureau; interview DG Environment). In this case too, there seems to have been little conflict among policymakers on the proposed measures of the revised Directive.

Despite the external conflict between industry and NGOs, the high levels of internal coherence meant that the scope of conflict was relatively clear to policymakers and perceived as surmountable (interview with Roudier). Affirming the ‘business as usual’ lobbying situation, one policymaker said in an interview that ‘industry always complains about costs being too high’, and NGOs ask for a higher ambition level. ‘We have to respond to [these complaints]’. With its own agenda to give the Directive more teeth, the Commission found the necessary backup in the clear and uniform position of the environmental NGOs.

The outcome of the legislative process is a policy which was in the end supported by both industry and the NGO community, namely that BATs are binding in issuing permits (interview DG Environment (b)). Derogation is allowed only on the basis of strict economic and technical arguments. The Directive obliges national competent authorities to provide a transparent and publicly available assessment of an operator's justification for derogation. In addition, and also based on the BAT Conclusions, the Commission set uniform maximum emission levels that must be observed and cannot be exceeded by any plant, which echoes the EEB's wish to keep the level of derogation within an acceptable range.

In line with our expectation on this conflict configuration, the IED case shows how even in the wake of high external conflict between NGOs and industry organizations, interest groups were able to have a significant impact on the policy process, because the opposing camps were unified in their demands. A unified NGO agreeing with the intended legislation could be used to back up the policymakers’ plans to improve the Directive's clarity as well as give it ‘teeth’, even if this was against the preferences of the industry coalition. Policymakers hereby had a strong incentive to consider both demands and come up with a solution satisfying both needs. As such, the fact that both camps stayed unified throughout the process strengthened their positions considerably and ultimately helped them both to have considerable impact on the policy process.

The child abuse directive (high internal/low external conflict)

Seeking to enhance international cooperation in the battle against child sexual abuse, in 2011 the Commission adopted a Directive on Combating the Sexual Abuse and Sexual Exploitation of Children and Child Pornography. Although the Directive covered all forms of sexual abuse of children, including for instance sex tourism and child prostitution, the main focus and source of controversy concerned the internet and measures to combat child sexual abuse online. As one of the main elements, the proposal stated that Member States should employ measures of filtering and blocking websites that display child sexual abuse material (i.e., child pornography; www.interpol.int).

Interest-group discussion on this issue followed three general lines of reasoning, the overarching frame of which concerns the effectiveness of the policy. The first argument against blocking websites that display child sexual abuse material is that it merely combats the dissemination of the online problem, whereas this should be deleted at the source. Secondly, focusing on distribution of child sexual abuse material through the internet was argued to be mere symptom management while largely ignoring the real problem, namely the physical crimes that precede such material. Relatedly, a third concern about deleting this material from the web is that it destroys evidence that can be used to identify and support the victims, and to trail and prosecute the perpetrators.

The key feature of this case study is the (unexpected) internal conflict that occurred among NGOs, while there was hardly any external conflict due to only remote involvement of IT business organizations.² Child-protection NGOs, represented most vocally by the European NGO Alliance for Child Safety Online (eNACSO), favoured the blocking and filtering provisions arguing that although this may not be the ultimate solution, it is better than doing nothing (interview eNACSO). Also hotlines, whose work revolves around detecting and taking down child sexual abuse material from websites, favoured obligatory blocking and hence actively pushed for this measure in the legislation (interview Internet Watch Foundation). Fierce opposition however came from data protection NGOs. Privacy association European Digital Rights (EDRi) completely rejected the proposed measures, stating that they were superficial, ineffective, and did not take into account any notion of data protection and privacy issues (interview EDRi). The data protection concern was shared by the European Data Protection Supervisor in their comments on the draft Directive (interview EDPS). It is important to note that in this case policymakers had divergent perspectives. Especially in the EP and the Council conflict occurred, as national experiences with the proposed measures of blocking made especially Swedish, Austrian and German policymakers hesitant about the Directive.

As suggested by EDRi’s Joe McNamee, the Commission from the outset did not expect much discussion on the proposed measures, and above all did not consider the blocking and filtering measures to possibly conflict with data protection and privacy regulation:

The work that the Commission had done on the proposal was very very very bad. They didn't do their homework; they didn't understand the issue. […] [They felt] no need to do their homework, because they thought there would be no opposition from the Member States or from civil society! (Interview with EDRi, Spring 2013)

Reinforcing this statement, the Commission stated that ‘[t]here was no need for external expertise’ for the impact assessment (2010/0064 (COD): 5). Presuming it would find unanimous support, the Commission had not expected much conflict at all – let alone the internal conflict among NGOs.

The unexpected internal conflict among groups that policymakers perceived to be on the same side posed severe problems for pursuing this legislation. The interest-group debate about the effectiveness and side effects of blocking and filtering eventually led Members of the European Parliament to demand amendment of the wording of these measures. Whereas the initial proposal intended to make blocking and filtering mandatory, it ended up being a voluntary and optional measure in the 2011 Directive (EC, 2011, p. 6). Additionally, the issue of child sexual abuse material as evidence for tracking the criminals was adopted in the Directive, which now states that these materials can and should be used in investigative tools (Ibid., p. 10). Even though the Commission had a very clear and uniform position from the outset around the legislative process, the high internal conflict among NGOs made them uncertain with regard to the main provision of blocking and filtering. Their uncertainty in turn translated into a Directive that is much vaguer than the initial proposal.

With respect to the ability to influence policy outcome under the condition of high internal and low external conflict, it is important that none of the active interest groups consider themselves (or their ‘side’) to have won the argument. eNACSO felt ‘out-lobbied’ in the amendment making mandatory blocking and filtering by Member States voluntary, whereas EDRi claimed to have lost this issue because this is still the main measure to combat the problem. Indeed, the unexpected schism among NGOs made it difficult for policymakers to find the needed (and expected) support for the proposed legislation. In turn, despite the relatively small scope of active interest groups, and despite the absence of external conflict with business organizations, no one coalition was able to dominantly influence the policy and core issues remain basically unresolved.

In part as a result of the weak compromise, the Commission and other EU institutions have put little effort into its implementation and enforcement, making the Directive practically a dead letter (Truijens, 2021). The weak Directive and lack of enforcement suggest that policymakers simply postponed making a decision.

The general data protection regulation (high internal/high external conflict)

Typified as ‘one of the most heavily lobbied bills in the history of the European Parliament’ (EU Observer, 14 April 2016), the General Data Protection Regulation (GDPR) represents a case of high external and (extraordinarily) high internal conflict. The broad scope of this general data protection regulation, the introduction of a large number of new concepts by which policymakers sought to answer to rapid developments in IT, and repeated media salience³ of data protection and privacy issues all contributed to an exceptionally scattered interest-group landscape. The internal and external conflict was so fierce that it is nearly impossible to identify any form of unified position or coalition of interests (interview Simon Davies). In the GDPR process, there has also been significant and open conflict among policymakers. Although this did not in the end result in failure to propose the Regulation, the differences in priorities, preferences, and preferred measures also occurred among policymakers (with the rapporteur being a very outspoken advocate of privacy).

Although the issues that were at stake are too numerous to discuss here, three general lines of debate can be identified. First, there were issues related to the implementation and enforcement mechanisms of data protection rules. Besides the roles and competences of the national data protection authorities (DPAs) and their interaction at EU level in the Article 29 Working Party (WP29), this concerned enforcement mechanisms. A big concern for representatives of business in particular was the level of the fines for data breach. Fines of four per cent of a company's annual global turnover – proposed by rapporteur Albrecht – were argued to be grossly disproportionate (interview with ETNO; interview with EuroISPA). The second line of discussion revolved around definitions of various (new) key concepts in the policy field, such as big data, profiling and explicit consent (interview FEDMA). The final line of debate concerned a more general issue of the balance between proper data protection and room for manoeuvre, innovation, and trade of businesses.

Acknowledging the scattered spectrum of interests and preferences across this extensive scope of issues, European Parliament rapporteur Jan Albrecht proclaimed it as his ambition to make ‘everybody equally unhappy’ (see davelevy.info). Among the main elements that were in the end adopted are the introduction of the General Data Protection Board (replacing WP29 and enhancing cooperation between national DPAs), obligatory requesting permission to use personal data (‘explicit consent’), and, in spite of the strong anti-lobby, the possibility of sanctioning data breach of companies with a fine of four per cent of their annual global turnover. With this, the GDPR lives up to Albrecht's initial ambition, as NGOs are disappointed with the finally adopted Regulation because it is not clear and strict enough, while business is disappointed because they regard the rules as unclear and at the same time too strict for their activities and development (Truijens, 2021). This suggests, in line with the hypothesis, that in a situation with high internal and external conflict, interest groups have very low chances of influencing policy outcomes.

As to the policymakers’ side, the lack of any form of reconciliation between the interest groups in the GDPR process allowed rapporteur Albrecht to pursue his preference because interest groups did not manage to form effective coalitions to either support or block it:

We met hundreds of them. The main reason was to see what they'd tell the other groups. We already had our goals clear […] because we knew our shit already, you know, so it was important for us to see what the discussion was around it, and what industry was telling [shadow rapporteur] Axel Voss for instance. That was the main reason for doing these meetings. (interview with Ralf Bendrath, assistant of rapporteur Albrecht)

This remark highlights how policymakers can benefit from this conflict configuration in pursuing their own agenda. For the rapporteur, this scattered situation allowed him to largely stick with his initial position. Bendrath explains:

Of course it was an interesting insight into all the different varieties of the different industry sectors. […] But we still tried to narrow it down to where it is really necessary, and not open it up for everything.

As Jacob Kohnstamm, former chair of the WP29, stressed in an interview, policymakers tend to avoid doing business with individual organizations as that would compromise their impartiality. The ‘clear blue water’ between the different interests makes it impossible for a small number of interest groups to represent a broad, sector-wide position. With each company thus practically defending its own individual position, policymakers were not inclined to seriously adapt their position in the legislation (cf. interviews Bendrath; Kohnstamm; Davies). In other words, the combination of high internal and external conflict made it difficult for policymakers to identify positions of seemingly coherent groups in society.

In conclusion, this case shows how the configuration of high internal and external conflict significantly decreases interest groups’ ability to influence the policy outcome. Although the conflict among policymakers may have certainly contributed to the outcome, the case at least suggests that amidst the high internal and high external conflict between interests group, no single (set of) interest-group actors has been able to steer the GDPR towards their preference.

In sum, across the four cases we find that influence varies a lot and that an important reason for this variation relates to the internal and external conflict dimension (see Table 3). More specifically, internal conflict seems to limit influence chances of interest groups more than external conflict. We discuss the implications of this finding in the conclusion.

Table 3 . The influence per case and type of configurations.

	Internal conflict
	Low conflict	High conflict
External conflict
Low conflict	I. Pension funds Directive = much influence	II. Child abuse Directive = limited influence
High conflict	III. Industrial Emissions Directive = medium influence	IV. Data protection regulation = limited influence

Conclusion

Divide et impera – as the Romans knew all the well. But be sure to understand along which lines you divide to rule. For interest groups to succeed in the political arena it is critical they convey a united position (Nelson & Yackee, 2012; Rasmussen, 2015; Junk & Rasmussen, 2019). Unity seems especially important for actors which are perceived by policymakers as having rather similar interests and policy positions. Conflict among groups which are already perceived as being on the opposite sides seems to impact chances of lobbying influence much less. While this statement is perhaps the main take away, our paper also displayed the actual decision-making procedures of policymakers, providing insight into why unity among similar types of interest groups is an important variable for future research to consider. Our findings suggest that conflict is indeed an important heuristic for policymakers and that, consequently, their decisions are influenced by the level of conflict among interest groups. Yet, because conflict is a heuristic, the type of actors that are in conflict with each other matters. Conflict between business groups and NGOs is seen as ‘business as usual’, and will not necessarily hinder lobby efforts. If, however, NGOs fight among each other, or business groups are in (open) conflict, this confuses policymakers and they are more inclined to ignore the demands of the interest groups and instead follow their own course.

Given the limited number of cases we could study, we refrain from taking a too strong position on these matters; by way of conclusion we rather list several key observations from our case studies, which deserve further exploration. First, our case studies provide an important note for the literature on interest-group influence: scholars in the field should be more attentive to the nature of conflict in terms of the unity of lobby camps (see also Junk & Rasmussen, 2019; Chalmers, 2020). To put it simply, conflict matters, but some conflicts matter more than others. We therefore hope that future studies will test our findings in a quantitative setting to find out whether they apply beyond our cases as well.

Second, our limiting of ‘lobby camps’ to NGOs versus business organizations, limits the generalizability to other types of lobby camps. For example, there are cases in which lobby camps consist of NGOs and business groups (Beyers & De Bruycker, 2018 ; Junk, 2019). It is unclear whether the same mechanisms we identified in this paper apply to these types of camps as well. It seems that such coalitions emerging from heterogeneous camps are a high-risk, high-reward type of coalition. Unity among these contradictive voices should benefit interest-group influence, as policymakers should perceive such coalitions as a strong indication of societal support. Yet, if cracks emerge in such coalitions, it may quickly turn into the worst type of situation, namely conflict within a lobby camp. We hope future research will analyze conflict among different types of lobby camps to complement the findings we presented in this paper.

Third, future research should explore further variation across the key variables we introduced in our paper. For starters, we have limited variation in types of policymakers that is targeted by interest-group camps. It may very well be that elected officials are more vulnerable to unity among camps than civil servants, as the former have a clear incentive to avoid public scrutiny and opposition. In that same vein, the nature of internal organizational decision-making has not been studied in this paper. As Lowery and Marchetti (2012) argue, there is much conflict within organizations about the specific policy positions as well. Once such in-house conflict becomes known to the public, the consequences should be detrimental for a group to have an impact on decision-making. These are just a few examples of how different types of conflict may affect the influence groups have, and are interesting areas of research in the future.

A final pathway for future research relates to our conception of conflict as an exogenous factor. It may very well be that conflict is (also) strategically used by interest groups. While we did not consider this option in our cases, it would make sense to create an image of conflict among opposing lobby camps if interest groups aim to maintain the status quo. What comes to mind here are the many firm-funded NGOs. If these organizations convey radically different ideas than the other NGOs, this creates the image that the NGO community is divided, and this may seriously harm their potential influence. Examples include the smoking lobby during the 1970s and 1980s and the environmental debate in current times. In both instances, NGOs sponsored by the corporate world (have) created an image of uncertainty about the claims made by NGOs and the scientific community. This has certainly influenced the perception of policymakers, either directly or through electoral backlash. We hope future research will explore this as well. At the very least, policymakers should be aware of such strategies and filter out whether camps are truly divided or fabricated for other purposes. Otherwise, it may distort them from making the best decision.

Acknowledgements

We would like to thank two anonymous reviewers for the their excellent and constructive comments on earlier versions of the manuscript. We also thank the Amsterdam Centre for European Studies (ACES) for financial assistance related to this project. Finally, we thank Jonathan Zeitlin for his extensive support for this project.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Douwe Truijens

Douwe Truijens works as a Scientist Innovator at the Netherlands organisation for applied scientific research. Douwe's PhD project ‘New Modes of Lobbying for New Modes of Governance’ (defended in 2019) addressed the question how interest groups respond to new and developing policy-making circumstances in the European Union.

Marcel Hanegraaff

Marcel Hanegraaff is an Associate Professor in Political Science at the University of Amsterdam. His research focusses on the politics of interest representation in a transnational and EU context. He currently works on a large scale project focusing on the agenda setting power of interest group in the EU.

Notes

1 See https://www.pensionseurope.eu/group-nine for the list of members of the G9.

2 IT businesses including Facebook, Google, Twitter and T-Mobile are organized in the ICT Coalition for Children Online. Though concerned with the issue, the Coalition was not actively involved in the Child Abuse Directive (interview ICT Coalition).

3 The exceptional salience of data protection in itself may be an important factor that determined the extent to which particular interests were (un-)able to influence policy outcome. Dür and Matteo (2016) for instance suggest that policymakers may be less inclined to make more business-friendly policy decisions when the policy issue is very salient, as this is regarded a less popular policy outcome. Though we are aware of omitting this mechanism in this paper, it seems that amidst the high internal and external conflict, NGOs typically representing the more public interests have not been able to profit from this salience in steering the GDPR towards their preferences.

Appendices

Appendix 1. Why process tracing as a method to study influence

Lobbying influence is not one single concept, and neither is measuring influence. Authors like Klüver (2013) and Dür et al. (2015) study the extent to which lobbying preferences are attained in eventual legislation, in which they quantitatively analyse the extent to which wording of the legislative text has changed in line with the wording of a particular lobbying actor or coalition. In this explorative paper, we take a more generic concept of influence, defined as being able to steer (parts of) the policy towards one’s preference. This means that influence can also occur in the form of not having a proposed legislation changes, despite pressure from other actors. We gauge such influence on the basis of reflection of the involved actors: do they perceive themselves as being successful in a particular policymaking process? Focussing on (self-)perceived influence also helps accommodating the role of chance and contingency in seeing favourable outcomes realized – which the preference attainment method may overlook. Moreover, influence may occur in way beyond word-alignment, for instance when a policy measure is phrased differently, becomes less strict or is deleted. Such self-perception information is obviously prone to subjectivity and has a relatively low reliability. However, the number of interviews conducted per case study as well as the documented change in legislation allows for triangulating such statements.

We rely on process tracing as our research method to analyse which groups have been more influential than others, with qualitative, semi-structured interviews with key stakeholders as our primary source of data. We opt for this method, in addition to (self-)perceived influence, as we want to analyse the causal relations between multifaceted concepts (Dür, 2008, see Appendix 1 for more information on this choice). Moreover, going beyond connecting a cause to an effect, we seek to explore how the concept of internal and external interest-group conflict is related to influence on policy outcome. We consider process tracing most suitable for investigating empirically how different actors behave and contribute to policy outcomes in an already complex governance environment. This in-depth qualitative methodology allows for an examination of the successive stages of legislation and mutual interactions between and among policymakers and interest-group actors in each case (cf. Gerring, 2007, pp. 172–173; Beach & Pedersen, 2013, p. 1). More specifically, this paper utilizes theory-building process tracing, as it seeks to further explore what underlying mechanisms connect the X (interest-group conflict) to the Y (influence) (Beach & Pedersen, 2013, p. 16).

Appendix 2. List of interviewees per case study

The selection of interest groups that were active in each case, was mapped out on three main bases. First, reports of official consultations organized by the Commission showed which groups attended and participated. As not all interest groups seek to approach the Commission in such a direct way, this official consultation was only the starting point of the mapping exercise. The second step was to visit websites of the groups listed in the consultation report, and review their position papers on the issue. This not only helped in mapping out an overview of positions, preferences, and lobbying objectives – categorizing different interest groups by ‘side’ – but also provided information about possible coalitions or collaborations with other organizations. As such, also groups that did not engage in the official consultation of the Commission but that were active on the topic could be identified. The third step was to ask the interviewees who else was involved on the topic – be it ‘allies’ or ‘opponents’. In order to obtain a representative and balanced picture of each of the lobbying processes, the interviews per case were selected in such a way that all the different ‘sides’ were represented by at least one interviewee. In order to investigate our preliminary hypothesis empirically, we conducted 61 interviews in the four different case studies, with a minimum of twelve interviews per case.

Below, one can find the list of respondents per case. Note that at the time of writing, some of the interviewees may not any longer have the position in which they were interviewed.

Industrial emissions directive

Table

Organization	Abbreviation	Name	Position
Non-state actors
The European Chemical Industry Council	Cefic	Ann Dierckx	Policy adviser environment
Institute for European Environmental Policy	IEEP	Patrick ten Brink	Director of Brussels Office
The Union of the Electricity Industry	EURELECTRIC	Hélène LAVRAY	Advisor Renewables & Environment
The European Steel Association	Eurofer	Jean-Pierre Debruxelles	Environment director
The European Steel Association	Eurofer	Danny Croon	Environment director
Confederation of European Paper Industries	CEPI	Jori Ringman	Deputy Director General
European Environmental Bureau	EEB	Christian Schaible	Policy adviser
Public actors
DG ENVIRONMENT		NN	Unit Air & Industrial Emissions
DG ENVIRONMENT		Filip Francois	Former team leader IED
IPPC Bureau		Serge Roudier	BREF writer and director
Environment Agency (UK)	EA	Neil Emmott	Permit writer

General data protection regulation

Table

Organization	Abbreviation	Name	Position
Non-state actors
Bits of Freedom	BoF	Floris Kreiken	Researcher
European Digital Rights	EDRi	Joe McNamee
AccessNow		Estelle Massé	Policy Analyst (Brussels office)
European Privacy Assocation	EPA	Paolo Balboni	Founder / individual lawyer
Privacy International	PI	Simon Davies	Founder / individual advocate
Bitkom		NN	Policy adviser
European association of European Internet Services Providers Associations	EuroISPA	Andrea D’Incecco	Head of Policy
European Telecommunications Network Operators' Association	ETNO	NN	Public and Regulatory Affairs Officer
The European Consumer Association	BEUC	Agustín Reyna	Senior Policy Officer / Senior Legal Officer
Federation of European Direct and Interactive Marketing	FEDMA	Mathilde Fiquet	EU Legal Affairs Manager
Zentralverband des deutschen Werbewirtschaft	ZAW	Katja Heintschel von Heinegg	Policy adviser European affairs
Science Europe		NN	Head of Policy Affairs
BUSINESSEUROPE		NN	Adviser Internal Market
American Chamber of Commerce to the European Union	AmChamEU	NN	Policy adviser
Public actors
DG Justice and Consumers		Bruno Gencarelli	Head of the Data Protection Unit
Article 29 Working Party		Florence Raynal	Rep. of Chair Isabelle Falque-Pierrotin
Article 29 Working Party		Jacob Kohnstamm	Former Chair
European Data Protection Supervisor	EDPS	Christian D'Cunha	Head of Private Office of Director Giovanni Buttarelli
MEP Greens / Rapporteur		Ralf Bendrath	Assistant to rapporteur Jan Albrecht
MEP ALDE		Sophie in 't Veld
MEP EPP / Shadow Rapporteur		Axel Voss	Shadow rapporteur GDPR
European Commission		Viviane Reding
DG JUSTICE		Paul Nemitz	Principal adviser

Child abuse directive

Table

Organization	Abbreviation	Name	Position
Non-state actors
European NGO Alliance for Child Safety Online	eNACSO	John Carr	Adviser online safety
European Digital Rights	EDRi	Joe McNamee	Director
European association of European Internet Services Providers Associations	EuroISPA	Andrea D'Incecco	Head of Policy
Internet Service Providers Association Ireland	ISPAI	Paul Durrant	Chief Executive
ICT Coalition		Andrea Parola	Chair
INHOPE		Arda Gerkens	Chair
Internet Watch Foundation	IWF	Sarah Smith	Technical Researcher
Public actors
Europol		Jan Ellermann	Senior specialist
Interpol		Michael Moran	Assistant director Human Trafficking
Eurojust		Diana Alonso Blas	Data Protection Officer
DG HOME		Jakub Boratyński	Head of Unit on Crimes against Children
DG HOME		César Alonso Iriarte	Principal Policy Officer

Institutions for occupational retirement provision directive

Table

Organization	Abbreviation	Name	Position
Non-state actors
German Occupational Pension Association	aba	Klaus Stiefermann	Director General
Pensions and Lifetime Savings Association	PLSA	James Walsh	Chairman
Pensioenbeheer	PGGM	Nine de Graaf	Coordinator IORP II
De Pensioenfederatie		Sibylle Reichert	Head of Brussels bureau
European Trade Union Confederation	ETUC	Claudia Menne	Confederal Secretary
PensionsEurope		Pekka Eskola	Policy Adviser
BUSINESSEUROPE		Rebekah Smith	Chair Pension Working Group
Insurance Europe		NN	Policy Advisor, Pensions
		Fieke van der Lecq	Professor of pension markets at VU University
Public actors
DG FISMA		Anna Kadar	Head of Unit Pensions and Insurance
EIOPA		Sandra Hack	Head Pensions Unit
OPSG / CEO PensionsEurope		Matti Leppälä	Chair OPSG (since April 2016)