Chapter Six
Explaining the Use of Attack Behaviour in the Electoral Battlefield: A Literature Overview
Annemarie S. Walter and Alessandro Nai
Negative campaigning tactics, i.e. attacking the opponent on its issues or traits (Benoit 1999; Geer 2006) have received tremendous amounts of attention from scholars. However, scholarly research has focused largely on the question of what its effects are (see Lau et al. 2007). Notwithstanding the importance of this question, as some scholars (e.g. Ansolabehere and Iyengar 1995) suggest that negative campaigning might be harmful for democracy, i.e. negative campaigning would decrease turnout and increase levels of political cynicism towards politicians and the political system (see Chapter Fourteen), the equally interesting puzzle about why campaigns turn negative has received considerably less attention.
The scholars that do take on the challenge to explain the use of negative campaigning tend to study this phenomenon in the presidential, Senate and House races in the United States (US). As a result, the theory developed on the use of negative campaigning is somewhat limited in scope. However, in recent years (non-US) scholars have started to map and explain negative campaigning in other countries (see for instance Hansen and Pedersen 2008 and Elmelund-Prsestekser 2010 on Denmark; Van Heerde-Hudson 2011 on the United Kingdom (UK); Ridout and Walter 2013 on New Zealand; Walter and Van der Brug 2013 on the Netherlands). Although still in its infancy the study of negative campaigning in non-US context has already provided us with some knowledge on the extent to which negative campaigning is present in other geographical settings and hints to whether the theory developed on the basis of the US case travels. Unfortunately, almost virtually no research has examined negative campaigning from a comparative perspective (a notable exception is Walter et al. 2014), thereby limiting our understanding whether systemic characteristics help to explain the presence of negative campaigning in political campaigns (Fridkin and Kenney 2012).
The work on negative campaigning can be divided in two camps, namely studies developing formal theoretical models focusing on the incentives of the political actor to attack (see e.g. Skaperdas and Grofman 1995) and studies that test prepositions on negative campaigning empirically with a range of statistical techniques (e.g. Lau and Pomper 2004). The empirical studies foremost make use of data collected using content analysis methods on paid publicity (advertisements, posters, leaflets, direct mail, party websites, press releases, twitter messages) and free publicity (newspaper articles, televised election debates). However, some work relies on expert interviews, survey or experimental data of campaign consultants when testing hypotheses on the use of negative campaigning (Swint 1998; Abbe et al. 2001; Theilmann and Wilhite 1998).
This chapter aims to give an overview of the literature explaining the use of negative campaigning. Which candidates/parties are likely to go negative and under which circumstances? This chapter proceeds as follows: we will first discuss the decision-making process preceding a (negative) campaign strategy and then the attacker's characteristics and the target's characteristics that make the
EBSCO : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi
use of negative campaigning more likely. We then reflect on the contextual (election and systemic) factors that are likely to affect the use of negative campaigning, including developments that are changing the overall context in which campaigns are waged nowadays. Finally, we talk about what still needs to be done and introduce the chapters included in this section of the book. By doing so this chapter not only aims to provide an overview of what has been done so far explaining negative campaigning, but also to pinpoint the 'grey areas' in existing literature and suggest avenues for future research.
Negativity as a strategic choice
First we will discuss negative campaigning - who does it and under what conditions? The literature examining these questions generally takes a rational choice perspective and argues that the decision to attack is a strategic one (Riker 1996; Lau and Pomper 2004). Political actors engage in a cost-benefit analysis before deciding whether to attack or not (Lau and Pomper 2004). It is posited that only once the expected benefits outweigh the costs, will negative campaigning be used in an electoral campaign. This decision to go negative is driven by the desire to attain a certain goal, namely to win the elections and negative campaigning can potentially increase a candidate's or party's vote share. Thus, a candidate or party resorts to negative campaigning in an attempt to become the voters' preferred choice by diminishing positive feelings for opposing candidates or parties (Lau et al. 2007; Budesheim et al. 1996). The opposite strategy of negative campaigning is positive campaigning, in which parties engage in acclamation or self-appraisal (Benoit 1999; Geer 2006). Positive campaigning is associated with strengthening a candidate's or party's base and mobilising adherents to cast their vote. Negative campaigning is considered to be a dangerous choice, as one runs the risk that the attacks will generate negative feelings toward the attacker instead of the target (Garramone 1984; Johnson-Cartee and Copeland 1991; Roese and Sande 1993). This risk is the so-called 'backlash' or 'boomerang' effect. Whether or not candidates or parties decide to attack depends on the expected balance between the losses their opponent will suffer from the attack and the risk they face from being perceived negatively.
Recent studies in non-US contexts where the party is the main actor in the political arena (Walter et al. 2014; Ridout and Walter 2013; Hansen and Pedersen 2008; Elmelund-Prastekair 2008, 2010) claim that the party system complicates this decision calculus. Parties in a multiparty system face a different cost-benefit analysis than parties in a two party system. The US literature argues that all candidates are vote seeking, a perspective that is insufficient when examining negative campaigning in a multiparty competition. Str0m and Müller (1999) assume parties to be rational actors whose behaviour is guided by three overarching political objectives: office, policy and votes. Within a two-party system, especially one with majority electoral districts, parties can concentrate on vote-seeking goals, as a majority of votes enables them to achieve office and to implement policy as a result. In contrast, parties operating within a multi-party system have to carefully balance their vote-, office- and policy-seeking objectives. In these systems, obtaining the most parliamentary seats does not automatically translate into government office or influence over policy. As a result, the costs of negative campaigning are higher in a multi-party system than in a two-party system, as parties not only need to worry about possible backlash effects, but also about potential post-election bargaining costs (Walter and Van der Brug 2013; Brants et al. 1982; Hansen and Pedersen 2008; Elmelund-Prsestekser
EBSCO : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi
2010). A campaign, which is fought too aggressively and too negatively, may damage parties' ability to govern together (for example see Sjoblom 1968; Brants et al. 1982; Kaid and Holtz-Bacha 2006). In addition, the rewards of negative campaigning in a multiparty system are less certain than in a two party system; as the number of parties increases it is less likely that the attacker will benefit from an attack (Elmelund-Prssteksr 2008; 2010; Hansen and Pedersen 2008; Ridout and Walter 2013). Most models on vote choice rely at least implicitly on the concept of electoral utility and they assume that voters vote for the party from which they expect the highest utility (Van der Eijk et al. 2006). Voters have a choice set of preferred parties from which they choose at every election (Tillie 1995). As a result of negative campaigning, a voter's expected utility for voting for the attacked party may decline, but that may not result in an increase in the expected utility for the party doing the attacking (Tillie 1995). After the utility for the first preferred party is lowered, these voters are likely to vote for another party in their choice set. Especially in multiparty systems voters prefer or identify with more than one party, as it is likely that there are a number of fairly ideologically similar parties competing (Schmitt 2002; Tillie 1995; Van der Eijk and Niemoller 1983). Thus, parties in a multi-party system face a different cost-benefit structure and as a result have a more difficult task calculating the likelihood of success when going negative, than parties operating in a two-party system.
The decision calculus surrounding the choice whether to go negative in a political campaign is affected by various kinds of factors, respectively the political actors' traits, its opponents' traits and the characteristics of the context in which he operates (both race/election characteristics as well as systemic characteristics). Due to the dominance of single country studies examining negativity in the US context, where candidates are more important than parties, the literature has extensively examined candidate and race characteristics that affect negative campaigning, but not so much the influence of party and systemic characteristics. The more recent work on negative campaigning studying the phenomenon in other geographical contexts than the US has just scratched the surface. Characteristics of the attacker, characteristics of the target and contextual characteristics are insufficient to explain the 'rise' in negative campaigning, we therefore discuss more general developments that are argued to have altered the overall context in which (negative) campaigns are waged nowadays. In the rest of this chapter we will try to systematically discuss the type of factors that affect the candidate's or party's likelihood to go negative.
Characteristics of the attacker
The prime actor in American politics is the candidate, consequently the literature predominantly focuses on candidate' traits when attempting to explain the use of negative campaigning. Some of these characteristics can also be applied to settings where parties rather than candidates are the prime actors. First of all, there is a set of identity features that is argued to affect the likelihood of candidates using negative campaigning. The most prominent one is the gender of the candidate. A growing body of work focuses on topics relating to the differential use of negative campaigning by female and male candidates (e.g. Benze and Declercq 1985; Kahn and Kenney 2004; Lau and Pomper 2004) and the disparity in impact of negative campaigning for female and male candidates (and how this impact in turn affects male and female voters differently) (Brooks 2010; Chang and Hitchon 2004; Gordon et al. 2003; Fridkin et al. 2009; King and McConnell 2003). This body of work suggests that male and female candidates both campaign differently and experience different pay-offs. Female candidates who
EBSCO : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi
employ negative campaigning may find themselves acting at odds with public expectation of their behaviour, as societal stereotypes depict women as kind, helpful, sympathetic and passive in contrast to men who are depicted as independent, forceful and aggressive (Huddy and Terkildsen 1993; Fridkin and Kenney 2009; Kahn 1996). Negative campaigning might be seen as a campaign tactic more suitable for male candidates than female candidates, as negative campaigning is considered rather aggressive. Violating gender stereotypes can have negative electoral consequences, candidates might suffer a substantial backlash as voters tend to reject behaviour that conflicts with gender expectations (Trent and Friedenberg 2008; Kahn 1996). The work on gender and negative campaigning is inconclusive. Some studies support the idea that female candidates are hesitant to go negative due to societal gender stereotypes (Herrnson and Lucas 2006; Johnston and White 1994; Kahn and Kenney 2004), while other studies find no difference (Lau and Pomper 2001b; 2004) or even find that female candidates are more likely to go negative (Bystrom 2006; Evans et al. 2014). Walter (2013) found preliminary evidence that female party leaders were not more likely to go negative than male party leaders in Germany, the UK and the Netherlands, with exception of Conservative Party leader Margaret Thatcher. Thatcher was more prone than her male counterparts to engage in attack behaviour during parliamentary election campaigns. Maier {see Chapter Eight) examines attack behaviour in televised election debates aired in state and national election campaigns in Germany; he finds no evidence for different behaviour between male and female candidates, which is in line with the findings of Walter (2013). One could also argue that not the characteristics of the candidate himself matter, but those of his personnel that design and manage his election campaign. That is why Swint (1998) examines gender differences between male and female political consultants in their attitudes towards negative tactics and actual use. Overall, he finds only small differences in their perception of what constitutes negative campaigning. However, his survey data does suggest that female political consultants are in general less likely to go negative; in particular they are less likely than their male counterparts to attack an opponent on personal grounds.
Another candidate trait that is argued to affect the likelihood that the candidate runs a negative campaign is the candidate's race. Krebs and Holian (2007) suggest on the basis of the 2001 Los Angeles mayoral election that minority candidates are less likely to make use of attack behaviour than non-minority candidates. Candidates whose ethnic or racial status places them in a minority of the electorate must reach beyond their base of support to be able to win the elections, they often do so by waging a deracialised campaign, i.e. a campaign that de-emphasises racial issues and focuses on issues that transcend race. The core of a deracialised campaign is the aim to project a non-threatening image, which might be realised by avoiding confrontation and thus the use of negative campaigning (Krebs and Holian 2007).
Several US studies find that the party affiliation of the candidate matters (e.g. Theilmann and Whilhite 1998; Lau and Pomper 2001b; Druckman et al. 2010; Sigelman and Buell 2003). The work on this link between the candidate's party affiliation and tendency to attack is inconclusive and under-theorised. Some scholars (Theilmann and Wilhite 1998; Lau and Pomper 2001b; Grossmann 2009) argue that Republican candidates are more likely to attack than Democratic candidates, which they link to the findings that consultants who work for Republicans tend to be more open to attack strategies (Theilmann and Wilhite 1998; Francia and Herrnson 2007) and Democratic voters are less likely to approve of negative campaigning than Republican voters (e.g. Ansolabehere and Iyengar
EBSCO : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi
1995). The notion that Republican consultants are more likely to go negative is not supported by all (Swint 1998). Others argue that Democratic candidates are more likely to attack (Druckman et al. 2009 and Sigelman and Buell 2003; Kaid 2004; Benoit 2007; Evans et al. 2014). No similar results are found in the European context {see for instance Elmelund-Prsestekser 2010), with exception of Scammell and Langer (2006) who find that in the UK the Conservative Party is more likely to make use of negative campaigning than the Labour Party and the Liberal Democrats.
Second, a candidate's position in the campaign affects willingness to attack respectively challenger status and competitive standing. The main finding from several US studies (e.g. Lau and Pomper 2004; Druckman et al. 2009; Benoit 2007) is that the challenger status of a candidate matters as challengers are disadvantaged in the race compared to candidates that have an incumbency status. The latter can promote themselves, their work and their accomplishments on the basis of their official positions and duties. Incumbents are also more likely to receive media attention and as a result benefit from name recognition and established reputation (Swint 1998; Kaid and Holtz-Bacha 2006). As challengers do not often have a policy record which can serve as a strong basis for positive campaigning, they are in a position that they have to provide voters with reasons why they should vote for them (Benoit 1999; Kahn and Kenney 2004; Haynes and Rhine 1998; Harrington and Hess 1996; Theilmann and Wilhite 1998; Hale et al. 1996). However, they are also in a position that they do not have an office to loose and are therefore more prone to take a risk and wage a negative campaign. This finding also travels to other contexts such as parliamentary systems where parties are the prime actor and opposition parties are challenging the parties in government. Here it is also the case that opposition parties are disadvantaged and more willing to take a risk and go negative (both in two party systems as well as multiparty systems; e.g. Hansen and Pedersen 2008; Elmelund-Prsestekser 2010; Walter and Van der Brug 2013; Schweitzer 2010). Walter et al. (2014) did not find such an effect in the UK, on the basis of attack behaviour in party election broadcasts between 1980 and 2006, but that is in accordance with the finding of Scammell and Langer (2006) that the Conservative Party as an incumbent party favours negative campaigning.
Another important factor is the competitive standing of the candidate, which most often is operationalised as his standing in the polls (Hale et al. 1996; Skaperdas and Grofman 1995; Kahn and Kenney 2004; Harrington and Hess 1996; Benoit 1999; Swint 1998; Nai and Sciarini 2015). The rationale is that candidates that are trailing behind in the polls or are losing are more willing to bear the risks of a backlash effect than candidates that are frontrunners or gaining in the polls (Damore 2002; Sigelman and Buell 2003; Hale et al. 1996; Skaperdas and Grofman 1995). Although the competitive standing of a political actor is likely to matter mixed results were found beyond the US case. In the UK parties that are losing in the polls tend to make more use of negative campaigning than parties that are gaining in the polls (Walter et al. 2014). The more recent work on multiparty systems has different findings. Elmelund-Prsestekser (2008) shows that parties trailing behind are more likely to go negative, while Hansen and Pedersen (2008) and Elmelund-Prsestekser (2010) only find partial support for Denmark. Walter et al. (2014) did not find such an effect for the Dutch and German multiparty systems. The inconclusive findings might be the result of the fact that operationalising a trailing or losing party is much more complicated in a multiparty system than in a two party system. For the
EBSCO : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi
concept of losing different points of reference can be chosen, namely, the position in the previous polls, the previous election results, or the position of a competing party (Kleinnijenhuis and Takens 2011).
Another indicator of competitive standing is the candidate's campaign budget size. Some studies (e.g. Lau and Pomper 2001b; Johnson-Cartee and Copeland 1991) argue that candidates with relatively fewer campaign sources than their opponents will be more likely to engage in negative campaigning (Lau and Pomper 2001b). These candidates need to win more votes with a smaller budget than their opponent and since negative campaigning is considered to be an efficient campaign tactic to win voters, they would be more willing to bare the risks and likely to attack. However, other studies (Damore 2002; Druckman et al. 2010; Swint 1998) find that candidates with a relatively large budget size in comparison to their opponents are more likely to engage in negative campaigning, as foremost they can afford to go negative, they have the resources to buy the advertisements and hire political consultants to do opponent research and develop advertisements.
Finally, a factor that affects the likelihood of going negative in a multiparty system is the party's coalition potential (e.g. Elmelund-Prsestekser 2008; 2010; Walter and Van der Brug 2013; Walter et al. 2014; Hansen and Pedersen 2008). Parties in multiparty systems waging a negative campaign not only run the risks of backlash effect, but also the risk of alienating potential coalition partners. Consequently parties with low coalition potential have less too lose from negative campaigning, as their chances of being part of the government are slim to none from the beginning. Coalition potential is operationalised differently across various studies, such as the size of the party, years of government experience and the distance to the median party {see Walter and Van der Brug 2013; Elmelund-Prsestekser 2010). In Germany and the Netherlands Walter et al. (2014) find that parties with less government experience are more likely to go negative. They do not find such an effect for government experience in the British two-party system where until recently coalition governments were absent. In addition, they find in Germany and the Netherlands that the further away parties are positioned from the median party, the more likely they are to make use of negative campaigning. In Britain they find the opposite, namely that the more close parties are positioned to the median party, the more likely they are to make use of negative campaigning. The study of Elmelund-Prsestekser (2010) yielded similar results for the Danish case, parties that are positioned in the centre were less likely to go negative.
Characteristics of the target
The decision of whether to make use of negative campaigning is not only affected by characteristics of the attacker, but also by characteristics of the opponent, the potential target of the attack. Most studies modelling the decision to attack pay only limited attention to opponent's characteristics. However, characteristics of the opponent are modelled extensively in a separate strand of studies, which examines the question of whom to attack when going negative in a multi-candidate/party setting. Only a handful of studies have dealt with the decision-making processes of choosing a target when going negative {see Doron and On 1983; Elmelund-Prsestekser 2008; Haynes and Rhine 1998; Ridout and Holland 2010; Sigelman and Shiraev 2002; Skaperdas and Grofman 1995; Walter 2014a; De Nooy and Kleinnijenhuis 2013; Sigelman and Buell 2008; see also Chapters Five and Ten in this book). This is the result of the dominance of studies examining negative campaigning in the US context. The
EBSCO : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi
decision of whom to attack only comes into play with intraparty competition in presidential nomination campaigns and in three candidate races in the US (Walter 2014a). In the US independent and third party candidates have rarely any significant influence on the campaign (a notable exception being Ross Perot in the 1992 presidential race), as a result independent and third party candidates are in the game of negative campaigning often ignored by the major party candidates and thus seldom attacked (Benoit 1999). It follows that the choice of target in a two-party system is relatively clear-cut. Recent work (Walter 2014a; Elmelund-Prsestekser 2008; see also Chapters Five and Ten in this book) has started to examine this question in the setting of multiparty systems, where the choice of whom to attack is much more complex. The choice of target is just as important as the decision to attack when attempting to understand the use of negative campaigning and one could argue that these decisions are intertwined and should be modelled separately. Thus, whether a candidate or party decides to attack is dependent on the characteristics of the opponent he faces.
First of all, the opponent's identity matters. Several studies have shown that male candidates are less likely to go negative if they face a female opponent than a male opponent. These male candidates refrain from attack behaviour as they fear an electoral backlash for 'beating up' a woman (Kahn and Kenney 2004; Benze and Declercq 1985). Societal norms generally hold that aggressive behaviour by men towards women is unacceptable. Thus, female candidates are less targeted than male candidates. Since the number of women running for office is still limited, we have no knowledge about situations in which female candidates have a female opponent.
Second, the position of the other candidate in the race matters, such as incumbency status (Sigelman and Shiraev 2002), competitive position (e.g. Gurian 1996; Haynes and Rhine 1998; Skaperdas and Grofman 1995) and ideological position (Gurian 1996; Doron and On 1983; see also Chapter Ten). Sigelman and Shiraev argue on the basis of the 1996 and 2002 Russian presidential campaign that challengers will primarily aim their attacks at the incumbent. This is in line with the arguments that challengers have to provide reasons to voters why they should be in office instead of the incumbent (Kahn and Kenney 2004; Hale et al. 1996). In addition, the incumbent is the winner of previous election and therefore also potentially the candidate most votes can be won from. Walter (2014a) and De Nooy and Kleinnijenhuis {see Chapter Five) also find that government parties are more frequently targeted than opposition parties in the Dutch multiparty system, as do Doležal, Ennser-Jedenastik and Müller {see Chapter Ten) for the Austrian multiparty system.
Third, no candidate is expected to engage in negative campaigning against the weakest of his opponents - it is more advantageous to defeat a successful rival (Gurian 1996: 8). Negative campaigning will therefore almost always be directed against the so-called front-runner in the campaign, i.e. the largest candidate in the polls, and if the front-runner engages in negative campaigning it is against its strongest opponent (Skaperdas and Grofman 1995). This finding also seems to travel to the context of multiparty systems (Walter 2014a; Bol and Bohl, Chapter Seven; Doležal et al., Chapter Ten). Furthermore, several scholars argue that not just the relative success of rivals matters, but also their ideological proximity (Haynes and Rhine 1998; Gurian 1996). Based on the Downsian notion of relative competition, candidates will try to eliminate the candidates close to their position, as they are competing for the same group of voters. Ridout and Holland argue the opposite, namely that candidates will refrain from attacking an opponent with the same ideological base in presidential nomination campaigns as they do not want to alienate potential voters that are
EBSCO : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi
needed in a later stage of the campaign, they find that Conservative candidates attack Liberal candidates instead of Moderate candidates. A party's position in the polls and its ideological proximity are not two mutually exclusive factors (Gurian 1996). Doron and On (1983) find in four Israeli parliamentary elections partial support for the notion that parties in a multiparty system will attack the largest party close to their position as that is where the most votes can be won. They also point out that parties in a multiparty system often compete in a multidimensional space, so the largest proximate party can be attacked on multiple dimensions. However, Walter (2014a) finds that the effect for ideological proximity is larger for small parties than large parties in the Dutch multiparty system.
In addition, in a multiparty system where parties not only take vote-seeking but also office-seeking considerations into account when deciding whether to attack, parties tend to refrain from attacking potential coalition partners (Elmelund-Prsestekser 2008; Holtz-Bacha and Kaid 2006). Elmelund-Prsestekser (2008) finds exemplary evidence for this in the 2008 Danish election campaign. Walter (2014a) finds mixed results for the Dutch case in the period 1980 to 2006 and suggests that a party's coalition potential is of minor importance in comparison to the other characteristics of the attacked party, or the relationship between the attacker and the attacked party. Doležal et al. {see Chapter Ten) find for Austria that parties that share executive office, i.e. coalition partners are less likely to attack one another than other parties, with exception of parties that are part of a grand coalition. More work needs to be done in multiparty settings to test the relationship between attack behaviour and targeting potential coalition partners.
Finally, the behaviour of the opponent affects the likelihood of using negative campaigning as election campaigns are dynamic. When the opponent engages in negative campaigning, candidates themselves employ more negative campaigning (Lau and Pomper 2001b, 2004; Damore 2002; Krebs and Holian 2007; Haynes and Rhine 1998; Geer 2006). A similar effect is found in the Dutch multiparty system {see De Nooy and Kleinnijenhuis, Chapter Five). Political consultants also prepare for a negative campaign on both sides when they are involved in a campaign against a candidate with a history of negative attack behaviour (Swint 1998). This is likely to be the result of the conventional wisdom among campaign consultants that a candidate must respond to an attack with a counter attack (Haynes and Rhine 1998). This tit-for-tat effect, however, does not seem to apply across party lines in primaries; the negativity of one party's primary is not a good predictor of negativity in the other party's primary (Peterson and Djupe 2005).
Context characteristics
Candidates and parties do not compete in a vacuum, the electoral context affects the calculus candidates and political parties make with respect to negative campaigning (Damore 2002). Context characteristics are features such as the type of election, the competitiveness and openness of the race, the moment in the election campaign and political system characteristics. These context characteristics are often tightly intertwined with one another. As these context characteristics and their relation with negative campaigning have received less attention than the identity of the attacker and the target, we will only reflect on several context characteristics. Nevertheless, we are certain that there are more contextual features that affect the decision calculus of negative campaigning than mentioned in this section.
First of all, the type of race tends to affect candidates and parties likelihood to go negative. To
EBSCO : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi
illustrate, primary campaigns provide a fundamentally different element to the calculation that candidates make. Instead of weighing the benefits and risks in a single election, candidates need to forecast the impact this type of campaigning will have on their general election chances as well. The candidates want to win the primary, but they cannot alienate the supporters of their in-party opponent whose votes they need in order to win in the general election (Peterson and Djupe 2005). As a result, candidates might be more likely to make use of negative tactics towards the end of the election campaign as the number of candidates narrows (Haynes and Rhine 1998). Another example that the type of race matters, in open seat races the use of negative campaigning tends to be higher than in challenger and incumbency races (Kahn and Kenney 1999b, Lau and Pomper 2001b; Damore 2002; Druckman 2010). Often candidates for open seats behave as challengers because they lack the advantages afforded an incumbent (Kahn and Kenney 1999b).
Second, many scholars argue that candidates and parties are more willing to use negative campaigning while jeopardising a potential backlash effect when the race is competitive (e.g. Lau and Pomper 2004; Swint 1998; Damore 2002; Hale et al. 1996; Kahn and Kenney 1999b; Theilmann and Wilhite 1998; Elmelund-Prssteksr 2008; Freedman and Goldstein 2002; Franz et al. 2008; Nai and Sciarini 2015). However, some scholars (Lau and Pomper 2001b; Lau and Pomper 2004; Herrnson and Lucas 2006) do not find a significant relationship between competitiveness and negativity or even the opposite. Francia and Herrnson (2007) find on the basis of a survey of candidates who ran in a variety of races from 1996 to 1998 in the US that candidates in non-competitive races were most likely to endorse negative campaign tactics. According to Mair (1997) two factors are needed for electoral competition, namely the availability of an electoral market and rewards associated with electoral gains. Walter et al. (2014) did not find significant results in Germany, the UK and the Netherlands when testing the first prerequisite of whether the likelihood of going negative increases when the size of the electoral market grows, operationalised as the number of voters that are undecided between two or more parties. However, extensive evidence exists for the second prerequisite, when the rewards increase so does the use of negative campaigning. Haynes and Rhine (1998) show that negative campaigning increases prior to delegate rich contests (such as New Hampshire, Super Tuesday, New York) and important events (Haynes and Rhine 1998). In particular, the electoral rewards are high in close races when both candidates have a chance of winning. US-based research shows that in these close races, the overall negativity from both candidates increases as they are more willing to use negative campaigning while jeopardising a potential backlash effect (e.g. Buell and Sigelman 2008; Damore 2002; Druckman et al. 2009; Kahn and Kenney 2004; Lau and Pomper 2004; Theilmann and Wilhite 1998; Hale et al. 1998; Garamone 1984; Wicks and Soley 2003). The competitiveness of the race is not only measured by the position of the candidates in the polls, but also by the total amount of money spent (Grossmann 2009; Peterson and Djupe 2005), as the higher spending campaigns are expected to be the most hotly contested (Peterson and Djupe 2005). Peterson and Djupe (2005) also measure the competitiveness of the race by looking at the number of quality candidates in the race. A quality candidate is a candidate that previously held office.
Walter et al. (2014) did not find evidence that a close race in a multiparty system, operationalised as a race where the gap between the two largest parties in the polls is small, heightens the overall use
EBSCO : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi
of negative campaigning. In Chapter Seven of this book, Bol and Bohl yield similar findings for the Swiss case; they suggest the reason why studies in Europe fail to reproduce this finding from mainly US studies is the absence of pre-electoral alliances.
Third, the probability of a candidate going negative varies over the course of the campaign (Ridout and Holland 2010; Damore 2002; Haynes and Rhine 1998; Freedman and Goldstein 2002; Nai and Sciarini 2015). Candidates are more willing to take the risk of going negative as the election nears (Damore 2002; Haynes and Rhine 1998). The general idea is that at the beginning of the race candidates mainly make use of positive campaigning as a means to build a base of supporters and then make use of negative campaigning to contrast themselves from other candidates (Peterson and Djupe 2005; Elmelund-Prsestekser 2011). Doležal et al. {see Chapter Ten) and Bol and Bohl {see Chapter Seven) fail to find that negative campaigning heightens at the end of the campaign.
Fourth, political system characteristics affect the likelihood that candidates or parties go negative. Four general characteristics of the political system affect the nature of election campaigns: the institutional structure, the electoral system, the party system and the media system (Bowler and Farrell 1992; Farrell 2005). The few studies paying attention to the influence of political system characteristics on negative campaigning has primarily focused on the party system (Walter 2014b; Ridout and Walter 2013; Walter et al. 2014; Elmelund-Prsestekser and Svensson 2014; Elmelund-Prsestekser 2010). As explained in the paragraph on negative campaigning as a strategic choice the party system affects the decision calculus, the cost benefit structure of negative tactics is different in two party and multiparty systems. The decision regarding the use of negative campaigning is considerably complicated in a multiparty system in comparison to a two-party system, as the objectives that parties strive for do not coincide in a multiparty system. Where in a two-party system parties that win the votes at the election, will also get the office and thus the opportunity to implement policy, in a multiparty system this is not necessarily the case. In a multiparty system parties can become part of the coalition government that lost at the elections and parties that gained can be excluded. As a result the costs of negative campaigning are higher in a multiparty system as it might jeopardise a party's chance of being part of the coalition government and the rewards are lower as a third party might win the voters that are put off by attacking your opponent. Walter (2014b) indeed finds that the level of negative campaigning is significantly higher in the UK than in Germany and the Netherlands. Ridout and Walter (2013) test the proposition in New Zealand. New Zealand exchanged single member first past the post for a mixed member proportional electoral system. As a result of the change in electoral system, the country witnessed party system change (Mair 2006), the almost ideal two party system became a multiparty system. The overall use of negative campaigning in advertising was higher in New Zealand when it had a two party system than a multiparty system. However, some scholars (Salmond 2011; Bohl and Bol, Chapter Seven) argue that it is not the party system, but the electoral system that matters; it is the difference between winner takes all single member district systems and proportional representation. PR generally coincides with a multiparty system, however there are exceptions to Duverger's law. Ridout and Walter (2013) try to make a convincing argument as to why it is the party system and not the electoral system that impacts the use of negative campaigning, nevertheless more work needs to be done to filter out what system characteristics affect negative campaigning and in what way.
EBSCO : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi
144
Recent developments and the rise in negative campaigning
The decision calculus of candidates and parties of whether to make use of negative campaigning is affected by the context they operate in. In the previous paragraph we discussed context characteristics that cause the use of negative campaigning to fluctuate between different kinds of races/elections over the course of the campaign, over campaigns across time and geographical contexts. However, some scholars (e.g. Benoit 1999; Geer 2006; Abbe et al. 2001; Fowler and Ridout 2010) do not see the use of negative campaigning as a tactic which fluctuates from election to election; they argue that negative campaigning is on the rise. This rise in negative campaigning is considered a response to developments that are permanently changing the context in which negative campaigning takes place nowadays. Studies show that four general developments are linked to negative campaigning, respectively the professionalisation of election campaigns, the increasing electoral volatility, increasing polarisation of the party system and the mediatisation of politics {see Geer 2012; Walter 2014b). These developments are not independent of one another. In the next paragraph we will discuss these developments, their link with negative campaigning and assess the claim that negative campaigning is on the rise.
First of all, the use of negative campaigning is linked to the increasing professionalisation of election campaigns (e.g. Geer 2012; Walter 2014b). Scholars often use the term 'professionalisation' interchangeably with 'modernisation' and 'Americanisation' and they tend to use these terms to refer to a range of campaign practices {see e.g. Swanson and Mancini 1996; Strombäck 2007; Plasser 2000). These terms all essentially describe a process whereby many tasks that were formerly ascribed to party members are instead given over to outside agencies (Lilleker and Negrine 2002). This development whereby campaigns are increasingly run by professionals (sometimes even US agencies) is likely to stimulate the use of negative campaigning. Despite the fact that substantive evidence is lacking, campaign consultants increasingly believe that negative campaigning works and regard it as a key to electoral success (Iyengar 2011). A candidate who wants to win the election cannot refrain from attacking their opponent (Perloff and Kinsey 1992; Lau and Pomper 2004; Lau and Sigelman 2000). As a result consultants would encourage candidates to go negative. Abbe et al. (2001) put this relationship between campaign professionalism and negative campaigning to the test. They find that a fully professional campaign that employs political professionals to perform all major campaign activities (such as campaign management, media advertising, PR, polling, issue and opposition research, legal advice and fund raising) is 19 per cent more likely to go negative in US House races. Francia and Herrnson (2007) demonstrate that candidates who wage 'professional' campaigns (those who employ at least one paid consultant) are more receptive to using negative campaign tactics and profess a greater willingness to attack an opponent on issues concerning an opponent's professional conduct or legal infractions. Over time tests and examinations of this relationship in non-US contexts are still absent.
The second development that is linked to the rise of negative campaigning is the increased electoral volatility that political parties similar to US candidates are witnessing, although the level of electoral volatility varies across countries (Walter 2014b). As voters have loosened their ties with parties, leaving more votes potentially 'up for grabs', parties have become more inclined to run an offensive campaign (Andeweg and Irwin 2009; Mair et al. 2004). Negative campaigning as a campaign practice fits better with an offensive campaign than a defensive campaign. The first is aimed at volatile voters and the opponent's adherents and the latter at mobilising a party's own adherents.
EBSCO : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi
This link between electoral volatility and the use of negative campaigning is indirectly assessed in the study of Walter et al. (2014) who assess the relationship between the size of the electoral market operationalised as the number of undecided voters and the use of negative campaigning in Germany, the UK and the Netherlands. They do not find evidence for this relationship, however, this study tests this relationship only since 1980 and the instability of the electoral market originated already decades earlier.
Third, the 'so-called' rise in negative campaigning is associated with an increasingly polarised party system (Geer 2006, 2012; Elmelund-PrEestekser and Svensson 2014). They argue that when the ideological span in the party system widens, more differences between political parties arise, consequently political conflict increases and thus the use of negative campaigning. According to Geer (2006) candidates in the US tend to disagree more on policy than they did thirty years ago. He finds a strong correlation between polarisation and negative campaigning (r=0.88 in Geer 2006), but polarisation affects primarily issue negativity and not so much trait negativity. Ansolabehere and Iyengar (1995: 113) argue the opposite to Geer (2006) and Elmelund-PrEestekser and Svensson (2014), they reason that negative campaigning fuels polarisation in American politics. Attack behaviour would drive the independent voter from the polls and thus generate a more partisan and hence more polarised electorate. The whole notion that polarisation and negative campaigning are strongly linked is contested by Buell and Sigelman (2008). They claim that party polarisation might not be that important as widely thought. They found that even though voting choices in 2000 reflected the highest degree of polarisation up to that point in American recorded history, the overall negativity of the 2000 contest ranked lower than that of any other race in our study except the one in 1976.
Finally, the use of negative campaigning is linked to what some call mediatisation (e.g. Geer 2012; Walter 2014b). Nowadays, parties increasingly have to compete with one another and with other actors in the public domain for the attention of the mass media. To increase the odds of being considered newsworthy it is likely that parties adapt their communication to the mass media's standards. Conflict is one of those criteria used for news selection and the practice of negative campaigning can be regarded as a way for parties to create conflict (Walter and Vliegenthart 2010; see also Chapter Sixteen). Thus the growing role of the mass media may be stimulating the use of negative campaigning. This process by which political institutions are increasingly dependent on and shaped by mass media is called mediatisation (Mazzoleni and Schulz 1999). The empirical evidence is somewhat mixed as a study of the 1992 Democratic nomination found that races that get more media attention are more likely to have a high level of negative campaigning (Haynes and Rhine 1998), whereas primary campaigns from 1998 with more media attention were no more likely to be negative (Peterson and Djupe 2005). Ridout and Smith (2008) find that negative advertisements are more likely to be covered than positive advertisements. Geer (2012) reports that 75 per cent of the stories on the nightly networks are about negative advertisements, and that this percentage has increased over time. Moreover, political consultants do point out that an attack can provide the attacking candidate the needed media attention (Swint 1998). The link between mediatisation and negative campaigning is scarcely explored.
Therefore, when trying to explain the use of negative campaigning it is important to assess the claim that negative campaigning is on the rise (e.g. Benoit 1999; Geer 2006, 2012; Abbe et al. 2001). As last decade's electoral volatility, campaign professionalism and mediatisation have increased, altering the context in which candidates and parties wage campaigns, consequently, one can expect a
EBSCO : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi
rise in negative campaigning. If no such rise in negative campaigning exists, these developments cannot be the (sole) explanatory factor(s) for the use of negative campaigning. So, is there empirical evidence for this claim of an ongoing increase in negative campaigning? We searched for empirical studies examining the so-called 'rise' in negativity. We selected all studies based on data from more than two election campaigns, as three points in time is generally the minimal requirement to be able speak of a trend. We do not claim that this selection is exhaustive; it is likely that there are more studies, in particular studies written in a language unknown to the authors of this chapter. Table 6.1 provides an overview of the found studies examining the rise in negativity. In total we found twenty-two studies examining the use of negative campaigning, of which nine studies focused on the US. All studies measured negative campaigning on the basis of content analysis of campaign sources, the majority focusing on television advertisements. The studies differ in the coding method used, selection of data and weighting procedure, which makes them not directly comparable. Overall, the studies draw mixed conclusions, thereby providing not much support for the claim that the use of negative campaigning is growing. However, six out of the eight studies examining the US report an upward trend supporting the claim that negative campaigning is increasing over time. The studies examining negative campaigning in other countries barely provide evidence for this claim. The only exceptions are Sullivan (2008) who argues that negativity might be on the rise in Taiwanese presidential elections and Scammell and Langer (2006) who argue there is no rise in negativity in the UK in general, but the Conservative Party increasingly makes use of this tactic. Holtz-Bacha (2001) and Schweitzer (2011) even provide empirical evidence for a downward trend in negativity in Germany. In addition, Samaras and Papathanassopoulos (2006) argue there is a decrease instead of increase in the use of negativity in Greek political advertising. Thus, for now the state of the art seems to suggest that the use of negative campaigning is not increasing, but that its level fluctuates from election to election, with exception of the US. As a result, negative campaigning is likely not (solely) the result of these general developments.
Conclusion and avenues for further research
This chapter set out to give an overview of the literature explaining the use of negative campaigning. Although several literature overviews on the effects of negative campaigning exist (e.g. Lau et al. 2007; Fridkin and Kenney 2012), not much attention has been given to the decision to go negative. To our best abilities we have reflected on the most important features mentioned in studies, trying to explain which candidates and parties are likely to make use of negative campaigning and the circumstances under which they do so. Not an easy task as this part of the field is transforming as we speak. However, the decision to go negative seems to be foremost a strategic choice, one that is affected by characteristics of the attacker, characteristics of the target and contextual characteristics. All these characteristics are part of the complex decision calculus of candidates/parties when they decide whether the benefits outweigh the risks of negative campaigning. If this is the case they will decide to run a negative campaign. In addition, when summing up the studies that assessed the claim of a rise in negative campaign tactics, we lacked convincing evidence. Some evidence supporting this claim has been found, although only for the US. A finding that might relieve critics of negative campaigning that are concerned about its effects on the electorate and the political system {see Chapter Fourteen).
EBSCO : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi
Table 6.1: Overview studies examining level of negative campaigning across elections
EBSCO : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi
Study
Country
Election
Source
Trend
l.Benoit 1999
United States
1952-96
United States 1960-2008
2. Buell and Sigelman 2008
3. Druckman et al. 2010  United States 2002-06
4. Fowler and Ridout2010
5. Geer2006
6. Hákansson 1999
7. Haigron 2012
United States
United States
Sweden
2000-12
1960-2004
1948-98
United Kingdom 2001-10
8. Holtz-Bacha2001 Germany
1957-98
9. Jamieson et ah 2000 United States 1952-96
10. Kaid and Johnston United States 1952-96 2001
11. Lau and Pomper United States       1992-2002 2004
12. Elmelund- Denmark 1971-2011 Praestekasr and
Svensson 2014
13. Ridout and Walter New Zealand       1969-2011 2013
14. Samaras and Greece 1993-2000 Papathanassopoulos
2006
15. Sc amine 11 and Langer 2006
16. Schweitzer 2011 Germany
United Kingdom 1992-2001 2002-09
17. Sullivan 2008
18.Toros 2013
Taiwan
Turkey
1996-2004
1983-2011
Television advertisements
Newspaper articles
Campaign websites
Television advertisements
Television advertisements
Party manifestos, election debates
Television advertisements
Television advertisements
Television advertisements
Television advertisements
Newspaper articles
Letters, television advertisements, election debates
Television advertisements
Television advertisements
Television advertisements
Online news releases, party websites
Television and
newspaper
advertisements
Newspaper articles
Upward
None
Upward
Upward
Upward
None
None
Downward
Upward
Upward
None
None
None
Downward
None* Downward
Upward
None
EBSC0 : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi
149
	Study	Country	Election	Source	Trend
	19. Van Heerde-Hudson	United Kingdom	1964-2005	Television	None
	2011			advertisements	
	20. Walter 2014b	Germany	1980-2006	Television	None
		Netherlands		advertisements,	
		United Kingdom		election debates	
1—' -M -C CTI	21. Walter and Van der	Netherlands	1981-2010	Television	None
■rH i_ >~ CL O	Brug2013			advertisements	
OJ i—1 -O	22. West 2005	United States	1952-2004	Television	Upward
f3 -rH 1-1 CL				advertisements	
Note: All studies examining the US are presidential elections, with exception of Fowler and Ridout (2010) who study advertisements from Congress and Senate elections, Lau and Pomper (2004) who study advertisements from Senate elections and Druckman et al. 2010.
* With exception of the Conservative Party, their advertisements showed no increase in negativity.
EBSCO : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi
150
The overview illustrates the width of the field, the range of features that seem to impact the likelihood of running a negative campaign. However, it also shows its inconclusiveness on many aspects. There is need for research that not only looks for new theories, undefined explanatory variables, but that foremost assesses the current findings. The mixed findings in the field are most likely to be (partially) the result of different methodological choices made, one of them being the source of data collection. This issue also becomes more apparent with the more recent work in non-US settings and a handful comparative studies that is often not based on television advertisements, as these do not play a prominent role in the campaigns of these countries or they are subject to all kinds of government restrictions. Elmelund-PrEestekser (2010) and Walter and Vliegenthart (2010) point at the importance of measuring negative campaigning over various campaign sources as the level, content and targets of negative campaigning significantly varies across them. Therefore, patterns of attack behaviour that are found should be validated on a variety of campaign sources.
Non-US studies on the use of negative campaigning will bring us closer to a general theory on negative campaigning, one that for instance also covers campaigns in multiparty settings, PR electoral systems or less established democracies. Several chapters in this section already work on this. Chapter Seven looks at negative campaigning in federal and cantonal election campaigns in the Swiss PR system, Chapter Eight examines negative campaigning in national and state level elections in the German multiparty system, Chapter Ten examines negative campaigning in parliamentary election campaigns the Austrian multiparty system, Chapter Eleven studies attack behaviour in two round presidential campaigns in the Brazilian multiparty system and finally, Chapter Twelve explores negative campaigning in national and local election campaigns in the Turkish multiparty democracy. Chapter Nine reflects on a different matter, namely negative campaigning in direct-democratic campaigns, a type of campaign that can happen in most countries, but is particularly an incremental part of Swiss politics.
Many findings in the field are based on studies that examine single or few election campaigns, comparative work can help to advance present theory. A larger number of elections across time and space would ascertain that the results are not idiosyncratic to the time and place of examination, for example during the recent US presidential election campaigns. In addition, this would allow us to understand better which contextual characteristics affect the use of negative campaigning. We would be able to assess the relationship between electoral volatility, party system polarisation, mediatisation and campaign professionalism with a dataset covering election campaigns for a considerable number of years. In addition, with a large dataset covering numerous countries we would be able to tease out the exact role of systemic variables, such as institutional rules, the media system, political system and electoral system. Research on the influence of systemic variables on the use of negative campaigning can also shine more light on why the US might be a peculiar case when it comes to attack behaviour and why negative campaigning is possibly solely on the rise in the US.
Finally, the work on the strategic use of negative campaigning would benefit from a more interdisciplinary approach. Most work on negative campaigning stems from the field of political science and communication science. Nevertheless, research on negative campaigning can benefit from theories and methods developed in economics, psychology, sociology and linguistics. Chapter Thirteen exemplifies the fruits of a more interdisciplinary approach. This study tests whether people that score high on certain personality are more likely to make use of attack behaviour. An approach that might
EBSCO : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi
help us move beyond the more clear visible candidate characteristics, such as gender, race and party affiliation, and their relationship with negative campaigning.
However, foremost it is clear that still much needs to be done if we want to solve the puzzle on the use of negative campaigning and come to a general theory on negativity. This section hopes to contribute to this endeavour.
EBSCO : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi
152
Chapter Seven
Negative Campaigning in Proportional Representation (Yet Non-Coalition) Systems: Evidence from Switzerland
Damien Bol and Marian Bohl
During the last twenty years, many scholars have studied the causes of negative campaigning in American politics (for a full review, see Walter and Nai, Chapter Six in this volume). Unlike traditional party competition theorists, such as Downs (1957) or Str0m (1990), who assume that campaigns are bare channels for candidates and parties to communicate to voters about their own policy position, they argue that the reality of electoral campaign is also full of incentives pushing these candidates and parties to discredit their opponents (Geer 2006).
Along these lines, negative campaigning scholars working on American politics state that one of the key factors explaining the decision to go negative is the state of electoral competition. For example, various theoretical as well as empirical studies show that the closer the electoral race between candidates, the more these candidates attack each other (Lau and Pomper 2001b; Skaperdas and Grofman 1995). However, this argument originally developed to fit the United States' (US) context, is said to be of little relevance to explain negative campaigning in European proportional representation (PR) democracies. Recent evidence reveals that classic theories only poorly fit the reality of electoral campaigns in this context (Elmelund-Prsestekser 2008, 2010; Walter etal. 2014).
In this chapter, however, we argue that the impact of the electoral competition on the decision of parties and candidates to go negative in PR democracies should not be discarded too quickly. The existence of some form of pre-electoral coalition agreements, and the necessity to effectively bargain with other parties to reach coalition agreements after election day, which are the norms in these countries, is likely to bring noise to the empirical tests of theoretical models. To bring new insight on the topic, we offer a test of the classic American-based theories of negative campaigning in four electoral campaigns in Switzerland: The 2011 Zurich federal and cantonal campaigns, and the 2011 Lucerne federal and cantonal campaigns.1 Unlike other PR democracies, the Swiss federation and cantons are not entirely parliamentary governmental systems. The influence of coalition bargaining on the composition of the executive is therefore limited. In the next sections, we first review the literature on the electoral determinants of negative campaigning; second, we describe the nature of electoral competition in Switzerland and more specifically in the four elections covered; third, we explain carefully our data collection; and last, we report our findings and the implications following from them.
The electoral determinants of negative campaigning
In the US, a growing body of evidence has formed that electoral competition influences the tone of campaigns. In the mid-1990s, Skaperdas and Grofman (1995), followed by Harrington and Hess (1996), elaborated a theoretical model of negative advertising in which electoral competition plays a
EBSCO : eBook Collection (EBSCOhost) - printed on 10/19/2018 7:44 AM via MASARYKOVA UNIVERZITA
AN: 1360081 ; Nai, Alessan, Walter, Annemarie.; New Perspectives on Negative Campaigning : Why Attack Politics Matters Account: s8431878.main.bsi