Influencing public opinion : how interest groups dertermine the right strategy

INFLUENCING PUBLIC

OPINION

HOW INTEREST GROUPS DETERMINE THE RIGHT

STRATEGY

Master Thesis Political Science

Foreign Politics

Advocacy in EU and World Politics: who is

represented (and who is not)?

Mike Posdijk

10196609

29 January 2016

University of Amsterdam

Faculty of Social- and Behavioural Sciences

First reader and mentor: dhr. dr. M.C.

Hanegraaff

1. SUMMARY

Many researchers within the field of political science and interest groups in particular have

conducted studies on available strategies interest groups employ in order to achieve their goals (in most cases influencing policy outcomes). A common distinction is made between inside- and outside strategies. Most literature in this specific field, focusses on factors that explain the choice between inside- or outside lobbying. Specialized interest groups would most often choose the inside lobbying strategy, while public interest groups would mostly rely on outside approaches. Although all this research provides insight into strategy choice at the highest level, much less is known when things are taken a step further. Especially on outside lobbying, the literature seems to be lacking. Whereas several scholars acknowledge the importance of public opinion as part of the outside lobbying approach, few of them actually pay attention to the ways interest groups might be able to influence public attitudes. This thesis tries to fill this gap by examining the ways interest groups might be able to influence public opinion and more importantly, which factors they take into account when determining their final strategy. After conducting twenty, semi-structured interviews with lobbyists working for different interest groups, the argument is made that interest groups take three different kind of factors into account when constructing their strategy to influence public opinion. Firstly, group type seems to affect strategy choice in three different ways. Reputation seems to “force” producer interest groups to focus primarily on indirect approaches. Since indirect approaches influence the public indirectly, “through” third parties, producer interest groups choose them to provide credibility to their statements. Moreover, being reliant on members also seems to have implications for strategy choice as these members want visible proof that interest groups are spending their funds wisely. This causes interest groups to choose the mass media, as this allows them to reach their constituents easily. Furthermore, sector-wide rules and regulations restrict interest groups to certain strategies as they are for example not allowed to advertise on their products. Secondly, issue characteristics seem to influence strategy choice in multiple ways. Issue salience not only determines strategy choice but also interest group activity. Highly salient issues make especially the mass media an attractive tool since saliency makes it easier to get a message published. This window of opportunity also causes interest groups to become more active as

reaching the mass media increases chances to achieve policy goals. Moreover, when issues are highly technical, interest groups are often forced to make use of more elaborate media as especially social media only allow for short messages. Lastly, resource endowment seems to impact strategy choice to a certain extend. Resource-poor interest group have to be more creative in order to get their

message across. Indirect approaches, which appear to be cheaper therefore seem to be the better choice.

2. CONTENTS

1. Summary ... 1

2. Contents ... 2

3. Introduction ... 3

4. Theoretical Framework ... 6

4.1 Defining interest groups and public opinion ... 6

4.2 Why focus on public opinion strategies? ... 7

4.3 What moves public opinion? ... 9

4.4 What role can interest groups play? ... 12

4.4.1 Direct influence ... 13

4.4.2 Indirect influence... 14

4.5 Theoretical model ... 17

5. Methodology ... 21

5.1 Approach ... 21

5.2 Validity and reliability ... 23

6. What explains strategy choice? ... 26

6.1 The influence of group type ... 26

6.2 Issue as a factor ... 29

6.3 Resources ... 32

7. Conclusion ... 35

3. INTRODUCTION

On September 18 2015, the American Environmental Protection Agency (EPA) published a shocking report. In this report, Volkswagen was accused by the EPA and the California Air Resources Board (CARB) for transgressing the strict terms of the Clean Air Act (Anon., 2015). The Volkswagen diesel engines used in models such as the Jetta, Golf, Beetle and Passat which were built between 2009 and 2015 exceeded the norm. Most shocking was the accusation that these engines would contain software which was capable to mislead test results by restricting performance and emissions while being tested. This entails that Volkswagen was deliberately misleading American authorities. Consequently, this scandal received a lot of media attention igniting an intense discussion about the potential impact of the so-called ‘mighty car-lobby’. According to some media, especially the German and French car-lobby would be well known for its major impact on European policymakers, leading to more favorable climate policy (Anon., 2015; RDW, 2015). This “renewed” attention for lobbyism not only led to questions about the desirability of lobbyism in democratic policymaking, but also about the strategies interest groups employ to have such an influence on policymakers (van Leeuwen, 2015).

Many researchers within the field of political science and interest groups in particular have conducted studies on available strategies interest groups employ in order to achieve their goals (in most cases influencing policy outcomes). A common distinction is made between inside- and outside strategies. The former strategy usually contains a direct approach, the direct contact between lobbyists and policymakers. Groups utilizing the outside strategy on the other hand, focus on mobilizing citizens outside the policymaking community to contact or pressure public officials within the policymaking community (Kollman, 1998: 3). Most literature in this specific field, focusses on factors that explain the choice between inside- or outside lobbying. Specialized interest groups would most often choose the inside lobbying strategy, while public interest groups would mostly rely on outside approaches. This choice would depend on organizational characteristics (Binderkrantz, 2008; Dür & Mateo, 2013; Hanegraaff, et al., sd; Dür & Mateo, 2014; Schlozman & Tierney, 1986), issue-specific characteristics (Beyers, 2008; Dür and Mateo, 2013) and institutional context (Eising, 2007).

Although all this research provides insight into strategy choice at the highest level, much less is known when things are taken a step further. Especially on outside lobbying, the literature seems to be lacking. Several scholars for example, acknowledge the importance of public opinion as part of the outside lobbying approach (Kollman, 1998; Rasmussen, et al., 2014; Davison, 2015; Dür & Mateo, 2014; Dür, 2015), but only few of them actually pay attention to the ways interest groups might actually be able to influence public attitudes. Kollman for instance, states that interest groups focusing on outside strategies, need to be backed by a large amount of citizen support in order to be

able to influence policymakers. One way to increase citizen support is by affecting public opinion (Kollman, 1998: 61-77). A logical follow up question would be: in what ways are interest groups able to influence public opinion? Like other scholars, Kollman describes public opinion as part of the outside lobbying approach, but again remains at a fairly abstract level and does not address this follow up question. Therefore this thesis tries to fill this gap by examining how interest groups are able to affect public opinion and more importantly which factors they take into account when determining the final strategy. This leads to the following research question: What explains the variation in strategy choices, interest groups employ to influence public opinion?

By answering this research question this thesis might not only contribute to existing outside lobbying literature by examining the factors interest groups take into account when forming a public opinion strategy. Moreover, this thesis argues that the process of influencing public opinion might be more complicated than existing literature assumes. Until now, most literature assumes that

especially group type determines whether an interest group tries to mobilize the public in order to increase chances to achieve policy goals or not. Public interest groups would predominantly engage in outside lobbying while business interest groups would focus mainly on inside lobbying approaches (Schlozman & Tierney 1986: 431–432; Maloney, et al., 1994; Binderkrantz 2008; Berkhout, 2010, 115; Dür & Mateo, 2013: 663). The reason that business interest groups would be somewhat reticent towards applying outside strategies would partially be due to the fear of reputational damage, which might lead to a loss of customers (Dür & Mateo, 2014, p. 1204). Indeed, public interest groups might have an easier time mobilizing the public as their interests mostly benefit the majority of society and therefore are not that sensitive for reputational damage. However, Influencing public opinion might not necessarily mean “going public”, leaving room for interest groups sensitive to reputational damage. The process of influencing public opinion might therefore be a little more complex than expected. To take this factor into account, a distinction is made between direct- and indirect ways to influence public opinion. This distinction helps to understand that interest groups might not always have to go public when trying to benefit from steering public attitudes. Interest groups fearing reputational damage might be able to steer public opinion by using third parties without being accused of doing so (indirect ways). This way, businesses may still be able to increase public support without risking to damage reputation.

This thesis is divided into several parts. The first part conceptualizes the important concepts of ‘interest groups’ and ‘public opinion’. It also explores the relevance of public opinion for both interest groups and policymakers. The second part examines how public opinion might shift and what interest groups can actually do to steer public opinion in the desired direction. In the third part a theoretical model will be presented, which describes how several factors might determine which

strategy interest groups choose. The fourth part provides a deep exploration of the applied methodology. Subsequently, the research-findings will be presented followed by a conclusion.

4. THEORETICAL FRAMEWORK

4.1 Defining interest groups and public opinion

There is a lot of debate in interest group literature about how to define interest groups and the manners in which they differ from each other (Baroni, et al., 2014). As Baroni et al. point out, scholars have attempted to define interest groups in many different ways. We can identify a rough distinction between a behavioral definition and one based on organizational characteristics. Whereas the behavioral approach defines interest groups on the basis of their policy-related activities, the alternative sees organizational characteristics as key defining features, reserving the interest group term for membership-based organizations (Baroni, et al., 2014, p. 5 – 6). Although these options might work for some scholars, this thesis relies on an alternative way to define the concept for two reasons. Firstly, I believe that defining interest groups in a behavioral way is not strict enough, leaving too much room for debate. As several scholars rightly argue, how much political activity is needed before an organization may be regarded as an interest group? (Wilson, 1990, p. 7; Baroni et al., 2014, p. 5). On the other hand, the alternative organizational approach, mostly rests on the so-called voluntary stereo-type (Baroni, et al., 2014, p. 5 – 6). For this thesis however, both voluntary and non-voluntary based organizations should be taken into account as corporations for instance are also likely to influence public opinion to achieve policy goals. Therefore an alternative definition, based on the nature of interests being represented, fits this thesis best. As Beyers (2004) describes, interest groups can be divided into two group types. On the one hand the diffuse- so called public interest groups which lack a well-delineated and in socio-economic terms concentrated constituency. They promote diffuse or public interests that are linked to broad or general segments of society (Beyers, 2004, p. 216). Producer interest groups on the other hand, have very specific interests. These types of interest groups have a very clear-cut stake in the production process and defend the interests of well circumscribed and concentrated constituencies (Beyers, 2004, p. 216). These

interest groups are not focusing on defending the common good, but are mainly defending their own interests or the interests of a very specific group. Although Beyers does not explicitly mention businesses in his definition, I will include them in the producer interest group category. Businesses might not defend the interests of certain members, but tend to defend their own interests and have a clear-cut stake in the production process. To keep things understandable, I will use the distinction between public interest groups and producer interest groups throughout this thesis. However, it is important to note that the producer interest group category exists of both businesses and interest groups representing businesses.

The second concept that needs further attention is public opinion. Although public opinion as a concept is hard to grasp, Kollman’s definition puts the concept directly into the right context (Kollman, 1998). In his book on outside lobbying, he constructs the concept of public opinion around the relationship between constituents, interest groups and policymakers. The relationship between constituents and elected officials is mostly one of communication, intentionally or not. Voting, answering opinion polls and writing letters are a few different ways to communicate something valuable about constituents to policymakers. Most often, the constituency is divided around certain policy and therefore sends conflicting messages to policymakers. Although public opinion within this constituency is quite complicated and diverse, policymakers essentially only want to know two elements: popularity and salience. The former refers to the proportion of constituents that supports a given policy. Salience on the other hand, refers to the perception of importance constituents have of a certain policy (Kollman, 1998, pp. 24 - 25). The distinction between both elements is important for two reasons. Firstly, there is an enormous difference in the difficulty of collecting information about these two elements. Public preferences for instance, are relatively easy to discover by interpreting annual poll results. Relative issue importance is much harder to discover, partially because this is most often not integrated in opinion polls. Policymakers therefore rely on interest groups to inform themselves. Interest groups are likely to have more information, as most interest groups are specialized in a certain policy area. Secondly, issue salience changes much more frequently than public preferences do. Salience is assumed to change with news coverage,

presidential speeches, world events and the activities of interest groups (Kollman, 1998, pp. 24 - 25). Public opinion, thus, should be seen as the amalgamation of both public preferences and issue salience.

4.2 Why focus on public opinion strategies?

The reason why public interest groups attach great importance to public opinion is because public attitudes frequently affect public policy. In fact, public opinion is of substantial policy importance for at least a third of the time (Burstein, 2003, p. 36). According to interest group- and democracy literature, public opinion might affect policymakers and policy making in two different ways. The first, most direct way, is sometimes referred to as “the theory of democratic representation” (Agnone, 2007, p. 1595). The idea is that policymakers govern in accordance to public opinion because ignoring the majority of their constituents might jeopardize their re-election (Agnone, 2007; Stimson, et al. 1995). Secondly, highly salient issues are more likely to attract third-parties into the conflict, increasing chances to achieve policy goals (Dür & Mateo, 2014, pp. 1204 – 1205; Rasmussen, et al., 2014).

The first concept, democratic representation, was elaborated by Stimson who found that in the United States, a shift in public opinion presumably leads to shifting policy outcomes through the process of dynamic representation (Stimson, et al., 1995). This process might occur in two different ways. To begin with, through the replacement of legislators in the following elections if voter demands are not met. Furthermore, through the process of “rational anticipation”. In this process, legislators choose to support policies supported by most voters, in order to secure their chances of reelection (Stimson, et al., 1995). Although the process of rational anticipation sounds reasonable, Stimson does not explicitly include the role of interest groups into his reasoning.

The “outside lobbying model”, introduced by Kollman, fills this gap. He describes the relationship between public opinion and policymakers in a similar way, but incorporates interest groups as a central element. Through “the signaling function of outside lobbying” interest groups communicate information (signals) about the salience of issues to policymakers. Public opinion should be seen as an essential instrument for interest groups in order to influence policymakers. The foundation of the model lays in the assumption that policymakers want to be reelected. To increase chances for reelection they have to be careful with their policy choices. As described before, it is not very difficult for policymakers to determine which side of the policy conflict can count on the most supporters, since most opinion polls provide sufficient information. Issue salience on the other hand, is usually not included in opinion polls and therefore much harder to determine. If policymakers want to make the right decision in choosing the “correct” side of a policy conflict, being informed about which issues really matter to constituents is of critical importance. This is what makes policymakers so dependent on interest groups. These interest groups can be divided into two types. The “high type” represents constituents who care enough about the issue to directly threat the reelection of policymakers who disagree with them. The “low type” on the other hand, does not form such a direct threat since their constituents are much less concerned with this specific issue and therefore much less likely to take action. Implicitly this model assumes that policymakers only want to cooperate with interest groups of the “high type”. This makes public opinion extremely important for both policymakers and interest groups.

The second way in which public opinion is important for interest groups to achieve their policy goals stems from the lobby success literature. Again, the salience element of public opinion seems to be important. As Rasmussen points out, interest groups tend to mobilize around highly salient issues (Rasmussen, et al., 2014, p. 260). Following this tendency, Dür and Mateo (2014, p. 1204) argue that an increasing mobilization of third parties into the conflict, might increase chances of lobbying success. If interest groups (especially citizen groups) are capable of successfully

mobilizing public opinion, they might stimulate other groups to join the bandwagon which strengthens their position. From here a positive feedback effect can develop. Increased public

salience attracts additional groups to join the campaign. As a result more lobbying makes the issue again more salient (Dür & Mateo, 2014, p. 1204). For interest groups on the less popular side of the debate, this goes the other way around. With a larger amount of interest groups lobbying on one side, it is expected for interest groups on the other side to back down as chances for success decrease. For these interest groups on the (now) weaker side, it might be better not to have their own name associated with a highly unpopular stance (Dür & Mateo, 2014, p. 1204).

After clarifying the relevance of public opinion to interest groups and policymaking, it is essential to consider how public opinion might change and more importantly, what options interest groups have to direct public opinion in a beneficial way.

4.3 What moves public opinion?

As described in the previous chapter, public opinion is highly relevant to both interest groups and policymakers. In order to understand what interest groups can do to steer public opinion in a desired direction, it is essential to understand what factors affect public opinion. According to different scholars, fluctuations in public opinion would be caused by environmental factors, mass media, opinion leaders and events (Davison, 2015; Page, et al., 1987).

As Davison points out, environmental factors would play a critical role in the formation and development of public opinion. Especially social environments including: family, school, peer groups and religious communities would have a significant impact on the initial formation of public opinion. People usually adjust their attitudes trying to conform with those that are most prevalent within the social groups they belong to. The average American individual for instance, who was born in a pro-conservative environment will most likely tend to have more sympathy for conservatism and probably vote for conservative candidates (Davison, 2015). Another example is the critical role of religion in foreign policy opinions. Baumgartner et al. for example state that American evangelicals are more likely to approve Bush’s decision making during the Iraq war, the war on terror and overall foreign policy. This environmental impact on individual values might also occur when environmental factors suddenly change. For example when an American liberal would move to a more conservative neighborhood, an increase of conservative values might occur, potentially leading to a conservative vote (Davison, 2015). It is also important to note that most environmental influences are very case specific. Gelpi et al. for instance show that more war casualties shift public opinion towards policy concerns (Gelpi , et al., 2009). Furthermore, energy prices would have a significant negative impact on public acceptance of alternative energy sources over conventional, carbon-based sources (Bolsen & Cook, 2008). Lastly, Kahn and Kotchen (2010) indicate that the business cycle influences levels of environmental concern. Higher numbers of unemployment and lower levels of income both seem to

have a negative impact on measures of public concern regarding the environment (Kahn & Kotchen, 2010).

A second, highly influential factor, would be the mass media. Especially the agenda-setting theory, extensively researched by Maxwell Mccombs has received a lot of attention. Building on the idea of Walter Lippmann that the world that we have to deal with politically is just too far away for the ordinary citizen (Lippmann, 1922, pp. 3 - 10), Mccombs states that citizens have to deal with a second-hand reality. One that is structured by journalists’ reports on events and situations

(Mccombs, 2014, p. 1). The mass media therefore not only has a signaling function in which they are informing citizens about major events, but are “guiding” citizens’ perceptions on the most salient issues with their daily selection of news issues on the public agenda (Mccombs, 2014, p. 1). This primary level of agenda setting finds a lot of support among other scholars. Dunaway et al. for instance, demonstrated that an increase of news coverage shifted public opinion on the importance of the immigration problem in post-Katrina Louisiana (Dunaway, et al., 2011). Iyengar and Simon also found evidence for increased political concerns about the conflict in the Persian Gulf as a result of more television news coverage (Iyengar & Simon, 1993).

After scholars increased their efforts to research the agenda-setting theory, they found out that mass media was not only capable of telling people what topics to think about, but also to tell citizens which attributes they should find important. This second-level of agenda-setting becomes clear when considering elections. Candidates running for office are in agenda-setting terms a set of objects whose salience among the public can be influenced by news coverage and political

advertisement. Thus, media are able to influence the relative prominence of these candidates in the public mind. However, this is not where the media influence stops. Campaign managers are not only trying to secure news coverage and political advertising to increase the salience of their candidates. They are also trying to build an image of the candidate in which specific attributes are particularly prevailing (Mccombs, 2014, p. 45). With this second-level of agenda setting media, might not only be able to influence importance of objects and attributes, but also tilt public opinion towards a

particular perspective or preferred solution. Therefore Mccombs states:

“ Setting the agenda of attributes for an issue is the epitome of political power. Controlling the perspective of the political debate on any issue is the ultimate influence on public opinion“ (Mccombs, 2014, p. 51).

The impact of mass media on the public can even develop to the third level of agenda-setting. Third level agenda-setting builds on the concept of compelling arguments which recognizes the ability of the media to transfer relationships among the elements of the media-agenda to the public agenda.

News media would be able to bundle an object with an attribute and simultaneously make them salient in the public’s mind (Mccombs, 2014, p. 55). The news media transfer the salience of relationships among a set of elements to the public. These sets could be both objects or attributes, but more importantly a combination of objects and attributes, leading to a fully integrated set of objects and attributes (including attributes of the media message). At the third level of agenda-setting, there is no longer only attention for the quantity of media coverage but also for the actual content of news messages. The tone of a message would also affect the salience of an issue. Schoenbach and Semetko for instance show that in the national elections of 1999, the salience of problems in the former East Germany significantly declined among voters despite intensive news coverage (Schoenbach & Semetko, 1992). This decline was especially among readers of Bild, whose coverage of the integration of Germany was framed in highly optimistic terms (Schoenbach & Semetko, 1992). This focus on the content of a message also leads to the concept of framing. With framing, mass media are not only able to influence the importance of a certain issue among the public, but are actually able to steer public attitudes. Framing refers to selecting and emphasizing certain aspects of experiences or ideas over others (Andsager, 2000, p. 579). This process occurs mainly when journalists select some aspects of a perceived reality and make it more significant in a communicating text, in such a way as to promote a particular problem definition, causal

interpretation, moral evaluation, or treatment recommendation for the item described. According to Entman, frames in the media emerge as the presence or absence of certain key words, sources of information and sentences that form thematic clusters. Through their representation of news stories, journalists suggest attitudes and opinions for the public. Thus, with framing, the media goes beyond telling the public what is important and newsworthy. They also tell the public which opinions, interpretations and definitions of a controversial issue are most valid (Entman, 1991, pp. 7 - 9).

Thirdly, opinion leaders would have a significant impact on public opinion. They have the ability to define popular issues and influence individual opinions regarding them. Especially political leaders can turn a relatively unknown problem into a national issue, if they decide to give attention to the matter in the media (Davison, 2015). Page et al. show that in the United States especially, popular presidents would have a significant influence on public opinion. Public opinion would tend to shift in the direction supported by the president (Page, et al., 1987, p. 36). But presidents are not the only elites capable of affecting public opinion. Congressmen and lawmakers would also affect public opinion by contacting voters personally or by providing interviews in the mass media. Some scholars also point to the impact of party stances on policy issues which would in some cases just be copied by voters (Bullock, 2011)

Opinion leaders do not have to be extremely well-known public figures in order to influence public opinion. It can be any person to whom others look for guidance on a certain subject (Davison,

2015). Experts may therefore also be included into this category. Page et al. finds significant evidence for the impact of expert opinions in the mass media on public opinion. As an example they point to the Senate vote on the SALT II arms limitation treaty. Public support for the treaty dropped

significantly as a result of several testimonials from retired generals and arms experts against the treaty (Page, et al., 1987, p. 36).

Finally, some scholars describe that a changing public opinion results from the occurrence of a specific event. Brulle et al. for example point towards extreme weather events as a potential factor. Weather events might increase recognition and concern on the effects of climate change (Brulle, et al., 2012, p. 173). Page et al. also expected certain events to have a significant effect on changes in public opinion. In their article they expected to find significant effects of rapid fluctuations of the inflation and unemployment rate. They also expected that sudden events such as terrorist attacks and natural disasters would change public opinion significantly. Surprisingly they did not find a significant direct effect of these events on public opinion. However, they are quite convinced that these kind of events do have at least an indirect effect, for instance through opinion leaders (Page, et al., 1987, p. 38).

4.4 What role can interest groups play?

After this brief examination of the potential factors affecting public opinion, it could be argued that one important factor is not included: interest groups. Although some scholars argue that interest group efforts would not have that much of an effect on public opinion (Page, et al., 1987; Gerber, 1999) most of them agree that interest group activities matter for public attitudes (Davison, 2015; Page et al, 1986; Dür and Mateo, 2014; Dür, 2015; Keck & Sikkink, 1998; Berry, 1999; Kriesi, et al., 2007; Dür, 2008). The problem with most of this literature is however, that most scholars start their research with the assumption that steering public opinion is part of the outside lobbying approach and therefore focus their attention mainly on public interest groups. Berry for instance argues that in the United States citizen groups are successful in influencing public policy because they can affect election outcomes via their influence on voters (Berry, 1999, p. 3). Moreover, Dür and Mateo argue: “Especially business associations that are resource-rich and active in distributive policy fields should focus on inside lobbying, whereas citizen groups should consistently rely on an outside strategy” (Dür & Mateo, 2013, p. 661). Another consequence of this focus on outside lobbying is focusing

predominantly on means to manipulate public opinion that involve “going public” such as using the media. Keck and Sikkink (1998) for example state that NGOs’ credibility still depends in part on their ability to mobilize their own members and affect public opinion via the media. Furthermore,

abortion debate in 1995-1996 focusing on frames in press releases and news-stories. Accordingly, this results in the perception that influencing public opinion predominantly means “going public” which would be one of the reasons producer interest groups are somewhat reluctant towards influencing public opinion. Dür and Mateo (2014, p. 1204) for instance argue: “many members of business associations, namely firms that sell consumer goods and services, may be reluctant to see their names associated with highly unpopular campaigns for fear of losing customers”.

However, I believe that influencing public opinion might be more complicated than assumed. By separating direct- from indirect approaches to influence public opinion, this thesis not only provides more insight into alternative ways to influence public opinion, but also argues that

influencing public opinion as part of the outside lobbying strategy might fit producer interest groups better than expected. Whereas interest groups utilizing direct approaches try to steer public opinion without the “interference” of intermediary parties, indirect approaches influence public opinion “through” third parties. Steering public opinion therefore does not necessarily has to mean “going public” which might be beneficial to producers interest groups. Influencing public opinion indirectly, through the mass media or opinion leaders, might provide interest groups with some “cover” which enables them to steer public opinion, without being accused of doing so. Therefore, especially businesses struggling with reputational issues, might be able to “covertly” increase their chances to achieve policy goals by steering public opinion in a desired direction.

4.4.1 Direct influence

In order to increase chances to achieve policy goals, interest groups might attempt to directly

influence public opinion without the interference of a “mediator” or third party. In general, there are two important ways interest groups can do this. Firstly, they can directly influence the public by organizing events. An interest group might organize an information event in order to influence the visiting public by providing (biased) information. Environmental groups may for instance provide information about climate change whereas a corporate organization in the transporting sector may provide information about how companies can become more durable. By shedding their light on certain issues, interest groups might be able to steer public opinion (of the attendees).

Secondly, interest groups may try to directly affect public opinion by organizing and

conducting protest activities. Where some interest groups may use protest activities to make it into the news, the way it is used here, is focused on direct influence. A protest activity can be very

powerful in a direct way because it may affect the attendees. Furthermore, interest groups might use these activities as content for their magazines which also allow them to influence the public in a direct manner.

Finally, interest groups may also try to directly influence public opinion by deploying advertising campaigns. They can for example do so by publishing articles in their own magazines, publish articles in newspapers, and by buying advertising space on television, newspapers and the internet. It is important to note that this kind of influence is only direct when no third parties are involved. Buying commercial time on television for instance should be seen as a direct way to influence public opinion since interest groups are in full control of their commercial time as they have paid for it. The mass media is therefore not able to have a significant influence on the content of the message.

4.4.2 Indirect influence

Interest groups might also be able to influence public opinion indirectly. This indirect influence can be separated into two categories. To begin with, indirect influence might occur when interest groups are trying to steer public opinion by influencing journalists. This basically means that interest groups may try to increase the amount of attention mass media spends on a certain issue. Interest groups might be able to do this by trying to manipulate the media agenda in such a way that the media will increase their coverage of a certain issue. There are basically two ways interest groups can be influential in this regard (Thrall, 2006, p. 409). Interest groups might be able to place their issue on the media agenda by undertaking some form of dramatic action to get noticed as they generally lack the legitimacy of government officials (Thrall, 2006, p. 409). As Greenberg puts it:

“Dissenting voices, whether institutionalized within political parties, in NGOs or looser activist networks, often find that they have to brand their identities and overhype their interests and perspectives in order to ensure media access, to influence the terms of their public representation and to maximize public attention and political effect.” (Greenberg , et al., 2011, p. 69).

Furthermore, interest groups might gain access to the media by focusing on the relationship between journalists and their sources of information. Studies on the news making process have highlighted the importance of sources that are authoritative, credible, reliable, easy to get to and articulate and/or mediagenic (Thrall, 2006, p. 409). These kind of sources would be preferred by journalists because they give them a way to construct news stories more quickly and with greater confidence in their accuracy and relevance. Additionally, authoritative and credible groups would also provide journalists with a way to balance commentary from other political actors or to bring new data and information into the news while avoiding charges for being biased. Thus, journalists would choose some sources over others because they provide a better foundation for journalists while constructing

and balancing their news stories. Therefore more authoritative, productive and reliable interest groups should expect to make more news than those that are not. Groups with the scientific horsepower to provide authoritative information on economic, environmental and other trends provide input for many news stories and put themselves into a position to provide commentary in many more (Thrall, 2006, p. 409).

To make this point more concrete, interest groups may put this role of authoritative and reliable sources into practice with the process which Mccombs calls: "subsidization". Journalists are only able to observe a small fraction of daily situations and events. Therefore, they are mainly dependent on information provided by public information officers and other public relation

practitioners. Thus, these communication professionals subsidize the efforts of news organizations to cover the news by providing organized information, frequently in the form of press releases. Almost half of the news stories coming from the New York Times and Washington Post would be

substantially based on press releases and other direct information subsidies (Mccombs, 2014, p. 115). In short, interest groups are able to affect the media agenda by subsidizing journalists for example by providing news releases, phone calls with journalists, press conferences and so on.

The second way interest groups might be able to steer public opinion indirectly is by utilizing opinion leaders. As described in the previous part, there is some evidence that opinion leaders have a significant effect on public opinion. By defining popular matters they have the potential to influence individual opinions (Davison, 2015). In order to sway public opinion I expect interest groups to attempt to mobilize political elites in order to spread the arguments of these interest groups. This mobilization of political elites might occur when political elites are reaching out to the mass media (providing interviews or show up in television shows) or by directly contacting voters (visiting cities and call voters for instance). The most obvious way for interest groups is mobilization through direct persuasion. Most interest groups have regular meetings with political elites to inform them about certain issues. Political elites are to a certain extent reliant on information and expertise which most interest groups provide. On their websites, almost all interest groups argue to invest in knowledge and technology which would make them experts in certain fields. When interest groups are able to convince political elites of certain attitudes, these elites might take a stand and spread this argument through certain channels such as the media.

Additionally, interest groups may also try to mobilize “unbiased” experts like scientists or research agencies to spread their arguments as these experts may also fulfill the role of opinion leaders. Arguments automatically sound much more convincing coming from experts compared to arguments spread by certain interest groups.

Table 1.1 Schematic diagram presenting the options interest groups have to influence public opinion.

4.5 Theoretical model

Although it is very likely that most interest groups will use a combination of the aforementioned means to steer public opinion, it is not likely that all interest groups will use them to the same extent. The following factors might have a significant impact on strategy choice.

Firstly, I expect group type to be of critical importance in determining the right strategy to influence public opinion. As argued before, I believe that the major difference between interest groups is the broadness of their interests. On the one hand, the diffuse- so called public interest groups lack a well-delineated and in socio-economic terms concentrated constituency. They promote diffuse- or public interests that are linked to broad or general segments of society (Beyers, 2004, p. 216). Because these interest groups defend especially public interests and might therefore be conceived of as “defenders of the common good”, I expect them to act in most cases on the most popular side of the conflict. On the other hand “producer interest groups” have very specific interests. These types of interest groups have a very clear-cut stake in the production process and defend the interests of well circumscribed and concentrated constituencies (Beyers, 2004, p. 216). These interest groups are not focusing on defending the common good, but are mainly defending their own interests or the interests of a very specific group. Therefore I expect these interest groups to act mostly on the less popular side of the conflict and should therefore be much more careful with their strategy choice.

The chosen strategy when acting on the popular side is most likely very different from the applied strategy of interest groups acting on the non-popular side. In the first situation, interest groups are likely to exploit their position by gaining as much attention as possible. These interest groups do not have to worry about getting a bad reputation, since their positions are most likely supported by the majority of society. Interest groups, defending popular issues, are therefore not limited in their options. They will probably combine as much options as possible to achieve the biggest impact on society and most importantly, keeping their stances popular. Interest groups defending less-popular issues on the other hand are more interesting to consider in this regard. Since these interest groups do not have the advantage of being supported by the majority of society on their stances, these interest groups will probably try to influence public opinion without being “accused” of doing so. These interest groups will therefore most likely not utilize many of the direct strategies, since these strategies directly link their names to an unpopular issue. Indirect strategies fit these interest groups better since using a mediating third party might provide them with some disguise. This disguise enables them to still have impact on public opinion without being accused of trying to affect citizens towards their own interests. These types of interest groups will therefore probably try to manipulate public opinion through the mass media by subsidizing journalists with

information and not by trying to be newsworthy. These interest groups will probably try to influence journalists by supplementing information which may lead to favorable articles being published in newspapers or more favorable coverage on television without being directly attached to the subject. The mobilization of opinion leaders is also a great opportunity for producer interest groups as opinion leaders provide great coverage and above all credibility. If a large oil producing company for instance claims to produce oil in a very environmental friendly manner, it is unlikely that this

company will be able to directly convince opposing parties. However, if the same oil company is able to mobilize experts and political elites to spread the same argument, chances increase significantly to convince more citizens. Although, it is expected for these interest groups to act much more “off-the-grid”, they will probably also try to influence public opinion by using advertising campaigns because these can be very effective and above all provide these interest groups with full controllability. If we take the oil company as an example again, they can supplement their other, more indirect strategies with television advertisements. Although these advertisements might not provide the same

credibility as external experts, they still might be able to convince a certain part of the general public. Besides, this strategy offers full controllability as they have paid for this commercial time, third parties are not allowed to intervene or adjust their commercials.

H1: Whereas public interest groups, will use a combination of all outside strategies to influence public opinion (combinations direct and indirect), producer interest groups will focus more on indirect outside strategies supplemented with advertisements.

Additionally, group type might influence strategy choice in another way. As Dür and Mateo (2013) argue, member-based organizations are more inclined to use the “outside lobbying strategy” in order to survive than non-member based interest groups. Although “inside strategies” might sometimes be more efficient, these member-based interest groups are sometimes forced to “go public” in order to be visible to their members (Dür & Mateo, 2013, p. 663). Especially public interest groups that are member-based, or represent a certain sector might experience pressure to show their members that they are doing everything they can to defend their interests and are using their funds wisely. These interest groups are most often largely funded by their members or by the corporations they represent. Their existence relies to a large extend upon these members. This results in a certain “pressure” to be visible to their members and prevent them from recalling membership. Producer interest groups might experience this pressure as well. Although I agree that the urge to go public for producer interest groups might be slightly lower as businesses are better able to know where their funds are going, they still might experience the same kind of pressure (Dür & Mateo, 2013, p. 663). Especially when producer interest groups represent a lot of small businesses,

the mechanism is likely to work the same. Is it for example likely that small agricultural businesses will put a lot of energy into finding out where their funds are going? Member-based interest groups will therefore probably use strategies that allow them to reach their members easily. Moreover, the applied strategies should “prove” that earned contributions are being spend wisely. Although direct strategies might appear to be appropriate since these allow efforts to be directly attached to interest group names, they are restricted in their abilities to reach broad-based constituencies. Therefore the mass media (part of the indirect strategies) seems to be the logical choice since this allows interest groups to reach their members easily. However, interest groups will have to make sure that the mass media attaches their efforts to their names by explicitly mentioning the interest group. On the other hand, non-member based interest groups, will not experience this kind of pressure and are therefore not forced in this regard to use a certain strategy. They enjoy the freedom to pick whatever strategy they prefer.

H2: Member based interest groups are more reliant on indirect outside strategies (mass media) than non-member based interest groups.

Thirdly, the issue at hand might have an impact on strategy choice (Beyers, 2008; Dür and Mateo, 2013). Especially the amount of attention an issue receives may determine the right strategy. If for instance an issue is already highly salient there is no need for interest groups to put effort into increasing attention. Since the issue is already salient it makes much more sense to put more energy into guiding the attention in the right direction by focussing on strategies which allow them to strengthen their position in the debate and weakens the opponents’ position. Probably the best way to steer public opinion is by convincing the public that the stances of the interest group are correct. In order to do so, interest groups will probably rely on indirect strategies as these are very suitable to reach large audiences and are open for expertise. The role of expertise becomes very important to them as the recruitment of third parties (especially experts) enables them to promote their stances in a much more convincing way and therefore might add credibility. In order to mobilise this

expertise, interest groups will probably focus on subsidizing expert knowledge to mass media and by mobilising opinion makers. If on the other hand an issue is not very known to the public, interest groups will probably focus more on increasing attention for the issue in the first place. To be able to increase attention, interest groups will probably combine as much possibilities to increase issue attention as possible.

H3: On low salient issues interest groups rely on both direct and indirect outside strategies, whereas on highly salient issues interest groups focus on indirect outside strategies.

Lastly, the potential role of resources deserves further examination. Although resources are considered by some scholars in a broad sense, I consider them more narrow. Thrall for instance, considers resources in his article as “organizational resources” which are basically a combination of tangible (money, staff and members) and non-tangible resources (reputation, expertise and

intellectual capital) (Thrall, 2006, pp. 409 - 410). Although, I agree that resources should normally be considered broader, this theoretical model pays attention to non-tangible resources when taking the effect of group type into account. Naturally, I expect the availability of financial resources to

determine which means interest groups have at their disposal. Accordingly, I expect direct strategies to be more resource demanding than indirect strategies. Organizing protest activities or

informational events are mostly demanding activities that require both time and money. Protest activities might not seem very resource demanding. However if interest groups want to make it into the media, these protest activities need to have a significant size which asks for a great deal of organizational capacity. Organizational capacity requires a large staff which again entails a lot of financial resources. Especially when it comes to advertisements, financial resources are crucial. Again, the amount of money needed to conduct an advertisement campaign depends on its size, but as Greenberg points out one of the most aggressive media campaigns (1997) conducted by the Global Climate Coalition cost around 13 million US Dollars (Greenberg , et al., 2011, pp. 70 - 71). Since resource-poor interest groups do not possess many financial resources, they are more likely to focus on indirect strategies. The media seems to be a relatively inexpensive alternative to influence public opinion. Reaching out to the media by sending them press releases or letters does not have to be expensive. Especially the uprising of social media provided resource-poor interest groups with new options to reach large audiences with “controlled” messages.

The fact that resource-rich interest groups possess a lot more financial resources does not necessarily mean that these groups automatically use resource-intensive means to steer public opinion. However, if it turns out that for one reason or another their demands are best met by for example conducting an expensive advertisement campaign, they have this option available.

Therefore I expect resource-rich interest groups to choose the options that suits their interests best, without worrying about the costs.

H4: Whereas resource-rich interest groups will use a combination of both direct and indirect outside strategies, resource poor interest groups will rely mainly on indirect strategies.

5. METHODOLOGY

In order to find out which factors interest groups take into account when setting up their final strategy, I decided to make use of semi-structured, in-depth interviews. Although I believe that a quantitative approach would have enabled me to find out which means certain interest groups prefer and which factors they take into account, I preferred the qualitative approach for two reasons. To begin with, I did not only want to find out which choices interest groups have to make when forming their strategy, but also wanted to understand the reasoning behind their responses. In-depth interviews provide the freedom to ask follow-up questions which allow interviewees to explain choices more in-depth. Secondly, I wanted to make sure that I was not ruling out any factors

beforehand. The in-depth interview allows the interviewee to make the “conversation” flow into the direction he/she finds important. This way, the interviewee does not only have the freedom to name any additional factors, but can also emphasise the importance of some factors over others. The quantitative approach is more restricted in this sense. More or less the same reasoning explains the preference of the semi-structured interview style over the completely structured interview style. Again, the semi-structured style provides more freedom and therefore allows the interviews to be more conversation-like. Off course, I could have chosen the non-structured interview style as this allows for even more freedom. However, for a relatively non-experienced interviewer, I believe it is better to rely on some structure. This made sure that I was not forgetting any important elements. Although, I prefer to interview in a face-to-face fashion, I was forced to conduct about half of the interviews over the telephone. This was due to the fact that some of the responding interest groups were situated in cities outside my reach. However, I do not believe that interviewing by phone has a significant negative influence on the results as the interviews were still “conversation-like”.

5.1 Approach

To make my results as complete and reliable as possible I contacted as many Dutch interest groups as I could. At the beginning, I started emailing around twenty interest groups from a few different sectors as I did not know what type of reactions I would receive. In a couple of days the responses turned out to be very disappointing. The first twenty emails only brought me three interviews. As a result, I decided to change my approach. Every day, I picked one sector and tried to email as many active interest groups (both public- and producer interest groups) as possible using google as my primary source (both public interest groups and producer interest groups). As the response rate stayed fairly low, I kept doing this for at least two weeks. After sending out around 70 emails to interest groups active in twelve different sectors, I was able to arrange 15 interviews. The responding interest groups turned out to be active in the food sector, energy sector, the pharmaceutical sector,

the transporting sector, the agricultural sector, the tobacco industry and the environmental sector. Interestingly enough, especially interest groups active in the food-, soda-, and tobacco industry turned out to be very reticent. Almost all businesses active in these sectors responded to my emails by stating that they had to deal with too many interview-requests and therefore had decided to stop participating. Additionally, it should be noted that it was surprisingly hard to arrange interviews with public-, environmental interest groups as they also explained to be too busy. This may seem to be a little surprising since several scholars argue that public interest groups are to a large extent reliant on outside lobbying (Dür & Mateo, 2013; Kollman, 1998; Dür & Mateo, 2014; Binderkrantz, 2008). Therefore I expected a high response rate among these types of groups. However, it should be noted that this thesis was written in the period between November 2015 and January 2016. This was right around the United Nations Climate Summit in the beginning of December in Paris, France. This might explain their high rejection rate. In the end I was able to arrange five interviews with employees working for public interest groups and ten interviews with employees working for producer interest groups spread among the different sectors. It should be noted that although I conducted 20

interviews, I was only able to include 17 of those into my transcriptions. This was due to technical complications which disturbed my recordings.

After arranging these interviews, I started emailing consultancy agencies, specialised in public affairs. As these agencies turned out to be highly responsive, I easily arranged five interviews. The reason I wanted to include these agencies into my analysis, is that employees working for these agencies most often have a lot of experience defending the interests of both public- and producer interest groups. As a result, these employees might have a better understanding of the differences between public- and producer interest groups and the different factors involved. With this

experience, working with both interest groups in mind, I figured that these interviews could add a relatively unbiased input. Although these agencies also have their interests (appearing qualitatively good), I believe these interest are not conflicting with the input given in my interviews and therefore can be interpreted as “unbiased”. Conducting these interviews in the end, also allowed me to check the correctness of findings from the first fifteen interviews.

All interviews were based on a few basic questions to make sure every concept was covered during the interview. This allowed me to ask follow-up questions on topics the interviewee found important. The first few questions were set up to get a better understanding of the origins of the interest group, the members it represented and the goals they pursued. The first question was: can you tell me something about the organization? In most cases this question already covered the aforementioned information, but sometimes a follow-up question was needed: What does the organization try to achieve with their lobby efforts? After these basic introductory questions, a few basic questions were asked to understand the general strategy of the interest group: What is the role

of public opinion in the general lobby strategy? What is the organization doing to influence public opinion? After answering these questions I asked a few follow-up questions about other possible strategies to make sure that the interviewee was not forgetting one of them and to discover the reasoning behind these strategies. If for example the interviewee did not mention anything about protest activities or organizing events, a follow up question was asked. For example: what about protest activities, is the organization using those to influence public opinion? After achieving a better understanding of the means interest groups used to influence public opinion, I asked some follow up questions in order to find out if my hypothesized factors seemed to influence their choices.

Therefore I structured these questions into three categories. Firstly, I tried to find out whether interest type has an influence on their strategy or not. Therefore I asked: Do you think that the type of interests this interest group defends has any influence on the means to choose from? To find out more about the influence of issue salience I asked: Are highly debated issues, in both media and society treated differently than issues that did not receive a lot of attention? In the third category, resources, I asked the basic question: What do you think of budget, does it play a role in the decisions you make to influence public opinion? After these basic questions were asked, different follow up questions were asked in order to get a better understanding of the reasoning behind the answers given to the basic questions. I am aware that especially the last few basic questions are slightly direct and might have primed the interviewees to a certain extent. However, after the first few interviews I found out, that the interviewees were not always aware of the factors they took into account when forming their strategy. This resulted in very short, non-informative answers. Therefore I had to ask the questions in a slightly more direct fashion to be able to gauge whether group type, issue and resources had some sort of influence or not. The interviewees seemed to take these factors unconsciously into account, but did not include them into their answers. By asking the

aforementioned questions, their answers suddenly became much more valuable.

5.2 Validity and reliability

In order to test my hypotheses, I distinguished five different variables. The dependent variable in this thesis is the chosen strategy to influence public opinion. Building on theory, I expected three

independent variables to influence strategy choice: group type, issue and financial resources. In the group type variable I distinguished two elements: the type of interests (public- or producer interest group) and member- or non-member based. In order to structure my analysis I coded the

transcriptions with five different codes using different colours. I highlighted all answers given regarding group type, the effect of membership, issue-related answers, resource-related answers and “unexpected” answers.

I believe the way this research was set up provides a decent validity. By interviewing

lobbyists and public-affairs specialists from interest groups active in many different sectors I believe that I can achieve a better understanding of the factors interest groups take into account when choosing their strategy to influence public opinion. By not focussing on interest groups active in one sector, but spreading the interviews over a decent amount of sectors I was better able to identify general- and more sectoral specific factors. Additionally, the semi-structured interview style made sure that I was not forgetting important elements while leaving more than enough room for

interviewees to take the conversations into the direction they found important. Therefore I was not ruling out any important factors beforehand. This also contributed to the internal validity of this thesis because if other intervening variables would be present, interviewees would probably pointed at the existence of these factors. Furthermore, the last few interviews with consultancy agencies, provided a relatively unbiased input. This allowed me to make sure that interest groups did not leave out specific factors on purpose and allowed me to check earlier findings. When it comes to external validity, I believe that as a result of the diversity in interest groups and sectors, the results can be generalised to a certain extent. However, as pointed out before, I was not able to arrange interviews with interest groups from all different sectors. Therefore I might not have been able to identify sectoral specific factors. Additionally, due to practical reasons, I was only able to arrange interviews with Dutch interest groups which might compromise generalisability to other countries. Country-specific institutional circumstances, might cause interest groups to take different factors into account when choosing the ways to influence public opinion. In the Netherlands, the pharmaceutical sector has to deal with totally different rules and regulations than the interest groups active in for example the United states. Therefore we should be careful taking these results to other contexts. However, apart from the purely sectoral differences, most factors hypothesized in the theory do not seem to be very context specific. The effect of group type for instance does not seem to be very context related. Interest groups defending specialised interests will always have to deal with reputation, whether they operate in Germany or the Netherlands. Moreover, a lot of rules and regulations are applied on the European level which causes me to believe that contextual differences within the European Union should not be that significant.

Lastly, when it comes to reliability I did everything I could to make sure that the interviewees provided honest answers to my questions. Therefore, I guaranteed them full anonymity (no names of both interviewees or companies). Although many of the interviewees explained not to have

problems with their names being mentioned (as long as they could see citations) I still assured them to provide full anonymity. This way I made sure that they could speak freely. Secondly, as mentioned before, the interviews conducted with consultancy agencies made sure I could check previous findings and provided some relatively unbiased information. Finally, after roughly ten interviews, the

given answers turned out to be repetitive which increases my confidence in the reliability of the results.

6. WHAT EXPLAINS STRATEGY CHOICE?

A thorough analysis of the conducted interviews revealed several interesting insights into the factors interest groups take into account when forming their final strategy to influence public opinion. For the sake of structure, I present the findings in the same order as the theoretical model, categorized on the potential factors: group type, issue and resources.

6.1 The influence of group type

Firstly, group type seems to have a significant impact on strategy choice. Although all interest groups explained to use a combination of both direct- and indirect means to influence public opinion, producer interest groups seemed to focus solely on indirect approaches. Public interest groups mostly focus on a combination of both strategy types to influence public opinion. As far as direct approaches goes, most public interest groups, declared not to be as activist orientated as an interest group such as Greenpeace, but did declare to join or support protest activities. Especially public interest groups seem to organize or support events from time to time, but not always with the purpose to influence public opinion (sometimes events were organized for networking purposes or for the members). When it comes to indirect approaches, all public interest groups said to use the mass media in one way or another. All of them were subsidizing media by sending out press releases, letters or by personal contact. Being newsworthy was most of the time important to them, but not in an activist way. They almost all declared that newsworthiness is of utmost importance to get your article published. Messages should therefore contain newsworthy information if you want to make it into the news. Especially interest groups with a very broad membership base declared that polls held among their own members are very effective because these polls provide journalists with unique information. The distinction between subsidizing journalists by providing information or making it into the news by being newsworthy, might therefore not be as clear-cut as argued beforehand. Last of all, it is hard to find a pattern in the use of opinion leaders. Only one public interest group, which seeks to combat poverty in third-world countries, explained: “we are “using” politicians and

celebrities for example as ambassadors which subsequently express themselves in television shows about our projects” (Attachment, p. 12 – 13). However, most public interest groups declared not to “use” politicians for these purposes. They declared to approach politicians regularly to persuade them into certain stances. If these politicians spread these arguments afterwards, they naturally welcomed this. However, this spreading of arguments through politicians was more of a wishful side-effect than a goal in itself. When it comes to experts, most public interest groups declared that expertise is very important to them. They all argued that providing correct information is crucial if

you want to be taken seriously in the future. Most of them explained to see themselves as experts, since they mostly have a lot of experts working in the field.

Producer groups on the other hand, indeed seem to focus a lot more on indirect approaches. Most of them declared not to participate in protest activities or events. The events that they did organize or support were not meant to influence public opinion, but were more networking related. Strikingly, advertisements were not used as much as expected. Most corporate interest groups explained that paid advertising is just not worth it. As one pharmaceutical company puts it:

“Advertising is very expensive, if I would start a large campaign, my budget would be completely gone. Besides, advertisements are very volatile. Look at the health insurance companies for instance, they just launched a big advertising campaign. Did you see them? No, that is why I do not use them.” (Attachment, p. 48).

Just like public interest groups, producer interest groups rely mainly on mass media (newspapers, magazines and social media) if they find it necessary to go public. Contrary to public interest groups, producer interest groups declared to rely heavily on third parties and opinion makers in their efforts to influence public opinion. They declared other stakeholders to be of crucial importance to send messages to the public as these stakeholders were able to strengthen their messages with more impartiality and expertise. Where most public interest groups conceived themselves as experts, producer interest groups argued that they had the feeling that the general public would not take a message seriously coming from a commercial organization. Therefore external experts and

stakeholders were found to be of crucial importance. Producer groups were also not using politicians as opinion leaders, but again expertise, especially provided by external stakeholders seems to be of crucial importance.

Although these patterns seem to confirm the first hypothesis, interest groups do not seem to reason similarly to arguments made in the first hypothesis. Although indirect approaches might provide producer interest groups with some cover, this does not seem to be the primary motivation to rely mainly on indirect approaches. As became clear from several interviews, interest groups only go public if they have the feeling that this would benefit them. Therefore there is no reason to hide. Admittedly, sometimes interest groups are forced by other actors to react to a certain issue (for example when another interest groups blames a producer interest group for some matter) despite the fact that this corporate group was initially not willing to go public on this issue. However, in those situations, the identity of the interest group is already uncovered, so even in these kind of situations it does not make sense to use indirect approaches to provide cover. The true motive to focus on