SEARCH

SEARCH BY CITATION

Keywords:

  • negative campaigning;
  • professionalisation;
  • personalisation;
  • election campaigns;
  • multiparty system

Abstract

  1. Top of page
  2. Abstract
  3. Defining Negative Campaigning
  4. The ‘Rise’ of Negative Campaigning
  5. Party System and Negative Campaigning
  6. Case Selection, Data and Coding Procedure
  7. Results
  8. Concluding Remarks
  9. References
  10. Biography
  11. Supporting Information

This article describes how political parties in parliamentary election campaigns in Western Europe make use of negative campaigning and examines whether their behaviour differs from that of candidates competing in US presidential election campaigns. Furthermore, it theorises how the differences and similarities between negative campaigning in these countries can be explained. First of all, this comparative study adds to the development of a more general theory on negative campaigning. Second, the study presents interesting new data measuring the use of negative campaigning by 31 political parties in 23 parliamentary election campaigns in Germany, the United Kingdom and the Netherlands between 1980 and 2006. Results show that there are no signs of an increase in negative campaigning and that the majority of attacks are issue attacks. However, there are systematic differences in the overall level of negative campaigning between these countries and we suspect that the type of party system could be the main explanatory factor.

The rise in negative campaigning is a recurring theme in the media coverage of recent election campaigns in Western Europe and also in political commentary. Whenever a party or politician attacks his or her opponent, journalists and scholars often speculate about an increase in negative campaigning and its consequences for democracy. This raises questions such as whether we are experiencing a growth in ‘American practices’ and whether or not this is a desirable development. The latter question is logical given that numerous US studies highlight the potentially corrosive effects of negative campaigning on democracy. Negative campaigning is said to lower turnout, decrease political efficacy, depress the public ‘mood’ and increase political cynicism (Ansolabehere and Iyengar, 1995). Nevertheless, various scholars contest these findings (see Goldstein and Freedman, 2002; Kahn and Kenney, 2000; Martin, 2004).

Despite the recurrent focus on the perceived increase of negative campaigning in Western Europe, research in the field of negative campaigning is directed primarily at the United States and research that goes beyond this scope tends to fall into the category of single-country studies that cover a single or few election(s) (e.g. Elmelund-Præstekær 2008; 2009; 2010; Hansen and Pedersen, 2008; Holtz-Bacha, 2001; Schweitzer, 2010; Van Heerde-Hudson, 2011; Walter, 2012; Walter and Vliegenthart, 2010; Walter and Van der Brug, 2013).1 This article presents one of the first comparative studies on negative campaigning in Western Europe. It sets out to answer three research questions: is there evidence of a rise of negative campaigning in Western Europe? Do the levels and characteristics of negative campaigning in Western Europe differ from those in the United States? Are there systematic differences in the incidence of negative campaigning between countries and how can these possibly be explained?

This article thereby contributes to the state of the art on negative campaigning in several ways. First of all, the study addresses the need for comparative research on negative campaigning. The current emphasis on the United States has led to a one-sided development of the theory on negative campaigning, namely towards the use of negative campaigning in a two-party system. Comparative work will provide scholars with an enhanced understanding of the conditions that foster and shape negative campaigning across different settings. Second, the study presents interesting new data on this campaign practice, as it not only examines negative campaigning in three Western European countries – Germany, the United Kingdom and the Netherlands – but also does so over an extensive period of time, looking at all 23 parliamentary election campaigns held between 1980 and 2006. Moreover, as the data have been collected using a method of content analysis based on the work of John Geer (2006), an indirect comparison with the United States is possible. Finally, this article aims to contribute to the discussion among politicians, journalists and scholars on the ‘increasing’ use of negative campaigning in Western Europe and its possible effects on our democracy by examining empirically whether there is such a rise.

This study shows that negative campaigning is not a typically American phenomenon. Negative campaigning also takes place in Western European parliamentary election campaigns. However, while the United States is experiencing a rise in negative campaigning, the results show that Western Europe is not. There is also no evidence that the nature of negative campaigning in Western Europe has changed. However, significant differences do exist between countries in the overall level of negative campaigning, indicating that the political system affects political parties' use of the strategy. We argue that the differences between the countries are linked to their different party systems.

The structure of this article is as follows. First of all the concept of negative campaigning is clarified and the context that is said to have generated an increase in negative campaigning in Western Europe is described, followed by a theoretical explanation of why systematic differences in the level and characteristics of negative campaigning between countries can be expected. Second, the case selection and data collection will be explained before presenting the empirical results and, finally, discussing the implications of the findings.

Defining Negative Campaigning

  1. Top of page
  2. Abstract
  3. Defining Negative Campaigning
  4. The ‘Rise’ of Negative Campaigning
  5. Party System and Negative Campaigning
  6. Case Selection, Data and Coding Procedure
  7. Results
  8. Concluding Remarks
  9. References
  10. Biography
  11. Supporting Information

Defining negative campaigning is a difficult exercise as different definitions are used among different groups of people and within the academic community (Johnson-Cartee and Copeland, 1991; Swint, 1998, p. 49). As a result, journalists, politicians, political consultants, voters and scholars often fail to understand each other when discussing negative campaigning. According to Stuart Surlin and Thomas Gordon (1977) the majority of voters see negative campaigning as any attack on the opponent. This directional definition is employed by most scholars conducting quantitative studies of negative campaigning; namely any criticism directed at one's opponent is a negative attack (e.g. Geer, 2006; Lau and Pomper, 2004; Surlin and Gordon, 1977). The opposite of negative campaigning is positive campaigning, which is constituted by self-praise and promotion.

According to Kerwin Swint (1998), consultants, politicians and campaign workers have a different perspective from most voters on what constitutes negative campaigning. The majority of them consider a campaign message to be negative only if it contains information that is untruthful, deceptive or irrelevant to the campaign, regardless of whether these are issue or trait attacks (Swint, 1998). As Swint (1998, p. 50) argues, political consultants recoil in horror at the suggestion that voters consider any criticism of the opponent to be a form of ‘negative campaigning’. Consultants and candidates usually argue that criticising the opponent is a justifiable and legitimate campaign practice and they do not want this to be labelled as ‘negative’. Some scholars in the field of negative campaigning agree with the definition of the concept employed by the practitioners in the field. This evaluative definition does not consider all criticism directed at an opponent as being negative, but only those critiques that can be considered unfair, illegitimate and dishonest and deal with trivial issues (e.g. Jamieson, 1992; Mayer, 1996).

The directional definition has a number of benefits (see Walter and Vliegenthart, 2010). First, it avoids subjective judgements, such as on whether a campaign is misleading, manipulative or illegitimate. The evaluative definition leaves the assessment of what is negative in the eye of the beholder, and this in turn is dependent on current political norms (Mark, 2006). Consequently, such a definition jeopardises validity and reliability when measuring negative campaigning. Second, the directional definition does not blur the distinction between negative and positive campaigning, as the evaluative definition does. Dishonesty and triviality are not exclusive to negative campaigning, as one can easily imagine positive campaign strategies that stretch the ‘truth’ (e.g. Geer, 2006; Jamieson, 1992). In this article negative campaigning is regarded as all criticism directed at the opponent, as this comparative study is in need of a concept of negative campaigning that is independent of place and time. Furthermore, as we do not enter into the ethical discussion of whether or not attack behaviour is legitimate, the directional definition of negative campaigning is sufficient for our purposes. We distinguish between two types of negative campaigning, namely trait and issue attacks. Issue attacks refer to criticising the plans or policies of an opposing party or candidate. Trait attacks refer to criticising the traits of an opposing party or candidate, that is, his or her integrity or competence.

Even though consultants, campaign managers and scholars do not always agree about the definition of negative campaigning, they do agree that it can be understood as a strategy to win over voters (Mancini and Swanson, 1996). Negative campaigning is aimed at diminishing the positive feelings voters might have for an opponent, thereby winning over the undecided (or even hostile) voter, in contrast to positive campaigning that is thought only to strengthen the loyalty of a party's supporters (Skaperdas and Grofman, 1995).

The ‘Rise’ of Negative Campaigning

  1. Top of page
  2. Abstract
  3. Defining Negative Campaigning
  4. The ‘Rise’ of Negative Campaigning
  5. Party System and Negative Campaigning
  6. Case Selection, Data and Coding Procedure
  7. Results
  8. Concluding Remarks
  9. References
  10. Biography
  11. Supporting Information

Several scholars (e.g. Benoit, 1999; Geer, 2006) argue that the level of negative campaigning in US presidential, congressional and senate election campaigns has risen considerably over time. However, others (Buell and Sigelman, 2009; Lau and Pomper, 2004) dispute these claims. A growth in negative campaigning in Western Europe is likely as political parties are experiencing increasingly unstable electoral markets, mediatisation of politics and professionalisation of election campaigns.

Similar to US candidates, political parties in Western Europe are witnessing increased electoral volatility, although the level varies across countries. 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. Given that nowadays parties are increasingly called upon to handle electoral instability, we can expect negative campaigning to be on the rise.

In addition, 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. 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).

Finally, election campaigns in Western Europe have become professionalised, a development that cannot be seen separately from increased electoral volatility and mediatisation. 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. However, they 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, p. 99). 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 in general believe that negative campaigning works and regard it as a key to electoral success. A candidate who wants to win the election cannot refrain from attacking their opponent (Lau and Pomper, 2004, p. 2; Lau and Sigelman, 2000, p. 13). Owen Abbe et al. (2001) show that in the United States the more professional a campaign is the more negative campaigning it uses. We formulate the following hypothesis:

  • H1: Negativity Hypothesis. The level of negative campaigning has increased in Western Europe.

Although we strive to measure the rise of negative campaigning in Western Europe, it must be mentioned that one should not regard negative campaigning as a completely new phenomenon. The practice of attacking the opponent is as old as politics itself. As Elmer Schattschneider (1960, p. 2) states, ‘At the root of all politics is the universal language of conflict’.

The ongoing debate on attack behaviour and its desirability for democracy not only focuses on the overall tone of the campaign but also on its content. Perhaps the single biggest concern about negative campaigning is the use of trait or character attacks (Geer, 2006; Swint, 1998). Trait attacks are more often considered illegitimate than issue attacks (Geer, 2006; Lau and Pomper, 2004; Swint, 1998). Critics are concerned that the focus on politicians' traits comes at the cost of discussing the issues at stake in the particular election. Informed voting behaviour is regarded as crucial for democracy, therefore voters must be provided with information about different parties or candidates' stands on the campaign issues (Geer, 2006; Johnston and Kaid, 2002). US politics is candidate centred and this becomes especially apparent during election campaigns in which they focus more on the candidate him or herself than the candidate as a representative of the political party (Dalton et al., 2000; Newman, 1994). As a result, US campaigns are associated with trait attacks. As attack behaviour is perceived to be on the rise, so are trait attacks, especially since the decline of party identification has led to an even greater focus on politicians in the US (Dalton et al., 2000; Newman, 1994). However, strong evidence for a rise in trait attacks in American election campaigns is lacking. Geer (2006, p. 145) concludes that the personalisation of politics has not worked its way into presidential advertising. William Benoit (1999) and Darrell West (2005) even state that there has been a persistent trend towards more reliance on policy attacks than character attacks in campaign ads.

Parliamentary democracies in Western Europe are currently witnessing a process of increased presidentialisation or personalisation (Farrell, 2005; Holtz-Bacha and Kaid, 2006; Mughan, 2000; Poguntke and Webb, 2005). Nevertheless, some scholars (e.g. Karvonen, 2010; Kriesi, 2011) contest this development. Several types of personalisation are said to be occurring; here we are interested in so-called media personalisation. This concept refers to a change in the presentation of politics in the media, as expressed in a heightened focus on individual politicians and a diminished focus on parties, organisations and institutions (Rahat and Sheafer, 2007). The increased focus on party leaders is the result of decreasing party loyalty, the growing prominence of electronic media and last but not least mediatisation, as parties themselves also increasingly choose to tailor their campaigns to the demands of the media and project themselves through the personalities of their leaders (Dalton et al., 2000; McAllistar, 2007; Mughan, 2000). These current developments make it likely that Western Europe will witness not only an increase in negative campaigning but also a specific increase in attacks on the opponent's traits. We would like to introduce a theoretical distinction between trait attacks in general and trait attacks targeted at a politician. Whereas in the US trait attacks are almost automatically linked to targeting a candidate, this is not the case in Western Europe. Next to individual politicians, parties are also attacked on their traits, as both actors also are on issues. To illustrate, in a Liberal Democrat ad of the 2005 election campaign voters said: ‘I feel Labour has lied, and they've gone against their promises’; ‘I don't think Tony Blair or Michael Howard can be trusted’. The first quote constitutes two trait attacks on the trustworthiness of the Labour Party and the second quote contains two trait attacks on the trustworthiness of Labour party leader Tony Blair and Conservative Party leader Michael Howard. Consequently, trait attacks can be defined in several ways. First of all, one can take all trait attacks together regardless of whether they are aimed at parties or individual politicians. Second, one can consider trait attacks solely as attacks on traits of individual politicians. We expect an increase in both types of trait attack, so we formulate the following hypothesis:

  • H2: Personalisation Hypothesis. The level of trait attacks has increased in Western Europe.

Party System and Negative Campaigning

  1. Top of page
  2. Abstract
  3. Defining Negative Campaigning
  4. The ‘Rise’ of Negative Campaigning
  5. Party System and Negative Campaigning
  6. Case Selection, Data and Coding Procedure
  7. Results
  8. Concluding Remarks
  9. References
  10. Biography
  11. Supporting Information

Under the influence of the volatile electoral market, mediatisation and the professionalisation of election campaigns, negative campaigning is thought to have become a common practice in Western Europe. However, we do not expect that the level and characteristics of negative campaigning will be similar across Western European countries or similar to the level found in the United States. From the campaign literature we know that 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 limited number of countries in our study prevents a statistical analysis of the influence of all these political system characteristics on negative campaigning. However, we have a strong theoretical basis for expecting the type of party system to affect the likelihood of negative campaigning, which is why we elaborate on this factor here. At the same time, we do not claim that other characteristics are of no or lesser importance for understanding the phenomenon of negative campaigning across countries.

Most studies examining the use of negative campaigning assume that candidates or parties are rational actors that engage in a cost–benefit analysis before deciding whether to make use of negative campaigning (e.g. Lau and Pomper, 2004; Riker, 1996). When the expected benefits outweigh the (potential) costs, parties will decide to make use of negative campaigning in an election campaign. However, these decisions are not made in a political vacuum; the cost–benefit analysis that parties make is not only affected by the features of the particular election they compete in (such as closeness of the race and type of election) but also the characteristics of the political system. This latter aspect has been neglected due to the absence of comparative work. Several non-US single-country studies on negative campaigning (e.g. Elmelund-Præstekær, 2008; 2010; Hansen and Pedersen, 2008; Walter, 2012) suggest that practising negative campaigning in a multiparty system is different from in a two-party system, thereby indicating the importance of this political system characteristic for the study of negative campaigning across countries.2 The party system is especially important if one aims to understand negative campaigning in Western Europe as most continental European countries have a party system characterised by multiple parties and coalition governments (Hobolt and Karp, 2010). The party system in which parties operate affects the cost–benefit structure of making use of negative campaigning. In a two-party system attacking parties fear so-called backlash or boomerang effects. Negative campaigning can cause negative feelings towards the attacker instead of the targeted party or candidate (Garramone, 1984; Johnson-Cartee and Copeland, 1991). Besides the potential risk of losing voters, negative campaigning can certainly help a party to win a race. In a two-party system negative campaigning is beneficial when voters turn away from the opponent, thereby making the attacking party the largest player in the field. Although desirable it is not a necessity to win over those voters to gain the upper hand in an election; in essence a party already ‘wins’ if the opponent's voters decide not to go to the ballot box.

In a multiparty system parties face a different cost–benefit structure when deciding whether to go negative; this is the result of two main differences between a two-party system and a multiparty system. First of all, multiparty competition provides a party with many more potential opponents. This larger number of competitors affects the expected benefits of negative campaigning. It increases the uncertainty of acquiring the benefits of attack behaviour as voters have a much broader range of parties to choose from. Voters who turn away from their party as a result of negative campaigning are not necessarily won over by the attacking party as they might decide to vote for another party (Elmelund-Præstekær, 2008). As a result, an unintended consequence of negative campaigning might be that it contributes to the success of a party other than one's own.

The second difference is the need to form a coalition government after the elections and how this affects the costs of negative campaigning. Parties that make use of negative campaigning not only face electoral backlash effects, but also potential coalition bargaining costs in a multiparty system (e.g. Brants et al., 1982; Elmelund-Præstekær, 2010; Hansen and Pedersen, 2008; Roper et al., 2004; Walter and Van der Brug, 2013). A campaign that is too aggressive, rough and negative may damage the opportunity to govern together (e.g. Brants et al., 1982; Holtz-Bacha and Kaid, 2006, p. 6). Therefore, negative campaigning within multiparty competition brings out a trade-off between the different goals that parties pursue. Within a two-party system parties can concentrate on vote-seeking behaviour as a majority of votes enables them to achieve office and implement policy. In contrast, parties operating within a multiparty system have to balance carefully their vote-, office- and policy-seeking objectives as obtaining the most parliamentary seats does not automatically translate into government office or policy influence (Strøm and Müller, 1999). Consequently, due to the absence of a clear connection between winning votes and obtaining office in a multiparty system, attack behaviour might not always be beneficial.

Overall, parties in a multiparty system thus face a different cost–benefit structure from parties in a two-party system. As a result of the different cost–benefit structure we expect that parties in a two-party system will be more inclined to make use of negative campaigning than parties in a multiparty system.

  • H3: Party System Hypothesis. The level of negative campaigning is higher in a two-party system than in a multiparty system.

Case Selection, Data and Coding Procedure

  1. Top of page
  2. Abstract
  3. Defining Negative Campaigning
  4. The ‘Rise’ of Negative Campaigning
  5. Party System and Negative Campaigning
  6. Case Selection, Data and Coding Procedure
  7. Results
  8. Concluding Remarks
  9. References
  10. Biography
  11. Supporting Information

For this article we measured negative campaigning on the basis of content analysis of campaign material of political parties in 23 British, Dutch and German parliamentary election campaigns between 1980 and 2006. We have chosen to study multiple elections in three countries instead of single elections in a larger number of countries. The multiple time periods are needed to be able to speak of a change in the use of negative campaigning.3 The Western European countries that we compare with each other and with the United States are the United Kingdom, Germany and the Netherlands. These Western European countries are all established parliamentary democracies with election campaigns that have become more candidate focused, but which in their fundamentals remain party coated. In addition, their election campaigns are increasingly led by campaign professionals (Farrell, 2005). Among these countries we find substantial variation between the party systems, and we expect this to be among the main factors affecting the use and characteristics of negative campaigning. The Netherlands and Germany have a multiparty system and Great Britain a two-party system.4 We examine the parliamentary elections, which are the first-order elections for these West European countries. First-order elections are elections in the main political arena of the country (Reif, 1985). In the US the presidential elections are the first-order elections. We restrict the data collection to campaign material from national political parties represented in the lower house of parliament in each of these countries within this time period. We do so for reasons of comparability across countries and the ability to access these materials.5 As a result, regional political parties, such as the Scottish National Party (SNP), Plaid Cymru and Sinn Féin, are not included.

We determine the degree of negative campaigning on the basis of party election broadcasts and televised election debates. Party election broadcasts are completely controlled and officially approved by the party leadership and therefore provide a reliable source to measure party or candidate behaviour. In addition, US research on negative campaigning is primarily based on television ads and we wish to make a valid comparison with John Geer's data (Geer, 2006). We acknowledge that negative campaigning is not restricted to television ads and that ads may not have such a prominent role in all Western European election campaigns as in the United States, as political parties in Western Europe often deal with more restrictive rules when it comes to purchasing advertising time and have fewer funds to buy them (Holtz-Bacha and Kaid, 2006; Van Praag, 2005). Therefore we complement the party election broadcasts with televised election debates. In countries in which party election broadcasts play a minor role, election debates have a more prominent place in the election campaigns. Although election debates are a semi-controlled source, as the moderator steers the discussion with his or her questions, the party leader is still present and in control of what he or she says and therefore they serve as a good indicator of party and/or candidate behaviour.

In order to ensure comparability between the West European countries, we include only party election broadcasts that are aired during free allocated broadcasting time. In the United Kingdom, the purchase of commercial airtime is prohibited, while in Germany and the Netherlands airtime is available and can be bought. In Germany, it has been possible to purchase airtime on commercial television since 1989 and in the Netherlands on both commercial as well as public television since 1998 (Holtz-Bacha, 2000; Holtz-Bacha and Kaid, 2006). While purchasing airtime is allowed in the Netherlands, it is still not a widespread practice (Van Praag, 2005). This being said, in all three countries free broadcasting time is granted on public television channels and in the United Kingdom also on the commercial channels (Holtz-Bacha and Kaid, 2006, p. 10). In these three countries there are no legal restrictions on the content of political advertisements; the only requirement is that they fit the provided broadcast time. As the main bulk of party election broadcasts are taped before the election campaign, the data source in general reflects the party strategy at the beginning of the campaign. After an extensive archival search, we are sure that we have all still existing ads; nevertheless for the Netherlands not all ads have been saved – primarily they are missing for the 1981 and 1982 election campaigns.6 In total 377 party election broadcasts were collected on tape or as transcripts from various public and private archives (see Tables S1, S2 and S3 in the online Appendix).

Televised election debates are traditionally part of the German and Dutch campaign culture, but not of the British. Until the recent 2010 British elections, no televised election debate had ever been held, as parties were always unable to agree on the terms under which it would be conducted (Mitchell, 2000). In Germany, there were televised election debates in which the main representatives from all parties represented in the Bundestag participated between 1969 and 1987. These debates are known as the Elephant rounds and were held three days before the elections (Maurer and Reinemann, 2003). For some time after 1987 these election debates took place after the elections and therefore they are not included in our study. Since 2002, debates between the two main political party leaders have also been held. In the Netherlands, televised election debates have been aired on public broadcast channels since 1963 and on commercial broadcast channels since 1994. Although various election debates always take place, two main debates occur during every election campaign: the public election debate between the main party leaders on the evening before Election Day and the election debate between the main party leaders on the commercial channel. In total nineteen televised election debates were collected on tape or as transcripts from various public and private archives (see Tables S4 and S5 in the online Appendix). Negative campaigning in the debates is measured by looking at:

  • the number of negative (any criticism against the opposition) appeals made in comparison to the total number of appeals. The latter is the sum of positive and negative appeals, where positive appeals consist of any self-praise of the party or politician;
  • the content of positive and negative appeals (are they issue or trait attacks?);
  • the target of the appeals (positive appeals are in general always directed at the politician or its party; negative appeals can be directed at the status quo,7 the government, the opposition, a specific opposing political party, a specific opposing politician or a cluster of opposing political parties or politicians).

The coding scheme will be illustrated by a quote from Hans Janmaat, party leader of the Dutch right-wing extremist party Centre Democrats, in his party election broadcast of 1989:

The Christian Democrats are only interested in money and power. For that Lubbers [CDA party leader] will sell his own mother to the devil. He will start by privatizing state companies and immerse the Netherlands in an uncertain Europe … We fight for our own society. We want to rescue the Netherlands. We want to point out to our youth the dangers of mixed racial marriages.

The first three sentences of this quote are coded as negative campaigning, in which the party CDA is attacked twice on its integrity, the party leader Lubbers is attacked on his integrity and then the party leader is attacked twice on his proposed policies with respect to the economy and the European Union (five negative appeals). The three following sentences are coded as positive campaigning; they would be coded as three appeals of self-praise of the Centre Democrats on immigration policy.

The transcripts of campaign broadcasts serve as the basis for analysis.8 The unit of analysis is a natural speaking unit, that is, the appeal, which is any mention of self-praise or criticism of the opponent. The complete ad is not the unit of analysis as ads often contain both positive as well as negative campaign messages (Geer, 2006). The same applies to sentences, which often contain more than one appeal. The coding method is based on the procedure developed by Geer (2006). Negative campaigning is measured by looking only at explicit and visible manifestations, that is, manifestations that the voter can see are taken into account. We do not code the images as they are even more subjective to the viewer's interpretation than text and for some ads we only have the transcript. One of the main advantages of Geer's (2006) method is that it is applicable beyond ads and can also be used to examine negative campaigning in election debates. In addition, using Geer's method enables us to make comparisons with his data on the United States.

The content analysis is executed by a group of native-speaking graduate students. Geer's (2006) coding method proved to be reliable both within and across countries. The most difficult coding category was to determine the unit of analysis, that is, what does or does not constitute an appeal and whether a text segment consists of one or multiple appeals. The inter-coder reliability (Krippendorf's alpha) was 0.66 for selecting the unit of analysis.9 For the tone (negative versus positive), kind of appeal (issue or trait) and its target (status quo, government, party, cluster of parties, politician, cluster of politicians) Krippendorf's alpha was 0.97, 0.69 and 0.82, respectively. The inter-coder reliability across countries was tested on a sample size of 5 per cent of the collected materials.10 In total 15,988 appeals were generated: 8,931 appeals from the party election broadcasts and 7,057 appeals from election debates. Geer's (2006) data from his book In Defense of Negativity will be used to compare our data with the United States. He coded a total of 795 ads for the period 1960–2000 which resulted in 5,414 appeals; we only make use of the data from the period 1980–2000 which is based on 514 ads.

As we lack complete records of how many times each party election broadcast was aired, where exactly and what the audience size of all election debates and party election broadcasts was, we are unable to perform frequency-weighted or rate-weighted analysis (Prior, 2001). However, we do know that all of these party election broadcasts and election debates were aired. Instead the data are weighted in two separate steps. First of all, as the number of appeals varies greatly across parties in an election campaign (partly due to the different number of ads produced by the parties) and we want to prevent a party with more appeals influencing the overall campaign image more than a party with few appeals, we weight the party election combinations equally.11 Second, we acknowledge that not all parties affect the overall campaign image equally; namely, large parties are more important for the overall image than small parties. As a result, we have weighted the data in the second step according to party size measured on the basis of the election results.12 To estimate whether there are country differences in the level of negative campaigning and whether there is an increasing trend over time, we run several logistic regression models with heteroskedastic errors. As the appeals are not independent from the party and the election, we cluster on the party election combinations. Three dichotomous dependent variables are estimated in the models. First of all Tone (0 = positive; 1 = negative), Trait Attack (0 = no; 1 = yes) and Trait Attacks Targeted at Politicians (0 = no; 1 = yes). The independent variables incorporated are first of all country dummies. For these country dummies the Netherlands is the reference category. In addition, we incorporate interactions between the country dummies and the time variable. As party election broadcasts and televised debates are of a different nature, we run all the analyses for both campaign means separately.

Results

  1. Top of page
  2. Abstract
  3. Defining Negative Campaigning
  4. The ‘Rise’ of Negative Campaigning
  5. Party System and Negative Campaigning
  6. Case Selection, Data and Coding Procedure
  7. Results
  8. Concluding Remarks
  9. References
  10. Biography
  11. Supporting Information

We turn to the empirical results to answer our research questions. First of all, are we witnessing a rise in negative campaigning in Western Europe? Based on the party election broadcasts (see Figure 1) there is no reason to believe that negative campaigning is on the rise for any of the three Western European countries. This finding is supported by the measurements of Germany and the Netherlands on the basis of televised election debates (see Figure 2). The overall level of negative campaigning in election debates is higher than in party election broadcasts. The results of logit models 1 and 2 presented in Table 1 support this finding. They show that there are no significant positive trends, only a significant negative trend for the United Kingdom. On the basis of party election broadcasts, the overall level of negative campaigning for the UK seems to have decreased over our research period. Similar to West (2005, p. 68) we argue that there are natural variations in the cycle of negativity, but that there appear to be no grounds for speaking of an upward trend. The Negativity Hypothesis (H1) is rejected. The findings are in accordance with findings of several single European country studies which all argue that while there is evidence of volatility in negativity, there has been no absolute increase in the number of negative appeals (Elmelund-Præstekær, 2009; Esaiasson and Håkansson, 2002; Holtz-Bacha, 2001; Van Heerde-Hudson, 2011). If there is a rise in negative campaigning at all (e.g. Buell and Sigelman, 2009; Lau and Pomper, 2004) then this seems to be a phenomenon peculiar to the US context.

figure

Figure 1. Amount of Negative Campaigning in Parliamentary Election Campaigns 1980–2006 (Party Election Broadcasts)

Note: Appeals are weighted on the basis of party size. N = 8,931 (NL = 3,504, UK = 3,102, GE = 2,325). See for the United States Geer, 2006.

Download figure to PowerPoint

figure

Figure 2. Amount of Negative Campaigning in Parliamentary Election Campaigns 1980–2006 (Election Debates)

Note: Appeals are weighted on the basis of party size. N = 7,057 (NL = 4,401, GE = 2,656).

Download figure to PowerPoint

Table 1. Country Differences and Trends over Time in Negative Campaigning 1980–2006
  Model 1 Negative appeals (PEB) Model 2 Negative appeals (debates) Model 3 Trait attacks (PEB) Model 4 Trait attacks (debates) Model 5 Trait attacks aimed at politicians (debates)
  1. Notes: PEB = Party election broadcasts. Table entries are logistic regression coefficients with White heteroskedastic standard errors in parentheses. For models 1 and 3 standard errors are adjusted for 141 clusters. For models 2, 4 and 5 standard errors are adjusted for 61 clusters. For the country dummies the Netherlands is the reference category. **significant at p < 0.01; *significant at p < 0.05; °significant at p < 0.10 (two-tailed).

Germany

−0.352

(0.337)

28.128

(46.206)

−0.324

(0.502)

50.808

(40.059)

84.801

(41.36)

United Kingdom

1.193**

(0.277)

0.196

(0.428)

Germany*

Time

−0.035

(0.019)

−0.058

(0.018)

−0.075°

(0.045)

−0.032

(0.017)

−0.032°

(0.018)

United Kingdom*

Time

−0.034*

(0.014)

−0.005

(0.026)

The Netherlands*

Time

0.003

(0.016)

0.008

(0.141)

−0.021

(0.019)

−0.007

(0.107)

0.010

(0.011)

Constant

−1.059

(0.234)**

−17.142

(28.181)

−2.850

(0.310)

10.964

(21.273)

−21.924

(21.602)

Wald X292.11**0.496.4817.6022.66**
McFadden's R20.0540.0010.0110.0330.0379
Correctly classified70.96%62.62%94.89%88.10%90.46%
N8,9317,0578,9317,0577,057

Recent electoral developments in Western Europe not only made us expect a rise in negative campaigning, but also a rise in trait attacks in particular. Figures 3 and 4 not only show that negative trait appeals are only a negligible part of all the appeals made, but that a clear increasing trend is not identifiable. Models 3 and 4 in Table 1 show that there are no significant linear trends to be found either in the party election broadcasts or the televised election debates. The only finding bordering on significance that can be discovered is for Germany in party election broadcasts and this is a negative effect. In addition, no signs of a linear trend are found when one looks solely at trait attacks towards individual politicians. These findings are not displayed graphically as such attacks are rare in general and particularly in party election broadcasts. In the United Kingdom the percentage of trait attacks on politicians varies between 1 and 3 per cent of the total number of appeals, in Germany it varies between 1 and 4 per cent and in the Netherlands between 1 and 3 per cent. Model 5 again only shows a borderline significant negative effect for Germany. Regardless of the increased focus on the party leader during current parliamentary election campaigns, no evidence for a rise in trait attacks during the period 1980–2006 can be found. Thus, the Personalisation Hypothesis (H2) is rejected.

figure

Figure 3. Amount of Trait Attacks in Parliamentary Campaigns 1980–2006 (Party Election Broadcasts)

Note: Appeals are weighted on the basis of party size. N = 8,931 (NL = 3,504, UK = 3,102, GE = 2,325). See for the United States Geer, 2006.

Download figure to PowerPoint

figure

Figure 4. Amount of Trait Attacks in Parliamentary Election Campaigns 1980–2006 (Election Debates)

Note: Appeals are weighted on the basis of party size. N = 7,057 (NL = 4,401, GE = 2,656).

Download figure to PowerPoint

On the basis of Figure 1 we have already seen that there are indeed country differences. When we calculate mean levels for the countries across election campaigns, we find that for party election broadcasts the average level of negative campaigning is 43 per cent for British election campaigns, 29 per cent for Dutch election campaigns and 18 per cent for German election campaigns. According to Geer (2006) the average level of negative campaigning in US presidential campaigns is 36 per cent. The British parliamentary election campaigns resemble US presidential election campaigns the most when it comes to the overall level of negativity. The country dummy variables in model 1, in Table 1, also show that for party election broadcasts there is a significant positive difference between the Netherlands and the UK, indicating that the Netherlands and Germany are more similar to one another than to the UK. The high similarity between these countries is supported by the results of the election debates as the mean level of negative campaigning only differs by 3 per cent – 38 per cent for Germany and 41 per cent for the Netherlands. Furthermore, no significant difference in negative campaigning between these countries is found when looking at the country dummy in model 2 in Table 1. To summarise, we have seen that the highest levels of negative campaigning can be found in two-party systems; the level of negative campaigning is considerably higher in the United States and United Kingdom than in the Netherlands and Germany. Thus, we do indeed find some evidence for a relationship between the party system and negative campaigning and cannot reject the Party System Hypothesis (H3).

Concluding Remarks

  1. Top of page
  2. Abstract
  3. Defining Negative Campaigning
  4. The ‘Rise’ of Negative Campaigning
  5. Party System and Negative Campaigning
  6. Case Selection, Data and Coding Procedure
  7. Results
  8. Concluding Remarks
  9. References
  10. Biography
  11. Supporting Information

This comparative study on negative campaigning outside the US covers 31 political parties in 23 election campaigns in three countries over a period of approximately 25 years. Although this is one of the first comparative studies and one of the most extensive studies on negative campaigning conducted so far, it is still limited in geographical scope, campaign means and time frame; further research is needed to strengthen the preliminary claims made and conclusions drawn in this study. This being said, the study not only offers scholars greater knowledge on the level and characteristics of negative campaigning in Western Europe, but also contributes to the recurrent debate in Western Europe on whether we are witnessing an ‘Americanisation’ of election campaigns. The study does not find empirical evidence for a rise in negative campaigning in Western Europe, thereby supporting several other single European country studies which suggest that there is little empirical foundation for this concern. For now, it suggests that if negative campaigning is on the rise in the US this seems to be a solely American phenomenon. Based on this result we would urge scholars studying the US system to look for specific country characteristics that may fuel this development.

In addition, no evidence has been found for a rise in trait attacks. Although some campaign literature argues that election campaigns are becoming more personalised in Western Europe we do not see such a development in negative campaigning. The absence of a rise in trait attacks in general and those targeted specifically at individual politicians in this study could also be regarded as evidence supporting the opponents of the personalisation thesis. This study thereby contributes to the field that discusses the personalisation of election campaigns. Given that issues are still the main focus of attack behaviour in parliamentary election campaigns in Western Europe, it appears that at this point in time there is no reason to be concerned that campaigns are becoming less substantial. Furthermore, we found some evidence for the notion that the party system constrains the use of negative campaigning. We found lower levels of negative campaigning in multiparty systems in comparison to a two-party system, thereby confirming a notion mentioned in the campaign literature (e.g. Farrell, 2005). A higher overall level of negative campaigning might simply be the by-product of the adversarial politics that comes with a two-party system.

As our study shows considerable differences between the West European countries as well as between Western Europe and the United States, the field would benefit from further comparative research. More comparative research is needed to provide a theory of how political system characteristics other than the party system affect the use of negative campaigning and to test this theory empirically. However, a case study of a single country that experienced party system change would also be useful to determine the impact of the party system on negative campaigning. Future work should include studies that try to explain the level of negative campaigning using a hierarchical model on the basis of party, election and political system characteristics.

What are the implications of our finding that there is no rise in negative campaigning? Is the perception of a rise in negative campaigning among journalists, politicians and voters wrong? Future research should examine this phenomenon across a wider time frame than the 25 years examined, but for now let us assume that this would not change our findings. The fact that we did not see a rise in negative campaigning over this research period does not necessarily mean that the perception of growth in this campaign practice is incorrect. It might well be the case that the perception of an increase in negative campaigning is not based on the behaviour of political parties, but on the way that the media cover election campaigns. Annemarie Walter and Rens Vliegenthart (2010) show that newspaper articles are four times more likely to contain trait attacks than party election broadcasts and the present study does not even take journalistic commentary into account. Second, a rise in negative campaigning might still lie ahead of us. Mark Franklin (1992) argues that although Western European countries are experiencing a weakening in the relationship between the voter's social position and party preference, this process of particularisation does not happen at the same pace in each country. It may be that negative campaigning is a similar phenomenon, and that the US example is a forerunner of what is yet to come.

While we are aware of the limitations of this study in terms of election years, number of countries and campaign means examined, we hope to have inspired scholars in the field to opt for studies that go beyond their own context.

Notes

I thank Wouter van der Brug, Catherine de Vries, Philip van Praag, Peter van Aelst, Huib Pellikaan, Travis Ridout and the participants of the session ‘Comparative Political Communication’ at the 2011 Dutch Belgian Political Science Association Annual Conference for their useful comments.

  1. 1

    Notable exceptions are two papers that have not yet been published. The first is a working paper by Desposato (2008) that develops a game model on the basis of several Latin American countries. The second is a conference paper by Salmond (2011) that examines among other characteristics the tone of YouTube ads in the twelve most recent elections in twelve different countries.

  2. 2

    Salmond (2011) suggests that it is the electoral system and not the party system that causes these differences in the level of negative campaigning across countries. We acknowledge that these two system characteristics are related; however the causal order between the electoral system and the party system is debatable (e.g. Benoit, 2007). In addition, there are exceptions to Duverger's law, such as one of the countries studied, that is, the Netherlands. The Netherlands has always been a multiparty system regardless of its electoral system (Andeweg and Irwin, 2009). However, we argue that although the party system might be caused by the electoral system, it is in practice the number of parties and the need for coalition government that matters for the level of negative campaigning.

  3. 3

    A longer period of study would have been preferable; however, pre-1980 campaign material is not available for all three countries in a systematic way.

  4. 4

    We define Great Britain as a two-party system since in the period studied there was no need to form a coalition government.

  5. 5

    Collecting data from all parties that participated in an election campaign would not be that simple. In the case of the Dutch multiparty system it means that one has to collect data from an additional 74 parties which do not exist anymore and no party archive is present.

  6. 6

    We deal with these missing ads by testing whether our findings hold when we run our analyses without the data of the Dutch parliamentary election campaigns of 1981 and 1982. This is the case.

  7. 7

    Status quo refers to comments that praise or critique the present situation and imply that a political actor is partly responsible for it. However the exact identity of this actor remains unclear, such as one of the ruling parties, the present coalition government or the previous government. Instead of assuming a certain target we code these appeals as directed at the status quo. For party election broadcasts 9.5 per cent of all the appeals fall into this category and for election debates 6.5 per cent of all the appeals do so. When we exclude the appeals coded with status quo as target from the analysis the overall results remain the same.

  8. 8

    From all the televised party election broadcasts and debates transcripts were made if they did not already exist.

  9. 9

    We will elaborate on how the inter-coder reliability of selecting the unit of analysis is calculated. Coders not only have to code the same text segments, but also identify the same number of appeals. See the following fragment of the 1994 RPF party election broadcast: ‘The RPF, list 8. A party that stands for its beliefs, a party with a clear foundation and practical policies. Vote May 3rd RPF, list 8’. These are three sentences of which only the second sentence should be coded and in this sentence three separate appeals can be found. Say that coder A does this correctly, coder A then has the score 0 1 1 1 0. The value 0 represents not coding a text segment and the value 1 represents coding a text segment. Coder B makes a mistake and codes the whole second sentence as one appeal instead of three, coder B then receives the score 0 1 0 0 0. Coder C who correctly codes these three appeals but wrongly identifies the vote call at the end as a fourth appeal receives the score 0 1 1 1 1. On the basis of these scores the Krippendorf alpha is calculated.

  10. 10

    The inter-coder reliability (Krippendorf's alpha) inside the three countries for the unit of analysis, tone, kind of appeal and target was, respectively, 0.77, 0.92, 0.83 and 0.92 for the Netherlands. For Germany the inter-coder reliability was for the same categories in the same order 0.71, 0.97, 0.81 and 0.94. For Britain the inter-coder reliability was 0.68, 0.94, 0.68 and 0.76. The inter-coder reliability within the countries is based on a sample size of 10 per cent of the country-specific materials.

  11. 11

    In general the length of ads has become shorter over time leading to a smaller total of appeals made in an election campaign. This does not affect our findings as it impacted on all political parties equally and we measure negative campaigning as the number of negative appeals relative to the total number of appeals made.

  12. 12

    We have not weighted the data on the basis of party size in a linear manner, but have taken the square root of the party size. This ensures that larger parties weight more heavily on the results than smaller ones, but that the results are not fully dominated by the larger parties. There is also an argument from inferential statistics to weight parties in this manner. Smaller parties have fewer members of parliament (MPs). Since the standard error of estimates is a function of the square root of the sample size, we also weight the parties by the square root of the proportion of number of MPs, so that our sample follows the sampling distribution when we interpret the sample to reflect the composition of the parliament. This second weighting only affects the Netherlands where various very small parties participate; these parties barely receive attention in the free publicity, or from other parties either during or after the election campaign. Nevertheless, they often make attacks and we do not want them to bias the overall image towards negative campaigning. The findings in this study still hold when they are conducted on the raw data.

References

  1. Top of page
  2. Abstract
  3. Defining Negative Campaigning
  4. The ‘Rise’ of Negative Campaigning
  5. Party System and Negative Campaigning
  6. Case Selection, Data and Coding Procedure
  7. Results
  8. Concluding Remarks
  9. References
  10. Biography
  11. Supporting Information
  • Abbe, O. G., Herrnson, P. S., Magelby, D. B. and Patterson, K. D. (2001) ‘Are Professional Campaigns More Negative?’, in P. S. Herrnson (ed.), Playing Hardball: Campaigning for the US Congress. Upper Saddle River NJ: Prentice Hall, pp. 7091.
  • Andeweg, R. B. and Irwin, G. A. (2009) Governance and Politics of the Netherlands. Houndmills: Palgrave Macmillan.
  • Ansolabehere, S. and Iyengar, S. (1995) Going Negative: How Political Advertisements Shrink and Polarize the Electorate. New York: Free Press.
  • Benoit, K. (2007) ‘Electoral Laws as Political Consequences: Explaining the Origins and Change of Electoral Institutions’, Annual Review of Political Science, 10 (1), 363390.
  • Benoit, W. L. (1999) Seeing Spots: A Functional Analysis of Presidential Television Advertisements, 1952–1996. London: Praeger.
  • Bowler, S. and Farrell, D. (1992) Electoral Strategies and Political Marketing. Basingstoke: Macmillan.
  • Brants, K., Kok, W. and van Praag, P. (1982) De Strijd om de Kiezersgunst, Verkiezingscampagnes in Nederland. Amsterdam: Kobra.
  • Buell, E. H. Jr and Sigelman, L. (2009) Attack Politics: Negativity in Presidential Campaigns since 1960. Lawrence KS: University Press of Kansas.
  • Dalton, R. J., McAllister, I. and Wattenberg, M. P. (2000) ‘The Consequences of Party Dealignment’, in R. J. Dalton and M. P. Wattenberg (eds), Parties without Partisans: Political Change in Advanced Industrial Democracies. Oxford: Oxford University Press, pp. 3763.
  • Desposato, S. (2008) ‘Going Negative in Comparative Perspective: Electoral Rules and Campaign Strategies’, Working Paper. Available from: http://swd.ucsd.edu/gn1.pdf [Accessed 10 September 2013].
  • Elmelund-Præstekær, C. (2008) ‘Negative Campaigning in a Multiparty System’, Representation, 44 (1), 2739.
  • Elmelund-Præstekær, C. (2009) Kammertoner og unoder I dansk valgkamp, Partiernes positive og negative, person og politikfokuserede kampagner 1994–2007. Syddansk Universitet: Syddansk Universitetsforlag.
  • Elmelund-Præstekær, C. (2010) ‘Beyond American Negativity: Toward a General Understanding of the Determinants of Negative Campaigning’, European Political Science Review, 2, 137156.
  • Esaiasson, P. and Håkansson, N. (2002) Besked ikväll! Valprogrammen I svensk radio och tv. Värnamo: Fälth & Hässler.
  • Farrell, D. M. (2005) ‘Campaign Modernization and the West European Party’, in K. R. Luther and F. Müller (eds), Political Parties in the New Europe: Political and Analytical Challenges. Oxford: Oxford University Press, pp. 6383.
  • Franklin, M. (1992) ‘The Decline of Cleavage Politics’, in M. Franklin , T. Mackie and H. Valen (eds), Electoral Change: Responses to Evolving Social and Attitudinal Structures in Western Countries. Cambridge: Cambridge University Press, pp. 383405.
  • Garramone, G. (1984) ‘Voter Response to Negative Political Ads’, Journalism Quarterly, 61, 250259.
  • Geer, J. (2006) In Defense of Negativity, Attack Ads in Presidential Campaigns. Chicago IL: Chicago University Press.
  • Goldstein, K. and Freedman, P. (2002) ‘Campaign Advertising and Turnout: New Evidence for a Stimulation Effect’, Journal of Politics, 64 (3), 721740.
    Direct Link:
  • Hansen, K. M. and Pedersen, R. T. (2008) ‘Negative Campaigning in a Multiparty System’, Scandinavian Political Studies, 31 (4), 408427.
  • Hobolt, S. B. and Karp, J. (2010) ‘Voters and Coalition Government’, Electoral Studies, 29 (3), 299307.
  • Holtz-Bacha, C. (2000) Wahlwerbung als politische Kultur: Parteienspots im Fernsehen 1957–1998. Wiesbaden: Westdeutscher Verlag.
  • Holtz-Bacha, C. (2001) ‘Negative Campaigning: In Deutschland Negative Aufgenommen’, Zeitschrift für Parlamentsfragen, 32 (3), 669677.
  • Holtz-Bacha, C. and Kaid, L. L. (2006) ‘Political Advertising in International Comparison’, in L. L. Kaid and C. Holtz-Bacha (eds), The Sage Handbook of Political Advertising. Thousand Oaks CA: Sage, pp. 314.
  • Jamieson, K. H. (1992) Dirty Politics: Deception, Distraction and Democracy. Oxford: Oxford University Press.
  • Johnson-Cartee, K. S. and Copeland, G. (1991) Negative Advertising: Coming of Age. London: Lawrence Erlbaum.
  • Johnston, A. and Kaid, L. L. (2002) ‘Image Ads and Issue Ads in US Presidential Advertising: Using Video Style to Explore Stylistic Differences in Televised Political Ads from 1952 to 2000’, Journal of Communication, 52 (2), 281300.
  • Kahn, K. F. and Kenney, P. J. (2000) ‘How Negative Campaigning Enhances Knowledge of Senate Elections’, in J. A. Thurber , C. J. Nelson and D. A. Dulio (eds), Crowded Airwaves: Campaign Advertising in Elections. Washington DC: Brookings Institution Press, pp. 6595.
  • Karvonen, L. (2010) The Personalisation of Politics: A Study of Parliamentary Democracies. Colchester: ECPR Press.
  • Keet, R. (1990) Lijsttrekkers, Analyses en Peilingen in Kro's Brandpunt-Politiek Café Verkiezingen 1989. Hilversum: KRO.
  • Klein, J. (1990) Elefantenrunden ‘Drei Tage vor der Wahl’: Die ARD-ZDF Gemeinschaftssendung 1972–1987: Einführung und Text-Transkription. Baden-Baden: Nomos.
  • Kriesi, H. (2011) ‘Personalization of National Election Campaigns’, Party Politics, 17 (1), 120.
  • Lau, R. R. and Pomper, G. M. (2004) Negative Campaigning: An Analysis of US Senate Elections. Lanham MD: Rowman & Littlefield.
  • Lau, R. R. and Sigelman, L. (2000) ‘Effectiveness of Negative Campaigning’, in J. A. Thurber , C. J. Nelson and D. A. Dulio (eds), Crowded Airwaves, Campaign Advertising in Elections. Washington DC: The Brookings Institution, pp. 1043.
  • Lilleker, D. G. and Negrine, R. (2002) ‘Professionalization: Of What? Since When? By Whom?’, Harvard International Journal of Press/Politics, 7 (4), 98103.
  • McAllistar, I. (2007) ‘The Personalization of Politics’, in R. J. Dalton and H. Klingemann (eds), Oxford Handbook of Political Behaviour. Oxford: Oxford University Press, pp. 571588.
  • Maier, J., Maier, M., Reinemann, C. and Maurer, M. (2006) Wortlaut der Fernsehdebatte im Bundestagswahlkampf 2005 und ihre Wahrnehmung im Ost-West-Vergleich: Eine Dokumentation. Kaiserlautern: Technische Universität Kaiserslautern.
  • Mair, P., Müller, W. C. and Plasser, F. (2004) Political Parties and Electoral Change: Party Responses to Electoral Markets. London: Sage.
  • Mancini, P. and Swanson, D. L. (1996) ‘Politics, Media and Modern Democracy: Introduction’, in D. L. Swanson and P. Mancini (eds), Politics, Media and Modern Democracy: An International Study in Electoral Campaigns and Their Consequences. London: Praeger, pp. 128.
  • Mark, D. (2006) Going Dirty: The Art of Negative Campaigning. Lanham MD: Rowman & Littlefield.
  • Martin, P. S. (2004) ‘ “Inside the Black Box of Negative Campaign Effects”: Three Reasons Why Negative Campaigns Mobilize’, Political Psychology, 25 (4), 545562.
  • Maurer, M. and Reinemann, C. (2003) Schröder gegen Stoiber, Nutzung Wahrnehmung und Wirkung der TV Duelle. Wiesbaden: Westdeutscher Verlag.
  • Mayer, W. G. (1996) ‘In Defense of Negative Campaigning’, Political Science Quarterly, 111 (3), 437455.
  • Mazzoleni, G. and Schulz, W. (1999) ‘ “Mediatization of Politics”: A Challenge for Democracy’, Political Communication, 16 (3), 247261.
  • Mitchell, A. (2000) ‘The Great British Exception’, in S. E. Coleman (ed.), Televised Election Debates: International Perspectives. London: Macmillan, pp. 104121.
  • Mughan, A. (2000) Media and the Presidentialization of Parliamentary Elections. Houndmills: Palgrave.
  • Newman, B. I. (1994) The Marketing of the President: Political Marketing as Campaign Strategy. Thousand Oaks CA: Sage.
  • Poguntke, T. and Webb, P. (2005) ‘The Presidentialization of Politics in Democratic Societies: A Framework for Analysis’, in T. Poguntke and P. Webb (eds), The Presidentialization of Politics: A Comparative Study of Modern Democracies. Oxford: Oxford University Press, pp. 125.
  • Prior, M. (2001) ‘Weighted Content Analysis of Political Advertisements’, Political Communication, 18 (3), 335345.
  • Rahat, G. and Sheafer, T. (2007) ‘The Personalization of Politics, Israel 1949–2003’, Political Communication, 24 (1), 6580.
  • Reif, K. (1985) ‘Ten Second-Order Elections’, in K. Reif (ed.), Ten European Elections. Aldershot: Gower, pp. 136.
  • Riker, W. H. (1996) The Strategy of Rhetoric: Campaigning for the American Constitution. New Haven CT: Yale University Press.
  • Roper, J., Holtz-Bacha, C. and Mazzoleni, G. (2004) The Politics of Representation: Election Campaigning and Proportional Representation. New York: Peter Lang.
  • Salmond, R. (2011) ‘MeTube: Politicians, YouTube, and Election Campaigns in Longstanding Democracies’, unpublished manuscript, University of Michigan.
  • Schattschneider, E. E. (1960) Semi-sovereign People. Hinsdale NJ: Dryden Press.
  • Schweitzer, E. (2010) ‘Global Patterns of Virtual Mudslinging? The Use of Attacks on German Party Websites in State, National and European Parliamentary Elections’, German Politics, 19 (2), 200221.
  • Skaperdas, S. and Grofman, B. (1995) ‘Modeling Negative Campaigning’, American Political Science Review, 89 (1), 4961.
  • Strøm, K. and Müller, W. (1999) ‘Political Parties and Hard Choices’, in W. Müller and K. Strøm (eds), Policy, Office or Votes? How Political Parties in Western Europe Make Hard Decisions. Cambridge: Cambridge University Press, pp. 135.
  • Surlin, S. H. and Gordon, T. F. (1977) ‘How Values Affect Attitudes toward Direct Reference Political Advertising’, Journalism Quarterly, 54 (1), 157172.
  • Swint, K. C. (1998) Political Consultants and Negative Campaigning: The Secrets of the Pros. Lanham MD: University Press of America.
  • Van Heerde-Hudson, J. (2011) ‘The Americanization of British Party Advertising? Negativity in Party Election Broadcasts, 1964–2005’, British Politics, 6 (1), 5277.
  • Van Praag, P. (2005) ‘De veranderde Nederlandse campagnecultuur’, in K. Brants and P. Praag (eds), Politiek en media in verwarring, de verkiezingscampagnes in het lange jaar 2002. Amsterdam: Spinhuis, pp. 2143.
  • Walter, A. S. (2012) ‘Choosing the Enemy: Attack Behaviour in a Multiparty System’, Party Politics, OnlineFirst. doi: 10.1177/1354068811436050.
  • Walter, A. S. and Van der Brug, W. (2013) ‘When the Gloves Come Off: Inter-party Variation in Dutch Elections, 1981–2010’, Acta Politica, Advance online. doi: 10.1057/ap.2013.5.
  • Walter, A. S. and Vliegenthart, R. (2010) ‘Negative Campaigning across Different Communication Channels: Different Ballgames?’, Harvard International Journal Press/Politics, 15 (4), 441461.
  • West, D. M. (2005) Air Wars; Television Advertising in Election Campaigns, 1952–2004. Washington: CQ Press.

Biography

  1. Top of page
  2. Abstract
  3. Defining Negative Campaigning
  4. The ‘Rise’ of Negative Campaigning
  5. Party System and Negative Campaigning
  6. Case Selection, Data and Coding Procedure
  7. Results
  8. Concluding Remarks
  9. References
  10. Biography
  11. Supporting Information
  • Annemarie S. Walter is Assistant Professor of Communication Science at VU University of Amsterdam. Her primary research interests are election campaigns, party strategy, political advertising, election debates and the relationship between media and politics. Her work has been published in international peer-reviewed journals such as Party Politics, Comparative Political Studies, Acta Politica and the International Journal of Press/Politics. This article is part of her dissertation entitled ‘Negative Campaigning in Western Europe: Beyond the Vote-Seeking Perspective’ written at the Political Science Department of the University of Amsterdam. Annemarie Walter, Department of Communication Science, VU University of Amsterdam, De Boelelaan 1081, 1081 HV Amsterdam, the Netherlands; email: a.s.walter@vu.nl

Supporting Information

  1. Top of page
  2. Abstract
  3. Defining Negative Campaigning
  4. The ‘Rise’ of Negative Campaigning
  5. Party System and Negative Campaigning
  6. Case Selection, Data and Coding Procedure
  7. Results
  8. Concluding Remarks
  9. References
  10. Biography
  11. Supporting Information
FilenameFormatSizeDescription
post12084-sup-0001-si.docx20K

Table S1: Distribution of Party Election Broadcasts included for Great Britain

Table S2: Distribution of Party Election Broadcasts included for Germany

Table S3: Distribution of Party Election Broadcasts included for the Netherlands

Table S4: Election Debates included for Germany

Table S5: Election Debates included for the Netherlands

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.