Political strategies in direct-democratic campaigns - NCCR Democracy

9 downloads 32044 Views 180KB Size Report
Our approach to the study of campaign strategies of political actors starts from a ... American military scholar, strategy is best expressed in terms of ends, means ...
National Centre of Competence in Research (NCCR) Challenges to Democracy in the 21st Century

Working Paper No. 8

Political strategies in direct-democratic campaigns

Hanspeter Kriesi, Laurent Bernhard, Regula Hänggli

July 2007

NCCR Democracy Stampfenbachstrasse 63 8006 Zurich, Switzerland Tel.: + 41 44 634 52 04 Fax: + 41 44 634 52 00

2 Abstract In this paper we make an attempt to provide a framework for the analysis of the political actors’ strategies in direct-democratic campaigns. We conceptually divide the set of choices into three subsets –coalition formation, mobilizing, and crafting of the message. Mobilizing choices include decisions about timing, targeting and the communication repertoire. Crafting the message refers to decisions about agenda-building (priming, framing) and rhetorical strategies. We formulate a number of hypotheses about the possible determinants, and the likely effects of the various choices involved in mobilizing and crafting of the message, at the level of the public sphere as well as at the level of the citizen public. Keywords: Political campaigns; strategies; direct democracy; Switzerland

3 Introduction The study of campaigning activities has long been neglected in political science (Farrell 1996, Schmitt-Beck und Farrell 2002). The disregard of campaigning can be explained by the fact that it has long been taken for granted that campaigns only have minimal effects. More recent studies have challenged this ‘minimal effect’ view of campaign effects. Indeed, Iyengar and Simon (2000) maintain that campaigns do matter and can be pivotal. Similarly, several recent studies (Nadeau et al. 2001, Lachat & Sciarini 2002, Fournier et al. 2004 or Holbrook & McGlurg 2005) provide evidence for substantial and systematic campaign effects on citizens’ voting behaviour. However, earlier expectations that campaign messages manipulate easily persuadable voters have been abandoned. Increasingly, campaigns are viewed as playing a key role for the information processing of the citizens, providing voters with the necessary information for making a choice in line with their preexisting preferences (Finkel 1993, Gelman and King 1993, Stimson 2004, Arceneaux 2005). As Stimson (2004: 112) observes, one of the best established findings in voting research is that voters rarely hold a preference for a candidate and then change it to the opponent. What happens instead is that the campaign, by increasing the information about the voting decision available to the voters, helps them update their beliefs regarding the issues at stake and to reach a decision in line with their preferences. Following Gelman and King (1993), Arceneaux (2005) calls this the ‘entlightenment hypothesis’. Under ordinary circumstances ordinary citizens are paying too little attention to and know too little about politics to reliably make coherent choices. Before the campaign starts, large numbers of people have not paid attention to politics, haven’t thought about it, and don’t really have any views about the specific issues at stake. Thus, polls taken early in a campaign often don’t measure real attitudes. What they often tap is a response to satisfy the needs of the interviewer. Later, when replaced by a real attitude, it will appear to have changed. During the campaign, the citizens’ attention becomes focused on politics more generally (during election campaigns) or on specific political issues (during direct-democratic campaigns), and the citizens learn a lot about the choice they have to make. Campaigns are information-rich events which allow the citizens to connect their preferences to the choice at stake and to make a decision in line with their preferences. As Sniderman and Levendusky (2007) point out, the clash of arguments during the campaign rather than confusing the citizens clarifies the choices before them. Our approach to the study of campaign strategies of political actors starts from a simple heuristic framework with three types of actors who are all involved in the communication processes that constitute a political campaign – political actors, the media and the public (Figure 1). We approach the interaction of the three types of actors in a political campaign from the point of view of the political actors and their strategies. These actors attempt to

4 control their fellow politicians, the media, and the public in order to impose their message in the campaign, which should contribute to their success in the popular vote. They are the ones who usually initiate campaign events, and who provide the key informational input in the campaign.

Figure 1: General framework of the module

Context

Political actors

Media

Public

Strategic action is a variant of instrumental (teleological) action that includes in the actor’s calculation of success the expectations about the decisions of at least one other goaloriented actor (Habermas 1981: 127). In game theory the term strategy denotes a plan for a player to play a game (Morrow 1994: 352). In elaborating their action plans, the actors are taking into account and exploiting the rules of the game as well as the possible reactions of their adversaries. That is, strategic action is strategic interaction, ‘in which you face other players who regard you strategically, just as you do them, and engage in a series of actions in response to others, anticipating their reactions in turn’ (Jasper 2006: 6). As Jasper (2006: 171) also observes, ‘in strategic action there are few rules …. but many choices’. This makes conceptualization of strategic action particularly difficult. According to Lykke (2001), an American military scholar, strategy is best expressed in terms of ends, means and ways. Ends refer to the objectives one strives for. Means are the resources designed to be used for

5 pursuing these objectives. Ways answer the crucial question of ‘how” the objectives are to be achieved. They link the means to the objectives by deciding what, where, when has to be done how. In short, any strategy consists of the articulation of objectives, the formulation of strategic concepts and the coordinated use of resources. Strategic political actors are confronted with media who are largely autonomous, and upon whom they increasingly depend for reaching out to the public, and for obtaining information about the public. As Koopmans (2004: 379) has observed, the information available to the political actors is prestructured by the media: ‘Paraphrasing the famous one-liner about the role of agency in history one might say that people make their own history, but on the basis of an information input not of their own making’. In the past, the increasing role of the media in politics and their increasing autonomy with regard to politics have been discussed mainly as a U.S. American phenomenon, but, more recently, European systems have observed similar trends (Swanson and Mancini 1996, Hallin and Mancini 2004). Although the political and the media systems are differentiated spheres with distinct logics, success in politics strongly depends on media goodwill and resonance. During heightened political conflict, in particular during intense campaigns, the media assume central importance. They bend politics to their own logic, their own sense of time, their own commodity-driven needs. They selectively confer public status on political actors. It is almost exclusively left to the news media to decide which actors and voices of society gain access to the public debate. The media also determine which political events come to the attention of the public and impose a systematic bias upon the media reality of politics. In other words, they assign political relevance and salience to social problems, and, by emphasizing certain issues and neglecting others, they set the public and, indirectly, the political agenda. The media construct the meaning of political events and personalities according to their own commercial logic. Moreover, the media themselves become key political actors of their own, who are able to influence political decision-making processes and, in rare instances, even as mobilizing agents. Political communicators are forced (and have learnt) to respond to the media’s rules, aims, production logics, and constraints (Mazzoleni and Schulz 1999). As a result, political actors and the media interact in complex ways in the public sphere which cannot simply be instrumentalized by one of the actors involved. Instead, we should assume a ‘symbiotic constellation of mutual dependence’, where all the participants are trying to optimize their control over the events (Schmitt-Beck and Pfetsch 1994: 115). In order for the political actors’ messages to reach the public of voters, they have to pass three selection mechanisms (Koopmans 2004: 371ff.). Access to the media is only the first one of them. This mechanism depends on the message’s news value (which, in turn, depends on such things as (geographical) proximity, prominence and prestige of the speaker, the level of conflict related to the message or the actor, the salience of the issue,

6 possibilities for dramatization and personalization, and the novelty of the story), the journalistic routines (news coverage is affected by the mundane constraints of a reporter’s job, specifically the problems of getting information and writing to a deadline), and the predispositions of the journalists (overt editorial policies, the ideological position of a newspaper that biases its reporting)(see Oliver and Maney 2000: 468ff.; Tresch 2007). Koopmans calls the number of communicative channels in which a message is included, and the prominence of such inclusion ‘visibility’. In addition, political actors need to evoke a reaction in the public sphere. If they do not succeed in provoking reactions from other actors – the media, bystanders, the public, their opponents, allied sponsors, their messages remain stillborn. The degree to which a message provokes a reaction we can – again following Koopmans 2004: 374) – call ‘resonance’. Messages that resonate travel farther and increase the actor’s chances of reproducing the message in the public sphere. Resonance can be either supportive (=’consonance’) or adversarial (=’dissonance’). The maxim that ‘any publicity is good publicity’ also holds for political messages. Dissonance includes any form of public action that condemns, expresses disagreement with, or actively counters an actor, his actions, or his aims, ranging from unfavorable verbal statements to various forms of repression or countermobilization. The balance between positive and negative responses, i.e. degree to which, on average, reactions by third actors in the public sphere support or reject an actor’s messages constitutes the third selection mechanism, which Koopmans (2004: 375) calls ‘legitimacy’. We prefer to call it ‘popularity’ or ‘support’. Koopmans suggests that, defined in such a way, ‘legitimacy’ (‘popularity’/’support’) is independent of resonance: highly legitimate messages may have no resonance at all because they are uncontroversial, while highly illegitimate messages may resonate strongly. The predicament of political actors is that the aspects of a political event (such as its ‘violence’ or ‘originality’) that create high news value and allow for visibility and resonance of their message, at the same tend to diminish its legitimacy. The distinctive premise of more recent analyses of a campaign’s impact on the citizen public is that ‘campaigns do not occur in vacuums but instead blend in with voters’ partisan motives and attitudes’ (Iyengar and Simon 2000: 160). According to the ‘resonance’ model of campaign effects, campaign messages work their influence in concert with voters’ prevailing predispositions and sentiments. In other words, the campaign effects are ‘contingent on the degree of fit between campaign messages and prevailing attitudes. New information intermingles with the old, and depending on the chemistry, voters’ choices will or will not be affected’ (Iyengar and Simon 2000: 158). Strong prevailing attitudes will be further reinforced by the campaign, which is why intensely conflicting campaigns tend to have a polarizing effect (Zaller 1992). Voters with strong prevailing attitudes – the ‘passionate’ in Stimson’s (2004: 163-165) terms –, however, never switch sides. Stimson distinguishes a second group

7 of voters – the ‘uninvolved’ –, which constitutes the exact opposite of the voters with strong attitudes. The ‘uninvolved’ do not pay attention to politics, under no circumstances, their views are not responsive to what happens in politics, and their movements have a random character. Accordingly, they are quite irrelevant from the point of view of the campaigners. From the campaigners’ point of view, Stimson’s third group – the ‘scorekeepers’ – is the crucial one. This group consists of ‘nonideological pragmatists’ who trust or distrust each side equally. They pay enough attention to and respond to the common signals of politics, yet they are not so involved as to be committed always to one side. They are what Schudson (1998) has called the ‘monitoring citizens’, i.e. citizens who, although not involved in politics on a daily basis, keep a watchful eye on the political scene and become more interested in politics as the political communication intensifies (such as in the context of a political campaign). For the possible outcome of a campaign, this group is crucial and a lot depends, of course, on its relative size, as well as on the distribution of the preferences across the three groups. The campaigning of the political actors is embedded in a specific political context, which determines to a large extent their action repertoire as well as their chances of success. In other words, the campaign and its possible outcome is highly prestructured by the institutional setting (including the media system), by the characteristics of the issues at stake, by the short-term events (exogenous shocks) intervening during the campaign, and by the configuration of actors who participate, their goals and resources. Against this background, in the course of the campaign, political actors craft and communicate their messages (appeals) with the purpose of persuading, activating, and mobilizing the public. In order to do so, they need to pass through the filter of the three selection mechanisms we have described above. The goal of our project is, first, to explore and to describe the different strategic choices made by the political actors – choices referring to the substantive content of their messages – their agenda-building and rhetorical strategies (appeals) –, as well as their ways of communicating their messages’ to the public – the timing of the messages, their intended targets and the communication repertoire the employ. Next, we shall attempt to explain these choices as a function of the context conditions, and the specific reactions of the various actors involved. Finally, we shall try to assess the effect of the different appeals in terms of their visibility, resonance and support in the public sphere and in the citizens public, as well as in terms of the citizens’ voting behavior. Figure 2 summarizes the general causal structure of our approach.

8 Figure 2: General causal structure of the approach

context conditions

coalition formation

mobilizing/ crafting the message

impact

- institutions - issues - exogenous shocks - predispositions

- configuration of actors - goals/beliefs - resources

- timing - targeting - communication repertoire - agenda-building - rhetorical strategies

- reactions (resonance) - visibility - support - voting behaviour - evaluation

Thus far most of the research on campaign effects has been done with data from US elections, although there is an increasing number of studies referring to campaign effects in a cross-national electoral context (Arcenaux 2005, Stevenson and Vavreck 2000, Banducci and Karp 2003, Schmidt-Beck and Farrell 2002). But there are to our knowledge very few studies of the effect of direct-democratic campaigns and even less studies of the campaign strategies in such campaigns. Given the overall notion that there is room for campaigns to matter, and to matter a lot, we follow Rohrschneider’s (2002: 308) suggestion, that ‘campaign decisions are an area too important for political scientists to ignore’.

The institutional context It is of crucial importance to keep in mind that any political choice and any campaign to influence such a choice is taking place in a given context. As Walder (2006: 713) points out, ‘a theory about political choice – or any theory about politics – can be no more valid than its claims about the contexts within which these choices are made’. We are proposing to study campaign strategies of political actors in direct-democratic campaigns in Switzerland. This means that we have to, first, take into account the specific institutional context of such campaigns. Political institutions and configurations define the rules of the game of a campaign as well as the logic of interaction between the political elites, the media and the citizens. Political institutions are, as Sniderman (2000: 69) points out, the ‘organizers of political choices’. To make reasonable choices, the citizens do not have to rely exclusively on their own limited informational resources and computational capacities. Instead, they get help from the political institutions which put them in a position where they can make coherent choices (Sniderman and Levendusky 2007). Thus, Swiss voters have grown up with the direct-democratic institutions and have acquired a working knowledge without making explicit

9 effort to understand them. Given this background knowledge, their task is not as complicated as it may seem at first sight. Moreover, their task is further facilitated by the fact that the institutional framework reduces the decision to a binary choice1, that the political forces usually line up in two opposing camps, and that the vote is preceded by a campaign which provides focused information and clarifies the issues at stake (Kriesi 2005: 41). The Swiss direct-democratic institutions come in different forms with very different kinds of logics. Very broadly, we can classify direct-democratic institutions according to two criteria: •

the source of a proposition: elite or citizens



the initiation of the vote: required by constitution or launched by citizens

To be sure, there are other criteria to classify direct-democratic institutions, for example the criterion of whether or not the direct-democratic decision is binding. Since all votes are binding in Switzerland, this particular criterion is not pertinent in our present context. Combining the two criteria, we can distinguish between the three basic direct-democratic instruments used at the federal level in Switzerland (Table 1). For each one of these three types, there are different subtypes, but for our purposes it is sufficient to distinguish between the three general categories. According to the source of the proposition, we can first distinguish initiatives from referenda: initiatives are propositions formulated by the citizens, or, more precisely, by organizations representing a circumscribed group of citizens, while referenda are propositions formulated by the government and the majority of Parliament. Initiatives and referenda follow entirely different logics: the initiative puts an issue on the political agenda and launches a debate in the public sphere. By contrast, the referendum only intervenes after the parliament has adopted a given proposition and serves as a sort of popular veto with respect to a decision taken by the political elite.

Table 1: Classification of direct-democratic institutions Required by

Source of proposition government

citizens

Constitution

Compulsory referendum

-

Citizens

Optional referendum

Popular initiative

1 In the case of an initiative, there is a possibility of a more complicated choice, when the governmental coalition formulates a counter-proposal that is submitted to the vote at the same time as the initiative, but such instances are rare.

10 The referendum comes in two basic versions that can be distinguished on the basis of the second criterion: Referenda are either required by the Constitution or by a group of citizens. Amendments of the Swiss Constitution require a popular vote. This is the compulsory referendum. Constitutional amendments are only adopted, if they obtain a double majority – a majority of the people and a majority of the Swiss cantons, i.e. the country’s member states. In addition, large parts of regular legislation are submitted to the optional referendum: in this case, once adopted by parliament, the legislative act passes into law by default, if a referendum is not required by a group of at least 50’000 citizens within three months after its adoption by parliament. If the required amount of signatures has been collected to qualify for a vote, the entire population has to vote on the legislative act, and it only passes into law, if it receives a simple popular majority. In case the legislative act fails to obtain a popular majority, it does not become law. In order to qualify for a vote, an initiative needs to be signed by 100’000 citizens within a period of 18 months. 100’000 signatures constitute a relatively low hurdly, given that they correspond to a bit more than 2 percent of the current number of Swiss citizens. Initiatives are only possible for constitutional amendments, which means that, just as compulsory referenda, they also have to obtain a double majority in the popular vote. Before they are submitted to the population, initiatives are debated in parliament and, with very few exceptions; the parliament recommends their rejection in the popular vote2. Given the binary nature of a direct-democratic choice, in each direct-democratic campaign, there is governmental coalition defending the majority’s position, and a challengers’ coalition defending the opposite position. The choice is always one between the status quo (SQ) and an alternative issue-specific proposal. However, which one of the two coalitions is defending the SQ depends on the type of instrument, which means that the logic of the directdemocratic instruments imposes quite different constraints on the two coalitions. In the case of a referendum, it is the challengers who attempt to preserve the status quo (SQ) by opposing the modification of the constitution or the legislative act adopted by the government’s coalition, i.e. the majority of parliament3. In this case, the challengers lead what Gerber (1999) has called a ‘direct preserving campaign’. The situation of an optional referendum differs in so far from that of a compulsory referendum as the challengers had to mobilize a first time in order to qualify the referendum for the vote. Therefore, in a campaign

2 There are only two exceptional cases, where a majority of the Parliament supported an initiative – the initiative launched by the nationalist right, which demanded that the Swiss national holiday be declared a day without work and which was subsequently accepted in a popular vote in September 1993, and the initiative for Swiss membership in the UN, accepted in a popular vote in March 2002. 3

With one exception, the government has endorsed all the projects adopted by the parliamentary majority and which were subject to the compulsory or optional referendum. The exception concerns the revision of the labor law which the parliamentary majority had modified to such an extent that the government refused to take a position in the subsequent popular vote required by the referendum successfully launched by the left in 1996.

11 on an optional referendum, the majority is certain to confront an intense minority, whereas in the case of a compulsory referendum, the issue may or may not be opposed by an intense minority. Compulsory modifications proposed by the majority of parliament need not necessarily meet with great opposition, since they may concern minor points which have made their way into the Constitution in the past for reasons which are no longer politically relevant today. By contrast, in the case of an initiative, it is the governmental majority who defends the SQ, and the challengers who have qualified the initiative in the first place, are leading a ‘direct modifying campaign’. As Gerber (1999) has argued, campaign hurdles are lower for direct preserving than for direct modifying campaigns, which means that – independently of other factors – challengers mobilizing against a decision submitted to a referendum are in a better starting position than challengers mobilizing in favor of an initiative. This is a consequence of the risk-aversion of the average citizens (see Bowler and Donovan 1998: 43). To understand the strategic situation of the two camps facing each other in a directdemocratic campaign, it is helpful to think of this choice situation in the form of a ‘spatial voting model’, a simplifying analytic device often used by political scientists (see Figure 3a). Such a model represents alternatives as more or less proximate to one another depending on their substantive content. We can think of possible proposals as placed on a continuum on which the proposals become progressively more extreme as compared to the SQ, as one moves from left to right. The most extreme form of an issue-specific proposal (EX) is placed at the right end of the scale. In the case of the asylum law, the example which will serve as an illustration for our argument throughout this paper, the most extreme proposition would have been to close Switzerland for all types of refugees. There are n voters, and they too are arrayed along this continuum, at their ideal points – that is, at the proposal they would like best. For the sake of simplicity we can assume for the moment that, when choosing between any two proposals, all voters prefer the proposal that is closer to their ideal – that is, preferences depend solely on the issue-specific distance between their own position and the proposal submitted to the vote.

12 Figure 3: Schematic diagram of the distribution of opinions on the proposal submitted to a direct-democratic vote

a) effect of issue-positions

SQ

P’

P’’

1

P’’’

m

EX

n

b) +effect of partisan loyalties

SQ

1

P’

P’’

m’

m

m’’

P’’’

EX

n

The proposal passes in the popular vote, when a majority of voters, i.e. n/2+1 voters, decide in its favor. The proposal should, in other words, be close to the preferences of the median voter m. The ‘equilibrium outcome’ would be a proposal corresponding to the preferences of the medium voter. Typically, this is, however, not the proposal offered to the citizen public, given that the governmental coalition (in case of a referendum) or the challengers (in case of an initiative) are in control of the agenda. Instead, in the case of an initiative, the proposal submitted to the vote is typically more radical than the preference of the median voter, i.e. it lies at P’’ or even at P’’’. The challengers launching an initiative are typically from the left, and since the left is notoriously minoritarian in the Swiss political system, its propositions are typically more distant from the SQ than the position of the median voter. In the case of a referendum, there are two likely possibilities: the proposal can be either closer to the SQ than the medium voter (e.g. at P’), or it can be more distant from the SQ than the medium voter (e.g. at P’’ or beyond). Sometimes, in the case of a proposal at P’, the governmental coalition is only capable of agreeing on a rather weak reform proposal, which, from the challengers’ point of view, is too close to the SQ. From their point of view, such a reform is likely to be

13 unacceptable, because it may serve as a pretext to prevent more far-reaching reforms for years to come. Sometimes, the governmental coalition adopts a more far-reaching reform, situated at P’’ or even beyond, which, for some challengers, may go way too far. In the case of routine legislation not subject to the compulsory referendum, the art of the governmental majority consists in formulating proposals in such a way that the instrument of the optional referendum is not used, i.e. to arrive at a compromise that includes all groups ‘capable of launching a referendum’ (Neidhart 1970). In the case of a compulsory vote as well as in the case where it may come to a popular vote, the art of the parliamentary majority is to formulate a legislative reform proposal that provides some significant improvement over the SQ without being too distant from the median voter. In the situation presented in Figure 3a, proposal P’’’ would have no chance of being accepted in a popular vote, while proposal P’’ would pass the test of the popular vote: although it is more extreme than the median voter would be prepared to go, if he could have his or her way, P’’ is closer to his ideal point than SQ. This means that governmental reform coalitions can, in fact, be more extreme than the medium voter and still get their way. This holds even more, when we take into account that voters not only decide on the basis of issue-specific considerations, but also on the basis of partisan loyalties as assumed by most models of voting. Following the ‘unified theory of party competition’, which tries to integrate nonpolicy factors such as party identities into the spatial model (Adams et al. 2005), we can capture the impact of such nonpolicy factors in our framework, too. Assuming that policy preferences are correlated with nonpolicy factors such as partisan identities or trust in government, and assuming that all parties can equally count on such nonpolicy factors, these factors shift the position of the median voter in the direction of the government majority, because by definition, a larger number of voters sides with the government majority with respect to such nonpolicy factors (see Figure 3b). In case the government proposes a reform, this means that the reform proposal can be more extreme than we would expect on purely policy-related grounds; in case the government defends the SQ, this means that the challengers’ reform proposal has to be less extreme. In other words, the nonpolicy factors tend to disadvantage the challengers’ minority in a direct-democratic vote. Finally, the institutional context defines the rules of the game for campaigning. The governmental coalition has some crucial advantages in the Swiss direct-democratic campaigns, because the government controls several aspects of the agenda (Kriesi 2005: 23f): First, the government sets the timetable for the votes. With regard to initiatives in particular, it currently has the possibility to delay a vote for more than three years when it does not present a counter-proposal, and for more than four years when it formulates an explicit counter-proposal. Such delays may serve to take the steam out of a proposal (Delley 1978: 102ff.). Second, the government determines the composition of the proposals to be

14 submitted to the vote on a given occasion; there are up to four voting dates per year and the government sets the agenda in such a way that the proposals to be submitted to the voters are spread throughout the year, and that important proposals are presented individually. The specific configuration of issues submitted to the vote on a given date may have an impact on the coalition formation, the various actors’ strategies: the issues submitted on the same date compete with each other for the attention of the media and the public; to the extent that they are related, the different issues provide an opportunity for divisions within a given coalition, and actors have to decide how to deal with the different issues. Moreover, related issues are likely to give rise to heuristic voting. Third, the government has the right to present its point of view in the ballot pamphlet which is sent to each citizen. The pamphlet also presents the challengers’ point of view, but the government’s position is always presented first and in more detail. The pamphlet is distributed together with the voting material, three to four weeks before the vote. Fourth, although the government does not have the right to spend public funds on publicity, it may still get involved in campaign activities. Thus, during important campaigns, members of the government endorse its position in public speeches held across the country. The members of the government are free to hold public speeches when and where they want to. In addition, a member of the government always gets the opportunity to address the citizens on TV during prime time in order to present the official point of view with regard to the submitted proposals. Summarizing the contextual conditions, the challengers typically face an uphill struggle in a direct-democratic campaign. This is especially true in the case of an initiative, where they lead a direct-modifying campaign, whereas in the case of a referendum, they have at least the advantage of leading a direct-preserving campaign. It is possible that the details of the rules of the game vary from one region (canton) to the next, since the exercise of political rights is governed by cantonal law. Thus, even the right to vote varies from one canton to the other. Although all cantons have universal suffrage now, the voting age varies from one canton to the other, and one of them – Schaffhausen – still has compulsory voting. Moreover, although postal voting is now equally facilitated in all the cantons, unconditional postal voting has been introduced in the last cantons (VS, TI) only in 2005 (Lüchinger et al. 2007 173). Finally, some cantons provide some special facilities for campaigning: Geneva, for example, provides the parties with free public space to present their posters4.

Coalition formation As already pointed out, a referendum campaign intervenes at the end of a policy-making process. This policy-making process, in turn, may only have been the last episode in a more 4

http://www.admin.ch/ch/d/pore/nrw07/pdf/kantonale_normen.pdf

15 extended legislative process that may have given rise to previous direct-democratic campaigns. This means that a referendum campaign does usually not start from scratch, but can count on the experience, and the alliances of previous campaigns. Similarly, an initiative may also intervene in an already ongoing policy-making process, and try to impose a new solution to a problem that has already been the object of extended legislative processes before. The case of the asylum law illustrates this point: there had been no less than three previous votes on asylum law over the last ten years – a vote in April 1987, one in June 1999 (involving two proposals), and one in November 2002. The first two votes concerned referendums by organizations from the left against the toughening of asylum law, the last one an initiative of the conservative right, which attempted to introduce a further turning of the screw into this legislation. In addition, over the same period of time, there had been eight more votes dealing with immigration issues, all of which were concerned with the question of the restrictiveness of immigration policy. A referendum campaign is triggered by the adoption of a legislative act by Parliament. In the case of an optional referendum, some challenger decides to launch a referendum, and starts to collect the required signatures. The first question to be asked about a referendum campaign is, of course, why it has not been possible to reach a viable compromise in Parliament and why the referendum had been launched in the first place. The answer to this question may help to elucidate the governmental coalition’s and the challenger’s goals and strategies. Did the governmental coalition believe that its project was a sufficient compromise to avoid a referendum, or did it consciously take the risk of a referendum? If it consciously took the risk, did it expect to win the referendum, i.e. did its members believe that they were at P’’, or were they not so sure, i.e. did they consider that they might be at P’’’? As for the challengers, did they believe the project was of the type P’’ or P’’’? That is, did they launch the referendum expecting to win, i.e. were they purely instrumental actors, or did they do so with the expectation of losing the vote, i.e. did they have other objectives to pursue? A challenger’s primary objective may not necessarily be to win the vote in the subsequent referendum, i.e. to block the legislative proposal of the parliamentary majority with the help of the people. Alternative objectives are possible. Thus, a challenger may anticipate a defeat and still decide to launch a campaign. He may expect to have an indirect impact on the issue-specific decision-making process: even if he fails to win a majority of the vote, he may still obtain the support of a sizeable minority of voters, signaling to the government that his position has widespread (although not majoritarian) support in the population. This may give him at least a procedural success (he may be coopted into the issue-specific political subsystem), or even a partial substantive success (he may obtain concessions in subsequent issue-specific reforms). A second alternative objective refers to the possibility that a challenger is not primarily concerned with the impact in the public at large, but with that

16 in his own constituency. In this case, he intends to signal to his own constituency, and to potential allies, that he is defending their cause even if the chances of success are virtually inexistent. This objective takes a longer-term view and counts on building up a reputation for defending a given cause, which will pay off in the future (in the next elections, for example). Related to this second alternative objective, and difficult to distinguish from it in practice, is principled, value-oriented action, which does not calculate the costs and benefits of a campaign, but mobilizes independently of such considerations. Value-rational action is action ‘involving a conscious belief in the absolute value of some ethical, aesthetic, religious, or other form of behaviour, entirely for its own sake and independently of any prospects of external success’ (Weber 1966: 115). Value-rational actors feel a moral obligation to do what they do, whatever the consequences. Theirs is a deontological, as opposed to a consequentialist, orientation. This third alternative is difficult to distinguish from the second one, because, indirectly, it may have the same effect of building up a reputation, even if it is not intended to do so. Table 2 summarizes five possible objectives based on the distinction between substantive, procedural and reputational success and the distinction between direct and indirect success.

Table 2: Classification of the actors’ objectives Success

Substantive

procedural

reputational

direct

Full success: win the

-

Building up a

vote indirect

reputation

Partial success:

Partial success:

Value-oriented action

concessions

cooptation

contributing to reputation

The challenger is at a disadvantage because he has to organize a campaign to collect the required number of signatures first. However, if the challenger succeeds in getting these signatures, he has a first-mover advantage in the campaign preceding the vote: Given that optional referendums are voted within a few months of their submission, the organizational capacity the challenger has built up for the purposes of qualifying the referendum can be carried over to the campaign preceding the vote. Moreover, the attention of the media has already been captured by the preceding mobilization for collecting the required signatures. In the case of an initiative, a challenger cannot count on these advantages, however, because of the delay between the campaign for collecting the signatures and the voting campaign, a

17 delay that is due to the government’s right to elaborate a position with respect to a popular initiative. Once the required number of signatures has been collected, the other actors are forced to take a position, too. They have to decide whether or not they want to side with the challenger or with the governmental coalition, and with which intensity they want to participate in the campaign. At this point, the art of the governmental coalition is to prevent the coalition that had adopted the proposal in Parliament from disintegrating. As Trechsel and Sciarini (1998: 118) have shown, the governmental coalition is often subject to significant erosion and fragmentation. Once a signature campaign has succeeded in collecting the required number of signatures for submitting the proposal to the popular vote, the political elite is under pressure to re-appreciate the situation in light of the chances of a popular success of the referendum, and realignment may take place. Kriesi (2005: 54f.) has pointed out that the seeds of any possible subsequent erosion of government support are already planted in Parliament: the more encompassing the parliamentary coalition in support of the legislative proposal, the more coherent the government’s camp will be during the campaign and the less it will be subject to disintegration. Depending on the type of issue and depending on the evaluation of the chances of success, the composition of the governmental coalition is more or less encompassing (see Kriesi 2005: 26-34). There are cases, where all four parties of the Swiss grand coalition government support the parliamentary proposal, and the challenge is launched by one of the minor parties. The four governing parties include the very conservative Swiss People’s Party (SVP), the two parties of the moderate right – the liberal Free Democrats (FDP) and the Christian-Democrats (CVP) – as well as the Social democrats (SP). Typically, however, the governmental coalition splits along the left-right axis with either the left (SP, Greens and some smaller parties) or the conservative right (SVP and some smaller parties) opposing the governmental coalition. In other words, typically, the governmental coalition is either composed of the three major parties of the right (centre-right coalition), or of the left in alliance with the two major parties of the moderate right (centre-left coalition). Less frequently, two of the four major parties join forces in opposing a parliamentary proposal. In the case we are using as an illustration here, asylum law, each one of the two major types of coalitions has already been formed in the past: the referendums of 1987 and 1999 have given rise to a centre-right coalition, while the initiative of the People’s Party in 2002 gave rise to a centre-left coalition. In the vote on the reform of the asylum law, which has taken place in September 2007, the governmental center-right coalition was once again challenged by the left. This policy domain is essentially structured along the cultural dimension of the two-dimensional political space that is characteristic of Western European party systems, with the left and the conservative right being diametrically opposed to each other (Kriesi et al.

18 2006). In each case, the challenger – either the left or the conservative right – is bound to be beaten by the governmental coalition. The question is, of course, why, in this particular case, the moderate right has chosen to join forces with the conservative right to adopt a tough reform of the asylum law. At issue is also the internal coherence of the governmental coalition. Quite untypical, in this case, the governmental centre-right coalition had to face a serious problem of internal cohesion, because a ‘third force’ composed of people close to the governmental centre-right coalition, but opposed to the governmental proposal constituted itself and mobilized quite intensely against the law. Surprisingly, too, this third force was led by an outsider who did not belong to the political, but to the business elite. Moreover, the coherence of both coalitions was taxed by the fact that the referendum on the asylum law had been linked to a parallel referendum on the new law on foreigners. The two proposals on immigration policy were related and put on the voting agenda on the same date, but they were not equally attractive to all of the actors included in the two coalitions. The coalitions facing each other in a direct-democratic vote can always be defined in terms of the party alignment. However, parties are not the only partners of these coalitions. In addition to the parties, various interest associations join the opposing camps. The configuration of associations on both sides is determined by the interests raised by the issue at stake. Economic and social policies as well as related issues typically mobilize the business interest associations and the unions, while policies more related to the cultural dimension of the political space – issues such as immigration, or questions related to cultural liberalism – bring in organizations connected to the new social movements (such as human rights’ groups), religious organizations, or professional associations. The interest groups play a crucial role during the campaigns because they provide the resources for mobilization. Swiss political parties are typically poor organizations, which heavily rely on the financial and logistic support by interest associations for the organization of direct-democratic campaigns. Given that these associations are generally embedded in partisan coalitions, it is very difficult to separate their impact from that of the partisan coalitions that form at the outset of the campaign (Kriesi 2005: 63). The coalition formation is crucial for the outcome of the campaign. The more encompassing and the more cohesive the governmental coalition, the greater its chance of imposing itself, independently of the intensity or the direction of the campaign, and independently of the characteristics of the issue at stake (Kriesi 2005: 65). The question is, of course, what this general finding means. On the one hand, the formation of an encompassing, cohesive coalition is indicative of the strategic capacity of the respective political actors (their salient knowledge, heuristic processing, motivation and resources in terms of finances, personnel and information) (Ganz 2000); on the other hand, it is also indicative of the large pool of

19 support such a coalition may count upon. The larger and the more cohesive the governmental coalition, the larger the number of voters whose predispositions are in line with the cause defended by this coalition, and the more likely it is that the coalition will succeed in linking the predispositions of a majority of voters to its own proposal.

Mobilizing Overall, a campaign can be characterized by its intensity and its direction. Both depend on the strategic decisions by the actors involved, within the constraints imposed on them by the availability of resources, by the issue characteristics, and by the preference distribution in the elite and in the public. The intensity of the campaign depends, first of all, on the extent to which the proposal in question is contested in the political elite. If there is consensus in the political elite with regard to a given proposal, there will be no campaign of any significance at all (Kriesi 2005). If the proposal is contested, the expected success becomes crucial for the intensity of the involvement of strategic actors. As is observed by Przeworski (1986: 55) in the context of democratic transitions, although interested in the outcome of a campaign, a strategic actor only mobilizes, if he can expect a certain amount of success. Strategic actors do not need to mobilize, when their success is guaranteed, and they will not be ready to mobilize when they stand no chance of success. This implies that the intensity of the actors’ participation in the campaign depends on the expected closeness of the outcome of the decision the campaign is designed to influence: the closer the expected outcome, the more intense the mobilization of the actors involved (Cox 1999). This has also been shown for direct-democratic Swiss votes (Kriesi 2005: 57). The closeness of the vote is, of course, not exactly known in advance, since it also depends on the campaign itself. However, based on related issue-specific votes in the past and on the results of public opinion polls, the actors involved may have a pretty good idea about the outcome of the vote in a given directdemocratic campaign. Note, however, that the expected success becomes less important, when we are dealing with principled, value-rational actors. They may mobilize whatever their chances of success. Or, as we have discussed above, they may define their success in terms which have nothing to do with the expected closeness of the outcome of the vote. Indeed, there are intensive minorities who may launch a referendum, and engage in intensive campaigns, even if they know in advance that they will not be able to obtain a majority of the vote. This is indicated by the fact that the relationship between the closeness of the vote and the intensity of the campaign is much weaker for optional than for compulsory referendums and for initiatives (Kriesi 2005: 58). The intensity of the mobilization by a given actor is, of course, constrained by the amount of resources available. Resource-rich coalitions may be able to dominate the campaign and

20 bias the information processing in their own favor. In other words, the direction of the campaign may be rather asymmetric. According to the ‘populist paradox’ (Gerber 1999), direct-democracy has been transformed from a tool of regular citizens to a tool of special interests to the detriment of citizens’ interests. Studies from the US (Gerber 1999) and from Switzerland (Kriesi 2005) show, however, that the relationship between the distribution of resources and the outcome of the vote are not as straightforward as the critics of the ‘populist paradox’ suggest. In addition to the handicaps we have already noted for challengers, they typically also lack funds. However, they have not lost the campaign from the start. A lot depends on the effectiveness of the respective campaigns, and their unanticipated effects. Thus, Kriesi (2005: 66-75) has shown that challengers benefit from intensive campaigns, even if they are outspent by the governmental coalition: spending brings more attention and more awareness to the issue raised by the challenger, which may not only benefit the government’s camp but also the challengers. The more attentive and the better informed the citizens are, the more likely they are to take note of the challenger’s positions and the less likely they are to follow the governing majority unquestioningly. In a similar vein, based on an analysis of American campaigns, Gerber (1999, 2004), Bowler and Donovan (1998) and others conclude that challenger campaigns are more effective than campaigns of incumbents. This conventional wisdom has, however, more recently been put into question by Stratmann (2006) and de Figueiredo et al. (2006). They show the diminishing returns of campaign spending (which partly explains the lesser effect of spending by the government camp that usually greatly outspends challengers), and the dependency of what in our situation would be the government’s spending on the expected outcome of the vote (the closer the expected outcome, the higher the spending of the government’s camp, i.e. the government’s camp especially spends a lot in ‘difficult cases’). Taking into account these two factors (as well as two other, minor ones), de Figueiredo et al. (2006) are able to show that the spending of both sides is equally effective.

Timing To make up their handicap, challengers have a number of choices to make, the first of which concerns the timing of their mobilization effort. Traditionally, direct-democratic campaigns in Switzerland have been rather short. They started around the time when the voting material was distributed, i.e. about a month before the vote. In the more recent past, Swiss campaigns have become longer. One reason is rather technical and consists in the spread of postal voting. Most voters in most cantons have adopted the procedure of postal voting, which means that they can cast their vote immediately after reception of the voting material. In order to be able to influence these early voters, the campaign has to start earlier. Note that there are no legal limits for the temporal extension of the campaigns. Such an extension is

21 also a result of the increasing professionalization of campaigning. For effective political marketing, it is crucial to operate in a long term perspective. As a result, ‘permanent campaigning’ has become a feature of Swiss politics, too. Perron (2007) shows that, compared to their respective opponents, successful challengers (in electoral contests) launch longer and more intense campaigns than unsuccessful challengers. It is important to start early to be able to set the agenda of the campaign. The citizens who originally lack attention for and information about the issue at stake have most to learn in the early phases of the campaign. The more there is to learn, the more important the occasion. This is how Stimson (2004: 129) explains the very important effect of party nominating conventions on the outcome of American presidential elections. His bottom line on the conventions is that they are times of intense political learning: ‘A public that is nearly always tuned out tunes in for a few days, and those are opportunities to learn about people and programs, times to change views’. The same reasoning explains why TV-campaigns between the presidential candidates barely have an impact on the outcome of these elections: the debates occur so late in the process that most voters have already decided, and the audience they attract will consist mainly of loyalists to each side (p. 133). There are also counter-examples, however. Thus, in the 2006 Italian elections, the incumbent Prime Minister, Silvio Berlusconi, launched his campaign very early and his was a very intensive campaign. For a long time, its effectiveness proved to be rather limited, however. Only in the very last phase of the campaign, Berlusconi succeeded in shedding his loserimage. His chances started improving when he finally succeeded in putting the question of the taxes on the agenda and in suggesting that the centre-left coalition was the ‘party of higher taxes’, which apparently allowed him to create uncertainty in an important part of the public and to draw many people into his own camp (Campus 2006). The timing may be influenced by exogenous shocks, i.e. by events which occur in the course of the campaign, and which cannot be influenced by the campaigners, but which can be picked up by them and used to their own advantage. Well known examples of such events include the bombing attack on a suburban train in Madrid during the Spanish election campaign in 2004, and the East German flood and the Iraq-conflict in the German election campaign in 2002. In the Spanish case, the incumbent government lost the elections largely as a result of its mishandling of the communication related to the Madrid attack. In the German case, the incumbent government decided the election in its own favour thanks to the fact that the two events allowed to deflect attention from the economic problems, and to mobilize and integrate part of its own sceptical constituency. Exogenous shocks may also have preceded the campaign. They may have created a ‘general mood’ that lends itself to being exploited by campaigners. A case in point is the referendum against the liberalization of the electricity market, where the challengers benefited from the widely publicized previous

22 policy failures of liberalizing reforms in the UK (train accidents) and California (electricity blackouts), and of spectacular failures in the Swiss economy (the ‘grounding’ of Swissair, among others) (Kriesi et al. 2003).

Targeting Political actors involved in campaigns also have to choose whether to focus on target groups. Target groups can be defined according to political, socioeconomic or geographical criteria. In direct-democratic campaigns concerning constitutional amendments, regional targeting may be very important because of the formal requirement of the double majority for the passage of constitutional amendments: as there are large differences between the Cantons with regard to their size, and since many of the smaller Cantons are rather more conservative than the median Swiss voter, it becomes possible that the modifying campaign may carry a popular majority, but falls short of a majority of the Cantons. In such campaigns, it makes sense for the campaigners to focus on swing Cantons, which may decide the outcome of the vote. More generally, a key distinction is the one between targeting one’s own constituency, and targeting the public at large. Rohrschneider (2002) calls the former strategy ‘mobilizing’, and the latter ‘chasing’. Chasing strategies (as opposed to mobilising strategies) involve the maximisation of the vote share, the focus on unaligned voters, the use of modern technology, the emphasis on leaders, and organizational innovations. A direct-democratic campaign usually involves ‘chasing’, since its goal typically is to gain a majority of the votes. This should apply even to principled actors, who may be above all focusing on the mobilization of their own constituency, but who still are likely to make an attempt to mobilize the public at large, too. The manoeuvering space for ‘chasing’ depends above all on the distribution of the predispositions in the general public, which, in turn, is a function of the characteristics of the issue at stake. Given that campaigns mainly serve to ‘activate existing political predispositions and make them electorally relevant’ (Finkel 1993: 1), the distribution of the predispositions is a crucial determinant of the targeting strategies. In this respect, we should distinguish between general political and issue-specific predispositions (see Table 3). Based on the general political predispositions, we can categorize the public into one’s own constituency, the adversaries’ constituency and the scorekeepers – those nonideological pragmatists who do not belong to either camp. We can define the constituencies on the basis of party identifications. In Switzerland, the proportion of citizens who have more or less close party attachments is comparatively small: it amounts to 36 percent, which compares with a high of 85 percent (Belgium), and a low of 19 percent (Slovenia)(Huber et al. 2005). The

23 weak overall attachment of the Swiss to the parties implies that a majority of the Swiss tends to belong to the group of the nonideological pragmatists, which opens up a lot of opportunities to campaigners. With regard to issue-specific predispositions, we can classify the citizens into those who hold predispositions favorable to the campaigner’s position, those who hold unfavorable predispositions, and those whose issue-specific predispositions are either ambiguous from the campaigner’s point of view, or very weak or non-existent. Predispositions may be ambiguous because one is torn between the positions of the two camps; they may be weak or non-existent, because one is not concerned by the issue, or because one has never thought about it, or both. The campaigner’s core constituency is composed of those who are generally attached to his camp, and who hold favorable issue-specific predispositions. Similarly, the adversary’s core constituency is composed of those generally attached to the opposite camp and holding issue-specific predispositions that are unfavorable to the actor in question. Although general and issue-specific predispositions tend to go together, their correlation need not be very close. To the extent that general and issue-specific predispositions are not aligned, voters are cross-pressured. As the analysis of political reasoning has shown that voters strive for consistency, we may expect that the cross-pressured will try to resolve this situation. Sniderman and Levendusky (2007) suggest that such resolutions will, more often than not, be guided by party loyalty. As implied by the Michigan model’s ‘funnel of causality’, party identification serves as the anchor, and the issue-specific views follow in tow. To the extent that issue-specific predispositions are not clear-cut, voters are ambiguous. The larger the shares of voters with ambiguous and cross-pressured predispositions, the greater the manoeuvering space of the campaigners.

Table 3: Classification of the public according to general and issue-specific predispositions Issue-specific

General political predispositions

predispositions

own constituency

scorekeepers

adversaries

favorable

Core constituency

Favorable

Cross-pressured

scorekeepers

adversaries

Ambiguous/weak or

Ambiguous/weak

Ambiguous/weak

Ambiguous/weak

non-existent

constituency

scorekeepers

adversaries

unfavorable

Cross-pressured

Unfavorable

Core adversaries

constituency

scorekeepers

24

The shares of cross-pressured voters, ambiguous voters, and of voters with weak or nonexistent predispositions, in turn, depend on the issue’s salience and familiarity. The less salient and the less familiar a given issue, the larger the corresponding shares of voters are likely to be. Salient and familiar issues have been politicized in the past and the respective constituencies are, therefore, likely to have formed clear predispositions about these issues and integrated them into their overall political views. Even in the case of a salient and familiar issue, voters may be uninformed at the outset of the campaign, and their predispositions will need to be activated by the campaigners, and connected to the specific proposition submitted to the vote. However, in such a case, issue-specific predispositions are no longer malleable to the same extent as they are in the case of an unfamiliar and non-salient issue. Based on these considerations, we expect the targeting strategies to differ according to the distribution of the predispositions: evidently, the larger one’s core constituency, the less chasing is needed. All the mobilizing actor needs to do, is to target his own constituency in order to inform it about the substantive links between the proposition submitted to the vote and the constituency’s relevant predispositions, and to bring out the constituency’s vote. Inversely, the larger the adversary’s core constituency, the less the mobilizing actor is able to do. Chasing only makes sense for mobilizing the cross-pressured, the ambiguous and the weakly predisposed. The larger their number and/or the closer the expected race, the more sense it makes to target the general public, not only to inform it about the issue at stake, but to enhance the salience of selected aspects of the issue, and to argue in favor of one’s own position.

25 Communicating The study of social movements has shown the usefulness of the concept of the action repertoire of challengers. Social movements in a given context tend to use more or less standardized repertoires of action (Tilly 1978, 1986, 1995). In an analogous way, we propose to introduce the concept of the communication repertoire, which may characteristically be used in a given type of campaign. Thus, in Swiss direct-democratic campaigns, political actors apply a limited set of established communication routines. Direct-democratic campaigns are something, Swiss collective political actors have learnt how to do over the decennies, and they have developed a set of communication strategies which they routinely implement. However, one should allow for the possibility that for specific issues, new actors may enter the campaign, who may lack the experience of the established organizations and who may, therefore, may be more likely to innovate with regard to the communication repertoire, but also more likely to make serious mistakes. Generally, campaigners can reach out to the public either via the media or via their own organization (see Figure 4). In either case, there is a direct and an indirect channel. With respect to their own organizational channels (or the channels of their allies, which may also be open to them), political actors are likely to increasingly rely on direct communications (e.g. by ‘direct mail’) with the members of their constituencies, given that the activists among their members are increasingly rare. To establish direct contact, the organization can use the various communication media available: direct mail, e-mail, SMS, own electronic media, newspapers and magazines, personal contact and so on. The second option is indirect contact, in which the messages are transmitted to members by activists who serve as middlemen in the communications process. In this case, the organization needs to contact only a small number of persons, each of whom can pass the information on to comparatively large number of recipients. To facilitate this kind of contact, a communication network can be established in which communication flows are regularized between the organization and its activists, and between the activists and particular sets of members (Moe 1980: 44). Alternatively, the campaigners can try to reach out to the public via the media. As we have already pointed out, in an audience democracy, the political actors increasingly depend on the media. Directly, they can reach out through paying for ads in the media or for posters in the public sphere. In reaction to negative publicity in the media, political actors learn how to deal with the media in an ever more sophisticated way. That is they try to find ways to reach the public directly without passing through the media (Swanson and Mancini 1996, p. 252): one possibility to do so is ‘the news news’ – popular interview programs or the use of unorthodox TV-channels such as the appearance in music channels on cable-TV Another possibility is political marketing – paid political advertizing. Morris (1999, p. 206f.) believes that citizens like this kind of advertizing: citizens distrust both journalists and advertisements,

26 but they use both sources to be able to check the bias in the respective other source. Paid advertisements are the best way, according to Morris, to get positive reporting. It allows to influence the public that then influences the reporting in the media, since the media have to take into account what the public wishes to see and hear. A final possibility to circumvent the media is the political communication by internet.

Figure 4: Communication repertoire

Communication channel

Organization (own constituency)

Media (general public)

indirectly (via activists)

directly

directly

leaflets stalls’ actions public meetings demonstrations

direct mail e-mail actions SMS-actions own media

paid media ads in press posters

indirectly

earned media press conferences press communiqué public events

Indirectly, political actors can get into the media by producing events (e.g. press conferences, press releases, public assemblies, demonstrations) that the media find attractive, and on which they are going to report and comment (earned media coverage). Some of these events are explicitly staged (so called ‘pseudo-events’) for the purpose of attracting the media’s attention and eventually influencing the voters’ decisions. The media play a crucial role in this process, not only because of their reach but also because of their limited carrying capacity (Hilgartner and Bosk 1988), and their tendency to impose their own logic of selecting and presenting information. In Swiss direct-democratic campaigns, all actors involved stage routine events which attract equally routine media coverage: thus, early in the campaign, all relevant parties organize an assembly of their delegates, where they adopt their official position and voting recommendation for the campaign. These assemblies are typically held on a weekend and are widely publicized in the press. In addition, public debates between protagonists of the

27 two sides are organized all over the country, either by the protagonists themselves, or by the local media. The national public TV-stations also organize highly popular debates on the most contested proposals submitted to votes. Upcoming direct-democratic votes generally focus media-attention on the issues at stake: this has been established by the studies of Höglinger (2006) and Tresch (2007) on the Swiss public debates on abortion and on European integration respectively. We do not test this hypothesis, but take its validity for granted. Paid media coverage constitutes a particularly popular element of the action repertoire of political actors. Since the campaigners do not get access to the electronic media, publicity in the press and on public posters becomes crucial channels for promoting their respective positions. Important sums are spent on this kind of communication. There are no limits with regard to contributions to the campaign by actors not belonging to the government, and contributors are not required to make public the amounts of money they contribute to the campaigns. The communication strategy of a given actor crucially depends on his resources. If paid media provide a very attractive communication channel for campaigners, they are expensive, and access to this channel is limited by the financial resources available to a given coalition. Campaigners lacking funds may depend more on their organization’s own channels. They may also choose to run a more personnel-intensive instead of a capital-intensive campaign. And they may rely more heavily on the production of events with news-value, which earn them media-access. In structuring an advertising campaign in particular, a coalition must anticipate not only the probable strategy of its opponent but also the evolving behavior of the news media (Iyengar and Simon 2000: 162). As in the US, the news media have started to publish ‘ad watch’ reports. The intent behind this new genre of campaign journalism is to deter campaigners from using false, distorted, or exaggerated claims. However, this form of news coverage also has the potential to provide the candidates with considerable free exposure by recirculating the campaign message to a larger audience. An example, which illustrates this type of events is provided by the posters produced by the unions in their challenge against the reform of the handicap insurance: their posters presented the members of government responsible for this reform (which the unions deemed detrimental to the handicapped) as seriously handicapped – in a wheel-chair, as blind, or otherwise disabled.

Appealing to the public We distinguish between two forms of appeals to the public – one indirect and one direct. Although they are not neatly separable, we are treating them as analytically distinct categories. Indirectly, the strategic campaigners may appeal to the public by selectively

28 emphasizing the salience of certain aspects of the issue at stake. In this case, the campaigner does not attempt to influence the public’s predispositions, but he tries to selectively draw the public’s attention to those aspects of the issue which favor his position, given the targeted public’s predispositions. Directly, the actor may use rhetorical strategies to formulate arguments allowing him to shape or modify the predispositions of some targeted part of the public. We shall first discuss indirect appeals, and then turn to rhetorical strategies.

Priming, framing, second-level agenda setting We start from the assumption that information is not a scarce good, but that the scarce factor is given by the attention for particular informations. At any given moment, the attention of the public can only be focused on a limited number of political problems. The public is, however, very sensitive with regard to new information. Viewed from the perspective of the agendasetting approach, such information is always ambivalent, which is why the selection, presentation and interpretation of information by the political elites and the media plays a key role. There is always room for ‘framing’ of political problems. The processing of information, in other words, provides the baseline for the attention management of both the media and the citizen-voters. Accordingly, the struggle for attention among the actors in the political elite constitutes the key element of campaigning, and, more generally, for democratic politics (Burstein 1998: vi), and attention shifts (Baumgartner and Jones 2002; Jones 1994) become a crucial mechanism in campaigning. In political science, Riker (1984, 1986, 1996) has introduced the notion of ‘herestetics’ to refer to the art of agenda manipulation, i.e. the structuring of the choice situation so that the actor wins, regardless of whether or not the other participants are persuaded. Heresthetics is about the structuring of the world so that you can win. At first sight, in the context of an issue-specific campaign, the agenda is all set. But issuespecific political decisions are typically multi-dimensional and often touch on so many relevant aspects that it is usually not possible to take into account all of them. In fact, most decision-makers and the general public in particular only consider a limited number of aspects. People are ‘cognitive misers’, whose attention is highly selective, and who ordinarily prefer heuristics when making their judgements. One simple heuristic is the accessibility heuristic. When asked to evaluate a proposal submitted to the vote, people do not consider everything they know, but they consider what comes to their mind (what is ‘on top of their head’, Zaller (1992)). This means that strategically minded political actors can have a dramatic impact on public debates by shifting the point of reference of the debate from one aspect to the other.

29 In communication science, the strategy of selectively emphasizing certain aspects of an issue has come to be known under various labels: second-level agenda-setting, priming, or framing. First-level agenda-setting is concerned with the salience of a given issue compared to other issues. Second-level agenda-setting by contrast concerns the communicative impact not only tied to the salience of issues, but also to specific aspects (i.e. attributes) of a given issue. Attributes can either be cognitive (position, arguments, related values, attributions) or affective/evaluative (likes-dislikes, preferences, support). Second-level agenda-setting refers to the capacity of political actors, media actors/content, and of the audience to increase levels of importance assigned to attributes of issues (e.g. Kiousis et al. 2006). Attribute agenda-setting posits that campaigning may not just influence what issues are covered in the media, but also how these topics are portrayed and ultimately how they are perceived and interpreted in public opinion. Similarly, priming refers to the determination of the standards (the issues or attributes of issues) that people use to make political choices (Iyengar and Kinder 1987: 63). Since voters’ decisions depend on what issues or aspects of issues are primed, a critical part of every campaign strategy concerns which issues or aspects of issues to prime. Priming occurs when a campaign actor’s emphasis on an issue causes voters to then base their evaluations on that issue. Heresthetics, in essence, is political priming (Druckman and Miller 2004: 502). Similarly, too, framing refers to selecting some aspect of a perceived reality and to making them more salient in a communicating text, ‘in such a way as to promote a particular problem definition, causal interpretation, moral evaluation and/or treatment recommendation’ (Entman 1993: 52). The frames applied in the framing process are key organizing ideas which are used for the interpretation and evaluation of different issues; frames are not identical with issues, but rather resemble belief systems that allow the actor to link different issues to each other and to give coherence to a set of idea elements (Ferree et al. 2002: 105; Dahinden 2006: 87). One final conceptual clarification is in order: as Druckman (2004) points out, the ‘framing effects’ refer to two distinct phenomena. In particular, we should distinguish between ‘equivalence framing effects’ and ‘issue framing effects’. The former refer to effects that occur, when different, but logically equivalent, phrases cause individuals to alter their preferences (Tversky and Kahneman 1981, 1987). This typically involves casting the same information in either a positive or a negative light. ‘Issue framing effects’, by contrast, refer to situations where, by emphasizing a subset of potentially relevant considerations, a speaker leads individuals to focus on these considerations when constructing their opinions. It is ‘issue framing effects’ with which we are concerned here. Druckman (2004: 673) maintains that, contrary to ‘equivalence framing effects’, ‘issue framing effects’ do not violate the

30 preference invariance assumption, since preferences do not change, when substantively different considerations are brought to bear on the issue at hand. Riker (1996) had maintained that ‘we have very little knowledge about the rhetorical content of campaigns, which is, however, their principal feature’. As Druckman and Miller (2004: 502) observe, the political psychological approach, with its focus on issue emphasis has changed this state of affairs. Compelling evidence of priming effects in the course of campaigns comes, for example, from the Canadian election study of 1988. Johnston et al. (1992) show how the free trade agreement between Canada and the US, as a result of the candidates’ and parties rhetorical posturing, came to the forefront of the public agenda. As the campaign progressed, voters’ preferences on the issue increasingly came to influence their vote choice. Rhetoric ‘does play an important role in campaigning, but not just by persuading people. Rhetoric also plays a role – possibly its biggest role – by directing voters towards a specific agenda and considerations surrounding that agenda’ (Johnston et al. 1992: 249). Another example of a priming effect comes from a study by Sniderman et al. (2004). Their survey experiments dealing with opposition to immigrant minorities in Western Europe, focusing on the Netherlands, indicate that what they call ‘situational triggers’, i.e. essentially priming devices, mobilize support beyond the core constituency already predisposed to oppose immigration. Such ‘situational triggers’ point to a mechanism for ‘flash’ politics, which are illustrated by the meteoric rise of Pim Fortuyn in the Dutch elections of 2002. The importance of this phenomenon implies, of course, that the simple one- dimensional model that we have presented in Figure 3 is too simple, indeed. If every issue has multiple aspects, and if it is possible to selectively emphasize the one or the other, we need at least a two-dimensional model. We can illustrate this with the case of the asylum law. In this campaign, two key aspects of the issue became salient – the abuse of the right to asylum by illegitimate refugees, and the humanitarian tradition of Switzerland. Each side emphasized one of these two aspects: the supporters of the law invoked the abuse of the current situation and pleaded for stricter rules to fight abuse, while the challengers maintained that the law violated the Swiss humanitarian tradition. Figure 5, which draws a two-dimensional space with the two aspects of the issue defining the two dimensions, permits us to formalize this particular situation. We can assume that the median voter m is both in favor of somewhat stricter rules to fight abuse, as well as in favor of maintaining the humanitarian tradition of the country. We can equally assume that the proposed reform P’’ went somewhat beyond of what the median voter might have ideally accepted in terms of tightening the rules, and that some of its measures actually violated the humanitarian tradition of the country. In other words, on the second dimension, the proposed reform P’’ was far removed from the ideal point of the medium voter. The interrupted line indicates how the proposed reform P’’ is situated in the two-dimensional space. The medium voter might be ready to accept some

31 other combinations of the two aspects of the law, which is indicated by the indifference curve. She might, for example, be willing to trade off the tighter rules proposed by P’’ against a somewhat reduced emphasis on humanitarian considerations, i.e. she might also be ready to accept the combination P’’ on the horizontal and P’ on the vertical dimension. But note that this combination is still very far off from the combination proposed by the reform. Given this situation, the incentive for the two coalitions are quite clear – the proponents of the reform will put the emphasis on the need to fight abuse, while the opponents will stress the violations of the humanitarian tradition by some aspects of the law.

Figure 5: Example of a two-dimensional issue-space in favor of maintaining humanitarian tradition

m P’

against stricter rules to fight abuse

SQ

m

P’=P’’

in favor of stricter rules to fight abuse

P’’

against maintaining humanitarian tradition

What kind of factors guide the priming and framing strategies of political actors? Riker (1996) has formulated two principles of heresthetics: The ‘dominance principle’ states that ‘when one side has an advantage on an issue, the other side ignores it’. According to the ‘dispersion principle’, both sides seek new and advantageous issues, when neither side has an advantage. As is observed by Sides (2006: 411), this formulation of the two principles begs the question of what provides actors with an advantage on any given issue. In order to answer this question, Sides connects the two principles to the theory of issue ownership which states that the advantage arises from reputations the actors (the parties in the case of this theory) have developed for effective policy making on certain issues. The actors (parties) have what Petrocik (1996) describes as a history of attention, initiative and innovation toward

32 these problems, which leads voters to believe that one of them is more sincere and committed to doing something about them. Thus, ‘issue ownership’ appears to provide parties with the kind of advantage that Riker would describe as ‘dominance’. The crucial mechanism behind this advantage is credibility, which is created by the accumulated historical evidence of the party activities related to the issue in question. With regard to the issue which constitutes the object of a given direct-democratic campaign, some actors may make the credible claim to its ‘ownership’. As Scammel (1999: 729) observes: ‘Reputation, based on record and credible promises, is the only thing of substance that a party can promote to potential voters’. Thus, in the case of the tightening of the asylum law, the SVP could make such a claim, since it has been mobilizing for a tougher asylum law for many years now. Accepting that each issue has different aspects, we can extend the idea of ‘issue ownership’ to the idea of ‘issue attribute ownership’. In other words, not only the SVP can make a claim to have a credible commitment to solving the problem of immigration by tightening the immigration legislation; the left can similarly make a claim to have a credible commitment to defending the immigrants against the tightening of the immigration legislation. To prove the credibility of its commitment, the left had launched the referendum in the first place. In other words, depending on the attribute of the issue that becomes salient, a given issue may be owned by different actors, which is why the dominance principle can also be applied to issue attributes. Sides contrasts this idea that parties (or political actors in general) focus on different issues based on their respective reputations with the idea that instead parties may focus on a common set of issues that are highly salient to the public. He refers to this alternative strategy as ‘riding the wave’. The motivation for such an alternative strategy may be to appear responsive to the public’s concern, regardless of a reputation or claim to ‘ownership’. Actors may thus have an incentive to ‘trespass’ on the opposing camp’s territory. They may do so in two different ways: either by talking about the issue in only the vaguest possible terms, or by focusing on a sub-issue (an attribute of the issue) on which they have an advantage. Sides shows in his analysis of the 1998 American presidential campaign that party ownership’s impact is weak, while ‘trespassing’ in one or the other of these two forms is widespread. Gilland Lutz and Marquis (2006) show that, at least in Swiss direct-democratic campaigns on European policy, the dominance principle does not play a role either. In the case of an issue-specific campaign, the two ‘trespassing’ strategies may, indeed, be combined: an ambiguous trespasser merely may endorse an aspect of the issue that is widely shared in the public without becoming more specific. Thus, the supporters of the asylum law made vague claims that they also were in favor of the humanitarian tradition. They maintained that the law was, in fact, compatible with the humanitarian tradition and that

33 they, the supporters, wished to maintain the humanitarian tradition just as well as the law’s opponents – all they wanted was to stop the illegitimate abuse of this tradition. Referring to Figure 5, they, in fact, maintained that, on the vertical dimension, the law was situated at P’ rather than at P’’. Given that the average voter was unlikely to have enough knowledge about the details of the law, and given that the consequences of the implementation of the law could not be predicted with absolute certainty, the supporters of the law did not risk very much by making such a claim. To meet the challenge of adversaries, political actors may try to avoid certain issues altogether. Avoidance strategies include the displacement of problems, the shifting of the debate to secondary arenas, endless talking without substance, and the transformation of substantive conflicts into moral ones. Such strategies may be adopted as a reaction to the strategies of the media and challengers or as an attempt to avoid issues that risk to split one’s own party, coalition or elite relations. They are not limited to electoral campaigns. In addition to the basic repertoire (‘stone walling’, ‘half-answering’, ‘not-remembering’, ‘disclosing drop by drop” or ‘suddenly and overwhelmingly”) the strategy of displacing problems includes also more targeted strategies attempting to undermine exclusive reports by disliked papers or to dilute the effect of investigative journalists’ research. Direct intimidations of such journalists and complaints about them lodged with their superiors belong to such strategies as does their targeted discrediting (Esser 2000: 22). Among the techniques of issue-avoidance, we also find actor-centered strategies, especially personalizing and negative publicity (‘negative campaigning’). Personalizing strategies allow to distract attention from political issues. Pfetsch (1993: 100) suggests that personalizing strategies used by opponents are quite apt to thwart the intended communication effects of political actors, because such strategies correspond to the selection criteria of the media and are readily picked up. Under what conditions is issue-management successful? Druckman (2004) studied the ‘equivalence framing effects’ and he found that this kind of framing effect is neither robust across political contexts nor particularly pervasive. It is less effective, if individuals are exposed to counter-frames, and if they engage in discussions with a heterogenous group (a group including participants originally exposed to different frames). Moreover, experts (i.e. individuals who have a high need for cognition) are less susceptible to framing effects. In other words, elite competition and heterogenous discussions limit and often eliminate framing effects. Sniderman and Theriault (2004) report similar results. In a direct-democratic campaign, the frames of the one side tend to be contested by the other side, which means that framing-effects may, indeed, be limited. We can expect more far-reaching framing effects in cases, where the campaign is one-sided, i.e. where one camp heavily dominates the campaign.

34 Tanner et al. (2007) studied the impact of framing on people with ‘protected values’. These people correspond to our core constituents: they are principled actors who act out of a moral obligation. Compared to people with consequentialist reasoning, i.e. people very much like our ‘nonideological pragmatists’, the people with ‘protected values’ were much less susceptible to framing effects. Probably, we can expect particularly strong framing effects among the uninvolved, i.e. among individuals who are politically little interested and whose issue-specific awareness is low – precisely those groups that are targeted by campaign chasing the general public. Our hunch is that issue-framing effects also depend on their resonance with the predispositions of the general public. Snow and Benford (1988) argue that resonance is a function of the empirical credibility (does the framing correspond to events in the real world?), the experiential commensurability (does the framing correspond to the everyday experience and the common sense of the average voter?), and the narrative fidelity (does the framing correspond to the existing cultural models?). In other words, the resonance (i.e. the impact) of a frame is expected to be greater, the greater its empirical credibility, experiential commensurability and narrative fidelity. This is why populist frames appealing to the sentiments of anxiety, disenchantment and ressentiment of the ‘common man’ and his allegedly superior common sense have an advantage over more sophisticated frames. This is also why introducing considerations that appeal to widespread stereotypes as well as to the received wisdom in a given society is likely to have a strong effect, while innovative, unconventional or unexpected framing is likely to have little effect at all. One pertinent example is provided by Bartels (2003: 59). In the U.S. General Social Survey 20 to 25 percent of the respondents in the years 1984, 1985, 1986 said that too little was spent on ‘welfare’, while 63 to 65 percent said that too little was being spent on ‘assistance to the poor’. As Bartels observes, this effect not only is spectacularly large but also is very hard to account for in any straightforward logical terms. One way to account for it, however, is to refer to the deeply unpopular connotations of ‘welfare’ for significant segments of the U.S. public. As to the nature of these connotations, Gilens (1999: 154-173) finds that the opposition to ‘welfare’ among white Americans is mainly race-based. While it does not reflect a general racial animosity or an effort to defend whites’ concrete group interests, it seems to reflect the most important non-racial basis of welfare opposition: the perception that welfare recipients are undeserving. The cynicism that white Americans express toward welfare recipients is fed by their stereotypical belief that blacks lack a commitment to work, in combination with their greatly exaggerated ideas about the share of welfare recipients who are African Americans.

35 Persuading: rhetorical strategies The selective emphasis of issues and their attributes is an indirect form of persuasion. As already pointed out, we treat rhetorical strategies attempting directly to influence the positioning of the public as an analytically separate category. Such strategies can be defined as the ‘instrumental use of arguments to persuade others of one’s selfish claims’ (Schimmelfennig 2000: 129). Such strategies can appeal to either systematic or peripheral information processing. As is well known, persuasion is either mediated by detailed processing of arguments, or it works through processes of classical conditioning or mere exposure. Dual-process theories integrate both theories of systematic processing and persuasion processes that are not based on systematic analysis of message arguments (Stroebe 2007). There are two versions of these theories, namely the elaboration likelihood model (ELM) and the heuristic-systematic model (HSM) (Eagly and Chaiken 1993). The likelihood that a recipient will critically evaluate arguments contained in a message (i.e. the elaboration likelihood) is determined by both processing motivation and processing ability. If recipients are motivated and able to process the information, they will engage in systematic processing (the central route to persuasion), if they lack motivation and ability, their opinion will be conditioned by the peripheral route to persuasion. The peripheral route entails cognitive processes such as the use of heuristic decision rules (e.g. ‘the government can be trusted’), affective processes (e.g. classical conditioning or mere exposure), and use of information about the attitude held by relevant others (e.g. ‘party recommendations’). The most important determinant of a person’s motivation to think about the argument contained in a message is its perceived personal relevance. Only if the issue is important (salient) to them personally should recipients of a communication be motivated to critically evaluate the arguments contained in a message. Processing ability, by contrast, is positively influenced by (moderate) argument repetition: repetition should enhance attitude change for messages consisting of strong (high quality) arguments and reduce attitude change for weak (low quality) messages. The HSM assumes that, at high levels of motivation and ability, both processing modes are likely to affect persuasion. In addition to the distinction between argument-based and heuristic strategies, it is useful to distinguish between positive and negative appeals. This leads to the classification of rhetorical appeals presented in Table 4. Arguments are the most elaborated form of appeals, followed by emotions and endorsements.

36 Table 4: Classification of rhetorical appeals General direction

Degree of elaboration Systematic

systematic/heuristic

Heuristic (peripheral)

(central) arguments

emotions

Actor-

Decision-rules

centered Positive

In support of

Excitement/

one’s own

enthusiasm

Endorsing

trust heuristic

Stigmatising

distrust

/

heuristic,

discrediting/

populist anti-

blaming

elitism

position Negative

Against adver-

Anxiety/stress/fear

sary’s position

In his analysis of the ratification campaign for the American Constitution in 1787/88, Riker (1996) observed that much of the arguments exchanged between the Federalists and the Anti-federalists were negative, which he tried to explain by making reference to theories of decision making under risk. To appeal negatively is to point out the danger implicit in the opponent’s program, while not emphasizing the advantages of the campaigner’s own program. Riker’s idea is that ‘an important element of campaigning consists of exploiting voters’ attitudes toward risk’ (p. 66). Speakers emphasize dangers rather than advantages because they believe some voters to be extremely risk averse. The emphasis on negative arguments, he believes, speaks to three groups of voters, and is, therefore attractive to the campaigner: It encourages the uninvolved to think about the choice to be made and to be activated into voting (negative arguments may arouse negative emotions in the group of the uninvolved, which may incite them to resort to systematic thinking about the issue, as is suggested by Marcus et al. 2000); it encourages the serious, but uninformed (part of the scorekeepers) to choose by the criterion of minimizing maximum regret (i.e. to vote against the implied risks of the adversary’s position), and it is also appropriate for the utility maximizers in the group of scorekeepers: negative appeals increase the cost of the adversary’s position, whereas positive appeals are likely to reach only the irrelevant audience already committed to the speaker. Accordingly, Riker (1996: 74) states: ‘Given potential marginal voters who are uncertain or indifferent, rhetors emphasize extreme and objectively improbable dangers in the opponents’ program’.

37 In this respect, we should, however, distinguish between the defenders of the SQ and the reformers (Riker 1996: 68f.). The camp defending the SQ has typically nothing to defend positively. The SQ is visible to all and not subject to transformation by rhetorical reinterpretation. Defenders of the SQ thus may devote almost all of their effort to negative arguments against the reformers. They typically use what Hirschman (1991) has called the rhetoric of reaction, which makes three types of negative arguments against reform: it points out the danger (‘jeopardy’), futility and even perversity of the reform. It does not make a difference, whether the defenders of the SQ are on the right or on the left; the structural setting of the campaign induces them to use this kind of negative rhetorical strategy. Thus, independently of the actor’s ideological position, in the case of the tightening of asylum law, the defenders of the SQ argued that the reform put the ‘humanitarian tradition’ in danger, that it would be inefficient in terms of fighting abuse, and that it might even, perversely, discriminate against certain groups of Swiss. The reformers offer an alternative to the SQ that is not completely understood and is vulnerable, therefore, to deliberate distortion by the defenders of the SQ. Although the reformers’ campaign may be mostly negative, it must also contain some positive elements. The modifying-campaigns of reformers have, on the one hand, to point out the deficiencies of the SQ, and, on the other hand, to point out that the reform proposal constitutes an opportunity to do something about these deficiencies and that the proposed reforms will be effective. They find themselves in a similar situation as a protest movement, that has to provoke a change in consciousness in its constituency. As Piven and Cloward (1977: 3f.) argue, such a change in consciousness involves a loss of legitimacy of the SQ, a sense that change is necessary, as well as a sense of efficacy, i.e. a sense that one can do something about the unsatisfactory state of affairs. Accordingly, the reformers typically use the rhetoric of change (Gamson and Meyer 1996), which also makes three types of arguments: it points out the urgency of reform (i.e. it attacks the deficiencies of the SQ), the opportunity for ‘agency’ (i.e. the window of opportunity), and the ‘new possibilities’ (i.e. the available solution). Thus, the supporters of the tougher asylum law argued that the SQ gave rise to ‘abuse’, they pointed out their willingness to do away with it, and they proposed measures designed to stop the criticized practices. Recall that in the case of an initiative, the challenger promotes reform, while in the case of a referendum, he defends the SQ. Defending the SQ comes rather naturally to the conservatives. By contrast, the left has traditionally championed the initiative and made rather little use of the referendum, because it attempted to modify the SQ and to introduce new advantages for its clientele. More recently, however, faced with retrenchment measures in the domain of social and economic policy, the left has also come to embrace the optional referendum and to defend the SQ, i.e. the achievements of past reforms (Kriesi and Wisler

38 1995:27f.). The left, however, has great difficulty in arguing in favor of the SQ, when the reform package which it attacks is not just a retrenchment of some previous social achievements, but includes some true reform elements, too, as was for example the case with the disability insurance attacked by a referendum launched by the left, and voted upon in June 2007. In such a case, the left needs to explain why the reform elements are not sufficient, and it needs to make a plausible argument that there are elements of the proposal, which would lead to a deterioration of the SQ for some key groups in society. Negativism has also become a selling option for both media owners and editors-in-chief. According to the theory of ‘media malaise’, dominant frames of politics in contemporary journalism are strongly biased towards negativism. They stress struggle rather than compromise, division rather than unity, bad news (about costs, defeats, failures, dangers, crises, sex and scandal, death and disaster, political incompetence and corruption, anything else that is sensational) rather than good news (benefits, triumphs, success-stories, opportunities, solutions). ‘Attack journalism’, said to be more common, encourages political actors to campaign negatively by attacking their opponents, not stating their own case. According to the theory of ‘media malaise’, ‘the combination of bad news, attack journalism and negative politics tends to create a pervasive sense of cynicism, distrust and suspicion of modern politics and politicians’ (Newton 1999: 318). Although the theory concentrates on television (hence video-malaise), the problem does not lie only with television but with all forms of modern mass media, both print and electronic. Negative campaigning in which one side directly attacks or seeks to discredit its opponents is becoming increasingly widespread. As Iyengar and Simon (2000: 161) observe, ‘practitioners generally acknowledge that it is the response of the attacked candidate that is more important. Generally, the attacked candidate is thought to suffer if he or she fails to rebut or otherwise discredit the attack’. Both positive and negative rhetoric may tend to overselling (Lowi 1969: 174-186): the negative version of oversell exaggerates the threat: it ‘is essentially the attempt to create the moral equivalent of war. It is the conversion of interactions into incidents, incidents into challenges, challenges into threats and threats into crises’. Overselling implies the escalation of meanings. Lowi is talking here about U.S. foreign policy, but the rhetorical strategy also works in other issue fields. In U.S. foreign policy, the President can mobilize the public with almost any international act with which he can clearly associate himself. This means that he is always tempted to prefer such actions. The danger of this kind of action is that it may be turned on at will, but cannot so easily be turned off. In its positive version, oversell exaggerates the qualities of the remedy: the danger of this kind of oversell is that when experiments are sold as sure things and specialized solutions are sold as cure-alls, frustration and failure are inevitable. An experiment may be partially successful, but after

39 oversell partial success seems like a failure. Failure leads to distrust and frustration, which lead to more oversell and to further verbal excesses, as superlatives become ordinary through use. The worst possible abuse of oversell, Lowi (1969: 180) maintains, is the rhetoric of victory. ‘It is the last stage before the end of politics’, he writes gloomily – several decades before the time of George W. Bush. In addition to being positive or negative, arguments can be of three types: they can be pragmatic, ethical, or moral arguments (Habermas 1991: 100-118; 1992: 196ff.). Pragmatic arguments justify an actor’s position with reference to the expected output. They refer to the standard of efficiency. In an ethical argument, justification relies on a particular conception of the collective identity and a particular idea of the values (the ‘good life’) represented by a specific community. Moral arguments rely on universal standards of justice, mutually recognized rights, regardless of the utility for a given actor or the specific values or the perceptions of the ‘good life’ embedded in the community in question. The utilitarian argument has much in common with the logic of consequence, while the ethical and moral arguments are related to the logic of appropriateness (March and Olson 1989). Actors of different ideological persuasion are likely to make different use of these three types of arguments: we expect the left tends to argue in universalistic terms, i.e. to use moral arguments more frequently than the right. The governmental coalition, who is always dominated by the moderate right, is expected to privilege pragmatic arguments. The Swiss moderate right argues with its problem-solving capacity and its search for compromise solutions. The conservatives and the new populist right are expected to argue in terms of collective (national) identity and in defense of the community values, i.e. it tends to argue in ethical terms. Just as in the case of issue-framing, the effectiveness of arguments depends on their resonance with the predispositions of the general public and we need not repeat what we have said in the previous section based on Snow and Benford’s (1988) distinctions. In addition, the effectiveness of arguments is also a function of the political sophistication of the public, which, in turn, depends on the familiarity of the issues at stake: the more familiar a project, the greater the effectiveness of argument-based reasoning. Moreover, the intensity of the campaign, i.e. argument repetition has also been shown to increase the effectiveness of arguments (Kriesi 2005: 184ff.). Heuristic appeals may be actor-centred or rule-based. Both can either be negative or positive. Endorsements by prominent and prestigious actors are examples of positive actorcentred appeals; stigmatization or discrediting are examples of negative actor-centred appeals. The SVP regularly uses negative actor-centred appeals by referring to its opponents as ‘softies’ (die ‘Linken und Netten’), or by blaming its political opponents for the alleged crisis. Rule-based heuristic appeals invoke the expertise, competence or

40 trustworthiness of the authorities or, by contrast, criticize the authorities for their lack of trustworthiness. Negative rule-based appeals are typical for populist challengers. As observed by Mény and Surel (2002: 11f.), the common denominator of populist movements puts an emphasis on the fundamental role of the people, claims that the people have been betrayed by those in charge, i.e. the elites are accused of abusing their position of power, and that the primacy of the people has to be restored. The distinction between argument-based and heuristic appeals is, of course, a continuous one. Some arguments are so simplistic as to come close to heuristic decision-rules (e.g. ‘the proposal is too expensive’). We need to develop a measure for the degree of elaboration (the ‘quality’) of an argument. Such a measure could take into account features of the argument such as its length (number of words), whether or not it is supported by examples and statistics (including graphical presentation of statistics), whether or not it is illustrated by pictures, whether or not it gives qualified justifications. A previous analysis of the citizens’ choices in Swiss direct-democratic votes (Kriesi 2005: 184ff.) demonstrates that argument-based voting is heavily determined by context characteristics – the familiarity of the proposal and the intensity of the campaign (which, in turn, contributes to the familiarity of the proposal – especially in the case of originally relatively unfamiliar proposals (p. 97ff.). We can, therefore, assume that rhetoric strategies will less heavily rely on (elaborated) argument-based appeals in campaigns for proposals lacking familiarity and salience, while they will rely more heavily on (elaborated) argumentbased appeals in campaigns for familiar and salient proposals. In addition, the more intense a campaign, the more the strategies on both sides are likely to rely both on argument-based and heuristic appeals. Finally, the effectiveness of rhetorical strategies heavily depends on the position they wish to support. Trent and Friedenberg (2004) observe that successful electoral challengers represent mainstream values, a notion that is confirmed by Perron’s (2007) comparative case studies. Coalitions typically maximize votes by representing policy positions that are similar to, but more moderate than, their partisan constituencies’ beliefs. The reason is that coalitions need to get support from the non-committed pragmatists in between the two camps, as well as from the ambiguous and cross-pressured members of the adversarial constituency. Accordingly, campaigners have to present their own position as a moderate one, while they have an incentive to describe the opponents’ position as extreme and beyond the scope of the mainstream values of society.

41 Conclusion In this paper we have made an attempt to provide a framework for the analysis of the political actors’ strategies in direct-democratic campaigns. This framework is based on three assumptions that need to be tested empirically. One, it assumes that the citizen public normally does not pay much attention to politics and knows little about it, but that large parts of the public are ready to learn about the issues submitted to the vote, once the voting date approaches. Two, our framework assumes that the political actors heavily rely on the media to reach the citizen public, i.e. to get its attention for their messages and eventually to win its support. Three, we assume that the political actors who make the strategic choices are embedded in a context with institutional, issue-specific, and actor-specific aspects, which all are decisive for the choices they are likely to make. Against this background, we have conceptually divided the set of choices into three subsets – concerned with coalition formation, mobilizing, and with the crafting of the message. Mobilizing choices include decisions about timing, targeting and the communication repertoire. Crafting the message refers to decisions about agenda-building (priming, framing) and rhetorical strategies. We have formulated a number of hypotheses about the possible determinants, and the likely effects of the various choices involved in mobilizing and crafting of the message, at the level of the public sphere as well as at the level of the citizen public. Given that little is yet known about the interdependence of these various choices, we have formulated our hypotheses for each one of them separately.

42 Bibliography Adams, James F., Samuel Merrill III, Bernard Grofman 2005. A Unified Theory of Party Competition. A Cross-national Analysis Integrating Spatial and Behavioral Factors. Cambridge University Press. Arceneaux, Kevin 2005. ‘Do Campaigns Help Voters Learn? A Cross-National Analysis’, British Journal of Political Science 36: 159-173. Banducci, Susan A. and Jeffrey A. Karp 2003. ‘How Elections Change the Way Citizens View the Political System: Campaigns, Media Effects and Electoral Outcomes in Comparative Perspective’, British Journal of Political Science 3: 443-467. Bartels, Larry 2003. ‚Democracy with Attitudes’, pp. 48-82 in Electoral Democracy, edited by Michael B. MacKuen and George Rabinowitz. Ann Arbor: The University of Michigan Press. Baumgartner, Frank R. and Bryan D. Jones 2002. Policy Dynamics. Chicago: University of Chicago Press. Bowler, Shaun and Todd Donovan 1998. Demanding choices. Opinion, voting, and direct democracy. Ann Arbor: The University of Michigan Press. Burden, Barry C. 2005. ‘Ralph Nader’s Campaign Strategy in the 2000 U.S. Presidential Election’, American Politics Research 33, 5: 672-699. Burstein, Paul 1985. Discrimination, Jobs, and Politics: The Struggle for Equal Employment Opportunity in the United States in the New Deal. Chicago: University of Chicago Press. Campus, Donatella 2006. ‘The 2006 election: more than ever, a Berlusconi-centered campaign’, Journal of Modern Italian Studies 11, 4: 516-531. Cox, Gary W. 1999. ‘Electoral Rules and the Calculus of Mobilization’, Legislative Studies Quarterly 24(3): 387-420. Dahinden, Urs 2006. Framing. Eine integrative Theorie der Massenkommunikation. Konstanz: UVK-Verlagsgesellschaft. De Figueiredo, John M., Chang Ho Ji, and Thad Kousser 2006. Do Initiative Backers Waste Their Money? Revisiting the Research on Campaign Spending and Direct Democracy, 1st Annual Conference on Empirical Legal Studies Paper, SSRN. Delley, Jean-Daniel, 1978. L'initiative populaire en Suisse - Mythe et réalité de la démocratie directe, Lausanne: L'Age d'Homme. Druckman, James N. 2004. ‘Political Preference Formation: Competition, deliberation, and the (Ir)relevance of Framing Effects’, American Political Science Review 98, 4: 671-686. Druckman, James N. and Joanne M. Miller 2004. ‘The Political Psychology of Electoral Campaigns: Introduction to the Symposium’, Political Psychology 25, 4: 501-506. Eagly, Alice H. and Shelly Chaiken 1993. The Psychology of Attitudes. New York: Harcourt, Brace Jovanovich. Entman, Robert M. 1993. ‘Framing: Toward Clarification of a Fractured Paradigm’, Journal of Communication 43, 4: 51-58. Erikson, Robert S. and Thomas R. Palfrey 2000. ‘Equilibria in Campaign Spending Games: Theory and Data’, American Political Science Review 94, 3: 595-609. Esser, Frank 2000. ‘Spin doctoring. Rüstungsspirale zwischen politischer PR und politischem Journalismus’, Forschungsjournal Neue Soziale Bewegungen, 13 (3), 17-24.

43 Farrell, David M. 1996. ‘Campaign Strategies and Tactics’, in: Lawrence LeDuc, Richard G. Niemi and Pippa Norris (Eds.). Comparing Democracies: Elections and Voting in Global Perspective. Thousand Oaks: Sage, 160-183. Ferree, Myra M., William A. Gamson, Jürgen Gerhards and Dieter Rucht 2002. Shaping Abortion Discourse. Democracy and the Public Sphere in Germany and the United States. Cambridge University Press. Finkel, Steven E. 1993. ‘Re-examining the ‘Minimal Effects’ Model in Recent Presidential Campaigns’, Journal of Politics 55: 1-21. Fournier, Patrick, Fred Cutler, Stuart Soroka and Greg Lyle (2004). Who Responds to Election Campaigns? The Two-Mediator Model Revisited. Montréal: Working Paper. Gamson, William A. 1990. The Strategy of Social Protest. Belmont: Wadsworth. 2nd edition. Gamson, William A. and David S. Meyer 1996. ‘Framing political opportunity’, pp. 275-290 in Comparative Perspectives on Social Movements. Political Opportunities, Mobilizing Structures, and Cultural Framings, edited by Doug McAdam, John D. McCarthy and Mayer N. Zald. Cambridge University Press. Ganz, Marshall 2000. ‚Resources and Resourcefulness: Strategic Capacity in the Unionization of California Agriculture, 11959-1966’, American Journal of Sociology 105, 4: 1003-62. Gelman, Andrew and Gary King 1993. ‘Why are American Presidential Election Campaign Polls So Variable When Votes are So predictable?’, British Journal of Political Science 23: 409-51. Gerber, Elisabeth R. 1999. The Populist Paradox: Interest Group Influence and the Promise of Direct Legislation. Princeton: Princeton University Press. Gerber, Elisabeth R. 2004. ‚Does campaign spending work? Field experiments provide evidence and suggest new theory’, The American Behavioral Scientist 47, 5: 541-574. Gilens, Martin 1999. Why Americans hate welfare. Race, Media, and the politics of antipoverty policy. Chicago: The University of Chicago Press. Gilland Lutz, Karin and Lionel Marquis 2006. ‚Campaigning in a Direct Democracy: Three Case Studies’, Swiss Political Science Review 12, 3: 63-81. Giugni, Marco G. 1998. ‘Was it Worth the Effort? The Outcomes and Consequences of Social Movements’, Annual Review of Sociology 24: 371-393. Habermas, Jürgen 1981. Theorie kommunikativen Handelns, Band 1. Frankfurt: Suhrkamp. Habermas, Jürgen 1991. Erläuterungen zur Diskursethik. Frankfurt: Suhrkamp. Habermas, Jürgen 1992. Faktizität und Geltung. Beiträge zur Diskurstheorie des Rechts und des demokratischen Rechtsstaats. Frankfurt: Suhrkamp. Hallin, Daniel C. and Paolo Mancini 2004. Comparing Media Systems. Three Models of Media and Politics. Cambridge: Cambridge University Press. Hilgartner, Stephen and Charles L. Bosk 1988. ‘The Rise and Fall of Social Problems: A Public Arenas Model’, American Journal of Sociology, 94, 1: 53-78. Hirschman, Albert O. 1991. The Rhetoric of Reaction. Cambridge, Mass: Harvard University Press. Höglinger, Dominic 2006. Drei Welten der politischen Kommunikation? Strukturen politischer Öffentlichkeit im Vergleich anhand der Debatte um den Schwangerschaftsabbruch in der Schweiz, Deutschland und den USA. Lizentiatsarbeit. Dept. of Political Science, University of Zurich.

44 Huber, John D., Georgia Kernell, and Eduardo L. Leoni 2005. ‘Institutional Context, Cognitive Resources and Party Attachments Across Democracies’, Political Analysis 13: 365-386. Iyengar, Shanto and Adam F. Simon 2000. ‘New Perspectives and Evidence on Political Communication and Campaign Effects’, Annual Review of Psychology 31: 149-169. Iyengar, Shanto and Donald R. Kinder 1987. News that Matter. Chicago: The University of Chicago Press. Jasper, James M. 2004. ‘A Strategic Approach to collective Action: Looking for Agency in Social-Movement Choices’, Mobilization 9, 1: 1-16. Jasper, James M. 2006. Getting Your Way. Strategic Dilemmas In The Real World. Chicago: The University of Chicago Press. Johnston, Richard, Andre Blais, Henry E Brady and Jean Cret 1992. Letting the People Decide: Dynamics of a Canadian Election. Standford: Stanford University Press. Jones, Bryan D. and Frank R. Baumgartner 2005. The Politics of Attention. How Government Prioritizes Problems. University of Chicago Press. Kahn, Kim und Patrick J. Kenney 1999. The Spectacle of U.S. Senate Campaigns. Princeton University Press. Kiousis, Spiro, Michael Mitrook, Xu Wu, Trent Seltzer 2006. ‚First- and Second-Level Agenda-Building and Agenda-Setting Effects: Exploring the Linkages Among Candidate News Releases, Media Coverage, and Public Opinion During the 2002 Florida Gubernatorial Election’, Journal of Public Relations Research 18, 3: 265-285. Kollman, Ken 1998. Outside lobbying public opinion and interest group strategies. Princeton, N.J., Princeton University Press. Koopmans, Ruud 2004. ‚Movements and media: Selection processes and evolutionary dynamics in the public sphere’, Theory and Society 33: 367-391. Kriesi, Hanspeter 2005. Direct Democratic Choice: The Swiss Experience. Lanham: Lexington. Kriesi, Hanspeter, Edgar Grande, Romain Lachat, Martin Dolezal, Simon Bornschier, Tim Frey 2006. ‘Transformation of the national political space. Western European politics in the age of denationalization’, European Journal of Political Research 45, 6: 921-957. Kriesi, Hanspeter, Tim Frey, Thomas Milic, Erwin Rüegg 2003. Analyse des Meinungsbildungs- und Entscheidungsprozesses zum EMG. Bern: Bundesamt für Energie BFE. Kriesi, Hanspeter and Dominique Wisler 1995. ‘Social movements and direct democracy in Switzerland’, European Journal of Political Research 30: 19-40. Lachat, Romain and Pascal Sciarini 2002. ‘When do election campaigns matter, and to whom? Results from the 1999 Swiss election panel study’, in: David M. Farrel and Rüdiger Schmitt-Beck (Eds.). Do Political Campaigns Matter? London: Routledge, 41-57. Lau, Richard R., Lee Sigelman, Caroline Heldman and Paul Babitt 1999. ‘The Effectiveness of Negative Political Advertising: A Meta-Analytic Assessment”, American Political Science Review 93, 4: 851-875. Lohmann, S. 1993. ‘A Signalling Model of Informative and Manipulative Political Action.’ American Political Science Review 87, 2: 319-333. Lowi, Theodore J. 1969. The End of Liberalism. Ideology, Policy, and the Crisis of Public Authority. New York: Norton. Lüchinger, Simon, Myra Rosinger and Alois Stutzer 2007. ‘The Impact of Postal Voting on Participation: Evidence for Switzerland’, Swiss Political Science Review 13, 2:167-202.

45 Lykke, Arthur F. 2001. ‘Toward an Understanding of Military Strategy’, in: Joseph R. Cerami & James F. Holcomb (Eds.). U.S. Army War College Guide to Strategy. Carlisle: Strategic Studies Institute of the U.S. Army War College, 179-185. Manin, Bernard 1995. Principes du gouvernement représentatif. Paris: Calman-Lévy. Marcus, George E., W. Russell Neuman and Michael MacKuen 2002. Affective Intelligence and Political Judgments. Chicago: University of Chicago Press. March and Olson 1989. Mazzoleni, Gianpietro and Winfried Schulz 1999. ‘Mediatization of Politics: A Challenge for Democracy?’, Political Communication 16: 247–261. Mény, Yves and Yves Surel 2000. Par le peuple, pour le peuple. Le populisme et les démocraties. Paris: Fayard. Moe, Terry M. 1980. The Organization of Interests. Incentives and the Internal Dynamics of Political Interest Groups. The University of Chicago Press. Morrow, James D. 1994. Game Theory for Political Scientists. Princeton: Princeton University Press. Nadeau, Richard, Neil Nevitte, Elisabeth Gidengil and André Blais 2001. Election Campaigns as Information Campaigns: Who Learns What and with What Effect? Montréal: Working Paper. Neidhart, Leonhard 1970. Plebiszit und pluralitäre Demokratie. Bern: Francke. Newton, Ken 1999. ‘Mass Media Effects: Mobilization or Media Malaise?’, British Journal of Politics 29: 577-599. Oliver, Pamela E. and Gregory M. Maney 2000. ‘Political Processes and Local Newspaper Coverage of Protest Events: From Selection Bias to Triadic Interactions’, American Journal of Sociology 106, 2: 463-505 Perron, Louis 2007. How to overcome the power of incumbency in election campaigns. Unpubl. PhD-thesis, Dept of political science, University of Zurich. Petrocik, John R. 1996. ‘Issue Ownership in Presidential Elections, with a 1980 Case Study’, American Journal of Political Science 40(3): 825-850. Pfetsch, Barbara 1993. ‘Strategien und Gegenstrategien – Politische Kommunikation bei Sachfragen. Eine Fallstudie aus Baden-Württemberg’. S. 45-110 in: Wolfgang Donsbach et al. (Hrsg.), Beziehungsspiele – Medien und Politik in der öffentlichen Diskussion. Fallstudien und Analysen. Gütersloh: Bertelsmann Stiftung. Piven, Frances Fox and Richard A. Cloward 1977. Poor People’s Movements. Why They Succeed, How They Fail. New York: Vintage Books. Przeworski, Adam 1986. ‚Some problems in the study of the transition to democracy’, pp. 4763 in Transition from Authoritarian Rule. Comparative Perspectives, edited by Guillermo O’Donnell, Philippe C. Schmitter, and Laurence Whitehead. Baltimore: The Johns Hopkins University Press. Riker, William H. 1993. ‘Rhetorical Interaction in the Ratification Campaigns’, in: William H. Riker (Ed.). Agenda Formation. Ann Arbor: University of Michigan Press, 81-123. Riker, William H. 1996. The Strategy of Rhetoric. Campaigning for the American Constitution. New Haven: Yale University Press. Riker, William H. 1984. ‘The Heresthetics of Constitution-Making: The Presidency in 1787, with Comments on Determinism and Rational Choice’, American Political Science Review 78: 1-16. Riker, William H. 1986. The Art of Political Manipulation. New Haven: Yale University Press.

46 Rohrschneider, Robert 2002. ‘Mobilizing versus Chasing: how do parties target voters in election campaigns?’, Electoral Studies 21, 3: 367-382. Scammel, Margret 1999. ‘Political Marketing: Lessons for Political Science, Political Studies 47: 718-739. Schimmelfennig, Frank 2000. ‘International socialization in the new Europe: rational action in an institutional environment’, European Journal of International Relations 6, 1: 109-39. Schmitt-Beck, Rüdiger and David M. Farrel 2002. ‘Studying political campaigns and their effects’, in: David M. Farrel and Rüdiger Schmitt-Beck (Eds.). Do Political Campaigns Matter? London: Routledge, 1-21. Schulz, Winfried 1997. Politische Kommunikation theoretische Ansätze und Ergebnisse empirischer Forschung zur Rolle der Massenmedien in der Politik. Opladen, Westdeutscher Verlag. Sides, John 2006. ‘The Origins of Campaign Agendas’, British Journal of Political Science 36: 407-436. Sniderman, Paul A. 2000. “Taking Sides: A Fixed Choice theory of Political Reasoning”, pp. 67-84 in Elements of Reason: Cognition, Choice, and the Bounds of Rationality, edited by Arthur Lupia, Mathw D. McCubbins and Samuel P. Popkin, Cambridge University Press. Sniderman, Paul M. and Matthew Levendusky 2007. ‘An institutional theory of political choice’, in Oxford Handbook of Political Behavior, edited by Russell Dalton and HansDieter Klingemann, forthcoming. Sniderman, Paul M., Louk Hagendoorn, and Markus Prior 2004. ‘Predisposing Factors and Situational Triggers: Exclusionary Reactions to Immigrant Minorities’, American Political Science Review 98, 1: 35-50. Sniderman, Paul M. and Sean Theriault 2004. ‘The Structure of Political Argument and the Logic of Issue Framing’, pp. in Studies in Public Opinion: Attitudes, Nonattitudes, Measurement Error, and Change, edited by Willem E. Saris and Paul M. Sniderman. Princeton: Princeton University Press. Stevenson, Randolph T. and Lynn Vavreck 2000. ‘Does Campaign Length Matter? Testing for Cross-National Effects’, British Journal of Political Science 30: 217-35. Stimson, James A. 2004. Tides of Consent. How Public Opinion Shapes American Politics. Cambridge University Press. Stratmann, Thomas 2006. ‘Contribution Limits and the Effectiveness of Campaign Spending’, Public Choice 129, 3-4: 461-74. Stroebe, Wolfgang 2007. ‘Strategies of attitude and behavior change’, in Klaus Jonas, Wolfgang Stroebe, Miles Hewstone (eds.): Sozialpsychologie, 5. Auflage. Berlin: Springer. Swanson, David L. and Paolo Mancini 1996. ‘Patterns of Modern Electoral Campaigning and Their Consequences’, in: David L. Swanson and Paolo Mancini (Eds.). Politics, Media and Modern Democracy: An International Study of Innovations in Electoral Campaigning and Their Consequences. London: Praeger, 247-276. Tanner, Carmen, Douglas L. Medin and Rumen Iliev 2007. ‘Influence of Deontological vs. Consequentialist Orientations on Act Choices and Framing Effects: When Principles are more Important than Consequences’, Dept of Psychology, University of Zurich. Tellis, Gerard J. 2004. Effective Advertising: Understanding When, How, and Why Advertsising Works. Thousand Oaks: Sage Publications. Tilly, Charles 1995. Popular Contention in Great Britain. 1758-1834. Cambridge, Mass: Harvard University Press.

47 Tilly, Charles 1986. The Contentious French. Cambridge, Mass: The Belknap Press of Harvard University Press. Tilly, Charles 1978. From Mobilization to Revolution. Addison-Wesley. Trechsel, Alexander and Pascal Sciarini 1998. ‘Direct Democracy in Switzerland: Do Elites Matter?’, European Journal of Political Research 33: 99-124. Trent, Judith S. and Robert V. Friedenberg 2004. Political campaign communication: principles and practices. Lanham MD: Rowman & Littlefield. Tresch, Anke 2007. Öffentlichkeit und Sprachenvielfalt. Medienvermittelte Kommunikation zur Europapolitik in der Deutsch- und Westschweiz. Unpubl. PhD-Thesis, Dept of Political Science, University of Zurich. Tversky, Amos and Daniel Kahneman 1981. ‘The Framing of Decisions and the Psychology of Choice’, Science 211: 453-458. Tversky, Amos and Daniel Kahneman 1987. ‘Rational Choice and the Framing of Decisions’, pp. in Rational Choice, edited by Robin M. Hogarth and Melvin W. Reder. Chicago: University of Chicago Press. Walder, Andrew G. 2006. ‘Amiguity and Choice in Political Movements: The Origins of Beijing Red Guard Factionalism’, American Journal of Sociology 112, 3: 710-50. Weber, Max 1966. Wirtschaft und Gesellschaft. Tübingen: Mohr. Wolfsfeld, Gadi 1997. Media and political conflict. News from the Middle East. Cambridge University Press. Zaller, John R. 1992. The nature and origin of public opinion. Cambridge University Press.