The use of humor and negative campaigning. An (un)happy marriage?
By Iris Verhulsdonk
10179569
Master Thesis
Graduate School of Communication – University of Amsterdam
Master Program Political Communication
Supervisor: dr. Alessandro Nai
29 June 2018 Wordcount: 7497
Abstract
Negative campaigning has received a large share of attention from scholars and politicians. However, research has not yet found definitive proof of the effectiveness of such campaigns. Moreover, concerns arise over the use of negativity, which could foster political cynicism. One of the reasons the jury is still out on the effectiveness of negative campaigning is taking into consideration the backlash effect, in which the sponsor of an attack is harmed by their own attack. This study puts forward the idea that the use of humor could potentially mitigate these backlash effects. It is hypothesized that humor can soften the blow of negative
campaigning in two ways: by lessening the backlash effect but also by decreasing
effectiveness of hurting the target. Lastly, it is expected that humor cannot alter the negative impact that attack-politics have on political cynicism. It is indeed found that humor can soften the blow, but that it is no silver bullet, as it also lowers attack effectiveness. Knowledge is an important aspect for the influence of negativity on cynicism, with those with high knowledge experiencing increased cynicism after exposure to an attack.
1 Introduction
Negative campaigning can be seen as one of the main trends within contemporary politics. Although attacking a political opponent instead of focusing on ones own favorable
characteristics might not be a new phenomenon, the attention this practice has gotten from both the professional as well as the academic field has seen a increase (Geer, 2012; Walter & Nai, 2015) In the USA a rise is seen in the amount of negative political campaigns. Although these results cannot be generalized to other contexts, there are indicators that American political practices are spreading across the world. Because of which politicians worldwide are more prone to use strategic methods to gain votes, such as negative campaigning (Plasser, 2000). In addition, the large scope of academic research on negativity shows that negative campaigning is a universal aspect of politics (Walter & Nai, 2015). An aspect which
academics have expressed concerns about in terms of the influence negativity might have on the health of democracy. Negative campaigning has been said to have a negative effect on the self-efficacy, political trust and cynicism of the electorate (Thornson, Ognianova , Coyle & Denton, 2000; Dardis, Shen & Hatfield Edwards, 2007). Negative campaigning does not only change the attitudes towards the people involved in the campaign, but also shows the public that the process of politics is flawed, which results in stronger feelings of cynicism towards politicians and the political field in general (Ansolabehere & Iyengard, 1995) Years later, research by Jackson, Mondak & Huckfeldt (2009) has investigated the claimed effects of negative campaigning on systemic political attitudes and found no empirical evidence to back up these claims.
This is not to say the discussion is over. The idea of negative campaigning eroding the pillars of democracy remains prevalent within negative campaigning research. Regardless of these possible worrisome implications of the use of negative campaigning, politicians seem to
2 remain interested in the use of this technique. That should not come as a surprise. Throughout the years, it has become clear that negativity matters in multiple ways. First and foremost, a negativity bias exists within politics. It is found that not only do negative impressions hold more weight than positive cues, but also voters base their decisions more strongly on voting against a candidate than for a certain candidate (Jasperson & Fan, 2002). Thus, going negative has a potential to benefit the political actor in pulling votes away from their opponent.
Moreover, media seem to favor negative campaigning and harsh attacks, by giving more attention to negative campaigns in news articles (Geer, 2012).
The potential of winning votes and increased media attention make for a solid argument for politicians to use this technique in their campaigns. Although politicians have their reasons to make use of negativity, literature has not been able to show exactly under which circumstances (and if) negative campaigning is actually effective in enhancing the position of a politician. For an attack to be effective, a different variables are at play. It matters who the sponsor and target are, in terms of gender, it matters whether the attack is person or issue-based, it matters whether the attack is civil or uncivil. Furthermore, all these different elements can have different effects on a variety of variables: either on short term electoral attitudes towards the politician, the party, or long term attitudes towards politicians and politics as a whole. These effects can be on explicit attitudes, or implicit associations. Moreover, the cognitive processing of information can be altered by different forms of attacks. In sum, there are a lot of variables that can influence and alter the effects of negative campaigning (Lau et al., 1999; 2007).
One of the most important aspects, however, is the potential for negativity to backfire to the sponsor by which the the attacker is also viewed as less favourable by the public (Roese & Sande, 1993). This so called ‘backlash effect’ can occur because voters in general do not particularly like negativity and so, the sponsor of an attack is ‘punished’ by the voter for using
3 such a technique. In this study, I will investigate whether it is possible to soften the blow of negativity, by adding a pleasant element to a negative message. In this case, humor is the device to add a positive note to an otherwise negative political message and could potentially change the way an attack is perceived.
Humor has been found to alter the way in which people perceive satiric messages (Boukes et al., 2010) It is expected that these effects will also occur outside of the satirical field, when politicians use humoristic messages. Specifically, good reasons exists to expect that humor will mitigate the backlash effect. If a message is perceived as funny, it could alter the way in which a message is perceived. It is possible that when a message is seen as funny, fewer negative feelings arise against the one attacking, because the message is perceived as less harmful and generally liked more. It is the purpose of this study to investigate whether humor in negative campaigning influences the effectiveness of this technique.
Research considering humor by politicians focuses mostly on specific cases, such as the self-satire of John McCain in Saturday night live (Becker, 2012) or the way in which Ronald Raegan used humor as an important rhetorical tool during his presidency (Meyer, 2009). Outside of the American context, research on online British humor during the 2005 UK elections shows a growing importance of humor within British political campaigns (Shifman, Coleman & Ward, 2007). This British study, however, is an exception to the rule. Most research campaigning focuses on the American context, with a specific preference for presidential candidates. This study will therefore also contribute to the field by focussing on negativity within the Dutch context.
Although the aforementioned studies show the importance and prevalence of humor within the political spectrum, little attention is given to the specific effects of humor on electoral attitudes outside of real-life case-studies. This study will contribute to the existing field of research by focussing specifically on the way in which humor can moderate the
4 effects of negative campaigning on electoral and systematic political attitudes concerning the candidates, in an experimental setting. This study then contributes to the literature on negative campaigning by investigating which effects negative campaigning has on electoral as well as systemic political attitudes. At the same time, this research will contribute to the study of humor in politics by showing which effects a humorous message can have on these negative campaigning effects.
Theoretical framework
Negative campaigning and the backlash effect
Interestingly enough, negative campaigning has not been found to be an effective means of winning votes per se (Lau et al., 1999; 2007) and the bulk of research does not offer
consensus on the effectiveness of political campaigning. Specific conditions under which negative campaigning yields results have been found. For instance, it is found that difference exists between explicit and implicated attitudes towards sponsor and target of negative campaigning (Carraro & Castelli, 2010a), between high or low volume of negative
campaigning (Nai & Seeberg, 2018) and between the gender of the sponsor (Krupnikov & Bauer, 2014). Moreover, the way in which citizens think and process information which is affected by negative campaigning has received a respectable amount of interest from
academia, with recent research suggesting that voters lower their cognitive effort when they are exposed to a negative message (Walter & Nai, 2015). In short, negativity has a potential to be an important player in the political field, but its effectiveness depends on the specific conditions in which it is deployed. One specific aspects of this, is the message itself. It is shown that relevance and civility of a message matter (Fridkin & Kenney, 2008) as well as issue or person – based attacks (Carraro & Castelli, 2010b) for the effectiveness of negative campaigning.
5 One of the major concerns in studying the effectiveness of negative campaigning, is that results may be inconclusive because they have been focusing solely on attitudes towards the target. However, this is not the only variable at play. Of course, the attack should foster negative feelings towards this target (Nai & Seeberg, 2018) but the attitudes towards the sponsor of the message should also not decrease if the attack really is successful. Evidence suggests that negative messages, specifically those who are not aimed at an issue (Fridkin & Kenney, 2008) can result in negative attitudes towards the source of the attack (Roese & Sande, 1993;Lau & Pomper, 2004; Carraro & Castelli, 2010; Carraro, Bertram, Gawronski & Castelli, 2010). To account for the backlash effect is an important task within negative campaigning research (Jasperson & Fan, 2002), the effectiveness of negative campaigning is two sides of the same coin: the assessment of the target and of the sponsor of the attack. More research needs yet to be done considering both of these variables (Lau, Sigelman, & Rovner, 2007). In this article I argue that the use of humor can alter the effectiveness of negativity on both the target and the sponsor side. Political humor is found to be effective in altering the way people perceive a message, by making people pay more attention to a message but at the same time lessen the perceived seriousness of a message (Boukes et al., 2010; Compton, 2012). This could potentially soften the blow of the backlash effect than negativity can produce.
The effectiveness of political humor and negative campaigning
Humor in politics has mostly been studied within the context of satire. The popularity of shows such as Last week Tonight with John Oliver and The Daily Show has attracted the attention of both audience and scholars. As Compton (2012) finds, empirical assessment of humor in political discourse is mostly focused on television, and specifically calls for continuing empirical research in the field. And with good reason, as the combination of humor and politics mostly has resulted in research on the effects of late-night comedy shows,
6 mocking politicians or the persuasive power of these show hosts (Becker, 2012; Duffy & Page, 2013). Although Compton (2012) makes a good case for investigating political humor further and to “explore the effects in all of its venues” (p. 54), the focus remains mostly on satire. This is a shame, since politicians themselves make use of political humor to a large extent and know the value of it within the political discourse. As formulated by Nilsen: “From the point of view of the politician, humor can be can be used to define political concepts, to disarm critics, to establish détente, to establish a position or make a point […]”(p. 35). In addition, Tsakona (2009) found that humor within the Greek parliament was used to construct and damage the target’s public image, and to attack an opponent without violating
parliamentary rules. She also suggests that humor is used as a means of attacking the
opponent in a mitigated and rule-obedient way (p.233), which means as much as attacking in a way that does not break with the (un)written rules of not coming across as too harsh or rude. Other research put forward that politicians can use humor to express opposition, their political identification and civic support for a cause (Davis, Love & Killen, 2018). Moreover, humor can be effective to for example change the topic of a debate, lighten up the mood or relieve anxiety (Bippus, 2007). implementing these functions of humor by assessing electoral or attitudinal effects are, however, scarce.
Baumgartner (2013), similar to the aim of this study, has investigated whether humor can mitigate the effects of negative campaigning advertisement. In Baumgartner’s
experiment, undergraduate students were exposed to humorous, negative advertising videos of the American elections of 2012 between Romney and Obama. After the exposure respondents were asked to rate the candidates on a 1 to 10 scale as well as score them on 11 positive traits. Lastly, they were asked how fair they thought either candidates campaign was. The study gave mixed results. Romney’s evaluations were lowered both when he was a sponsor as well as when he was the target of the attack. Obama’s evaluations, on the other hand, were barely
7 influenced by the attacks towards him, nor did Obama experience the backlash effect when he was the sponsor of an attack. These differences are found to be only partly due to the fact that the negative advertisements sponsored by Obama were more lighthearted, however two of the advertisements attacking Obama were aired and sponsored by third-party groups. The results of Baumgartner’s study can therefore not be isolated to solely the effects of humor and negative campaigning. In the discussion he calls for further research to isolate the humorous versus the negative effects by conducting an experiment. In light of this call for further research, this study will investigate in what way negative campaigning and humor can form an (un)happy marriage by investigating its effects on electoral and long term systemic political attitudes.
Hypotheses
I start from the assumption that humor can mitigate the backlash effect. Overall, the backlash effect exists because the electorate perceives an attack as ‘untruthful or unjustified’
(Garramone, p. 251, 1984) and citizens seem to dislike negative campaigning. Humor allows candidates to attack their opponents in a way that does not seem harsh, because it is put forward as a joke (Harris, 2009). Moreover, humor may distract the audience from the true purpose of the attacker, making the attack seem less vicious and preventing the backlash effect (Postelnicu & Kaid, 2008). Furthermore, in the context of political satire, it is argued that a humorous element makes attacks more accessible to an audience “opening them to judgement that they may otherwise be unwilling to accept” (Caufield, p. 52, 2008). So, even in circumstances when an attack could be described as uncivil, harsh or vicious, humor can mitigate this effect. Conforming this idea (although the results were not solely due to the humorous component) Baumgartner (2009) did find that under specific circumstances humor can result in less backlash for the sponsor. This results in the following hypothesis:
8
H1: The presence of humor will reduce the backlash effect (reduced evaluation of and voting intention for the sponsor) for an attack, when compared to the same attack without humor
Although humor is expected to mitigate the backlash effect, this does not mean a humorous message is the silver bullet of negative campaigning. I assume that humor might decrease the persuasive power of an attack. There are multiple, adverse, ways in which humor is thought to influence the way we process information. Although political satire has been found to lower argument scrutiny, which could result in easier persuasion (LaMarre, Landreville , Young & Gilkerson 2014), it is reversely found that satirical and humorous messages are taken less seriously overall (Boukes et al., 2015). Furthermore, when a political message is seen as ‘just a joke’, it is suggested that people do not further think about the message or arguments put forward (Innocenti & Miller, 2016). And although humor about social issues increases source liking, the message put forward is discounted for attitudinal judgement (Nabi, Moyer-Guse & Byrne, 2007). These examples of the effects of humor on message perception indicate that a humorous attack will be disregarded: the information in the attack that is supposed to lower target evaluations receives less attention and is therefore less effective in lowering target evaluations. In addition, humor lowers the persuasive power of an argument in itself
(Compton, 2007; Harris, 2009). These findings suggests that humor will make an attack less effective, because the argument that is used to attack the opponent is not considered to be information that should be taken seriously. This results in the second hypothesis:
H2: The presence of humor will reduce the effectiveness of the attack (heightened evaluation of and voting intention for the target) when compared to the same attack without humor
In addition to the short-term electoral effectiveness of negative messages, scholars have investigated long term societal effects of negative campaigning. These effects have been cause for concern. It has been found that negative messages can increase political cynicism,
9 (Thornson, Ognianova , Coyle & Denton, 2000; Schenk-Hamlin, Procter & Rumsey, 2000) especially amongst people with high involvement in politics (Yoon, Pinkleton & Ko, 2005). Moreover, negative messages – and this is even stronger for personal attacks – can increase anxiety and other negative emotions (Nai & Walter, 2015) whereas issue-oriented negativity can expose flaws in the political system and therefore increase distrust towards the system and politicians, fostering political cynicism (Dardis, Shen & Hatfield Edwards, 2007). Political cynicism is considered a lack of faith or trust in politics, governmental institutions, or politicians (Agger, Goldstein, & Pearl, 1961), and although it has been argued that a certain amount of skepticism about politics can enhance participation and therefore be an important asset to the democracy, that generally high levels of cynicism are associated with low political involvement (Van der Meer, 2017). Low political involvement is seen as problematic for a representative democracy in which members of parliament make decisions based on a solid mandate.
Although the use of a humorous message might decrease potential negative effects on the short term, countering backfiring of an attack, the main tone and direction of the message remains negative. Finally, negativity sticks. A taxonomy of the prevalence of negativity within human cognition shows that when an entity is negative, the positive elements are ‘contaminated’ by the negative (Rozin & Royman, 2001).
Being exposed to negative campaigns can make voters more cynical about the responsiveness of politicians, which in turn contributes to overall political cynicism
(Ansolabehere, Iyengar, Simon & Valentino, 2013). Although trust in the (political) system has been found to be a relatively stable attitude amongst voters (Banducci and Karp, 2003), being exposed to harsh messages can strengthen the sense of politics being ‘just a game’ to the politician to gain power. Even if a message is perceived as funny, this does not change that the message in itself was negative and can foster negative attitudes. Regardless if
short-10 term evaluation of a candidates are influenced in favor of the attacker, there is a reduction in respect for both candidates involved in a negative message (Shapiro & Rieger, 1989). In this paper it is argued that humorous messages cannot mitigate the long term negative societal effects of negative campaigning, because perceived funniness will only be there for a couple of seconds – maybe minutes, whereas the negative message will remain. The rationale to this idea is that negativity over clouds positive messages when exposed to both (Schenk-Hamlin, Procter & Rumsey, 2000). Continuing this rationale, the humor in the message cannot alter the effects of negativity on political cynicism, since the negative messages is prone to prevail. Laughing, and perceiving something as funny is a temporal reaction which is different from general disapproval of politicians behaviour. This results in the third, final hypothesis:
H3: Exposure to negative messages increases political cynicism, especially when that message is harsh, regardless of the presence of humor
Method
Experimental protocol
In order to test for the hypotheses that negative campaigning and humor have an effect on candidate assessment and political cynicism I use an experimental setting. In this setting participants are exposed to a mock-news article in which one remark is manipulated across the different conditions. These conditions consists of a combination of negative messages (attack) from sponsor to target and humorous components. In these news articles, real life candidates Jessica van Eijs from D66 (sponsor) and Aukje de Vries from VVD (target) are presented to the respondents.
Participants
For this study I rely on a convenience sample of 177 undergraduate students in
11 research credits for their cooperation, as part of their bachelor degree. Students were invited to conduct the survey by the researcher and they could find the survey online in the LAB of the University of Amsterdam.
Filter
Respondents do not always pay as much attention to a survey as is required off them. To filter out those who did not read the questions carefully, an ‘attention check’ was incorporated in the survey (Oppenheimer, Meyvis & Devidenko, 2009). Between questions on political cynicism, one of the items said ‘please answer ‘somewhat agree’ to this question’. Respondents who failed to answer accordingly to the instructions, were excluded from analysis. This measure resulted in 39 excluded cases.
Furthermore, the results on respondents knowing the candidates were extremely skewed: only 6.78% of the respondents knew Jessica van Eijs and 4.52% for Aukje de Vries. To prevent previous opinions about the politicians to influence the results, respondents who said they knew the candidates were excluded from the analysis. Because of this exclusion, the attitudes concerning the candidates were not biased by pre existing attitudes towards the candidates themselves. This resulted in a total of 121 respondents.
83.5% of the respondents were female, which is slightly higher than the gender division within communication science on the University of Amsterdam. Most respondents (38%) have lived in The Netherlands most of their lives, followed by Germany with 10.7%. Although 82.6% of the sample did not have English as their native language, only 1% rated their English level as ‘moderately low’, with all others reporting high levels of English language skill.
Treatments
Respondents were asked to fill in the questionnaire which started with general questions about themselves and their political opinions such as voting behaviour and opinions about Dutch
12 parties. The questionnaire was in English and respondents had the choice to read the stimulus in Dutch or English. Respondents were asked about their level of English to prevent language barrier bias.
20% of the respondents were randomly assigned to one of the following five conditions: (1) control (no attack) (2) non-harsh attack with humor element (3) non-harsh attack without humor element (4) harsh attack with humor element (5) harsh attack without humor element.1 The non-humorous conditions used the same frame as the humorous conditions and the specific attack was aimed towards the same thing (VVD-ers having no heart or VVD-ers can’t be taken seriously anymore because so many of their ministers have resigned), but the humorous component was removed, just leaving the criticism in itself. The control condition consisted of the news article in which no attack was present, but a general positive comment about D66 was made by the sponsor. All treatments (both in Dutch and English) can be found in Appendix A and B.
Aukje de Vries (VVD) and Jessica van Eijs (D66) are both in the current Dutch house of representatives. These two parties were chosen seeing as they are both liberal, and are part of the current coalition (Rutte III). This was done to ensure that the parties were not extreme towards an ideology and not in either opposition or coalition, to keep the parties as similar as possible. Although VVD is more right-winged than D66, their differences are not amongst the lines of traditional political left and right. Previous attitudes towards these parties are taken into account.
Lastly, both of these candidates are women. The choice to do this is based on the fact that in almost all other studies on effects of campaigns on electoral attitudes, men are used.
1
The difference between the harsh and soft attacks were based on creating a personal attack on the target and a party aimed attack on target’s party. The distinction between harsh and soft was chosen because it is found that negative campaigning is often perceived as more inappropriate, relevant and ‘harsh’ when the attack is aimed towards the person, than when it is aimed towards a party or specific issue (Fridkin & Kenney, 2008) which could result in a higher risk for backlash (Carraro & Castelli, 2010).
13 This study uses female candidates, as research has shown the electoral backlash effect is stronger for women (Krupnikov and Bauer, 2014) especially when the negativity is non-issue based, since these attacks are mostly seen as ‘unfeminine’ (Maier, 2015).
Pre-test
To test for the effectiveness of the treatments in terms of humor, a pre-test was conducted. For this I used a convenience sample, consisting of Dutch relatives and friends.
Respondents were asked to read eight statements, written together with a professional 2 comedian across different subjects such as the looks of the opponent and associations with extreme-right sentiment, as well as comments on a recent political crisis within VVD. The respondents were asked to rate these statements on funniness and harshness. The funniest statements were selected for the experiment. From these tests, the two stimuli with the highest mean scores on funniness were selected in combination with either high or low scores in the assessment of ‘harshness’. These two statements differed from each other, with the personal attack being perceived as significantly more harsh (M = 56.26, SD = 32.72) than the party based attack (M = 34.64, SD = 22.72) t(29) = 3.14, p = 0.004 d = 0.78.
Manipulation check Humor
As a manipulation check, the respondents were asked multiple questions on how funny they thought a message was and whether the message was meant as a joke. Furthermore, they were asked if the message was (too) harsh. Lastly, they were asked to rate to what degree the message was aimed personally against the target and towards the party of the targetone a 1 to 10 scale.
Respondents in the humorous conditions (M = 5.1, SD = 2.38), compared to the non-humorous conditions (M = 3.5, SD = 2.36) thought the message was meant to be funny to a
2
14 significantly larger extent. t(74) = -2.92, p = 0.005, d = 0.68. The difference in funniness between the conditions approached significance with the funny condition having higher mean scores on funniness (M = 5.05, SD = 2.77) compared to the non-humorous condition (M = 3.94, SD = 2.39). t(74) = -1.84 p = 0.07, d = 0.42.
Harshness
Respondents perceived the harsh condition as significantly more harsh (M = 6.49, SD = 1.99) Than those in the soft condition (M = 5.44, SD = 2.09) t(74) = 2.24, p = 0.028, d = 0.51. Moreover, respondent reported the harsh condition as being significantly more ‘too harsh’ (M = 5.24, SD = 2.49) than those in the soft condition (M = 3.8, SD = 2.09) t(74) = 2.71, p = 0.007, d = 0.63. Moreover, people in the harsh condition gave higher scores to the questions whether the message attacked the target personally (M = 5.84, SD=2.5) than those in the soft condition (M = 3.33, SD = 2.43) t(74) = 4.43, p < 0.001, d = 1.02. These results indicate that the harsh condition was perceived as such, but also specifically as an personal attack
compared to the soft condition.
Measurements Dependent variables
Voting intention. To measure the effects of humor on negative campaigning, short term-electoral results were measured by asking respondents their intention to vote on either candidate. This was measured after the treatment by ‘how likely would you be to vote’ for either of the candidates. The response was measured on an eight point scale ranging from 1 (“Extremely unlikely”) to 8 (“Extremely likely”).
Evaluation of the candidates. To further measure assessment of the candidate after the stimulus, respondents were asked how well certain words described the politicians on a scale from 1 to 10, where 1 means “Not at all” and 10 means “Completely”. The words were:
15 warm, competent, professional, honest, inspiring, experienced, shares my values, easy to like, provides strong leadership. There variables were based off the candidate assessment index used by Fridkin and Kenney (2008), adding the words ‘warm’ and ‘competent’ cause these characteristics have been found to be important predictors of voting intention (Costa & Da Silva, 2015). This set of variables is used to create of scale of positive candidate evaluation for both the target and the sponsor. Both scales were highly reliable with d = 0.92 for sponsor evaluation and d =0.95 for target evaluation.
Political cynicism. Lastly, political cynicism was measured with an eight-item seven point Likert scale. Respondents were asked to what extent they agreed with statement such as “Almost all politicians will sell out their ideals or break their promises if it will increase their power” on a Likert scale ranging from 1 (“Strongly agree”) to 7 (“Strongly disagree”). For the sake of the analysis, this variable was recoded so that a higher score would mean higher cynicism. This scale was taken from Pattyn, Van Hiel, Dhont & Onraet (2012). The scale was reliable d = 0.83. On average the sample scored moderate on political cynicism with a mean score of M = 4.43. SD = 0.94.
Covariates
Some might suggest that the way the conditions are perceived has to do with knowledge of the political landscape on the one side – since the humurous conditions relate to the specific political reality of The Netherlands – and the tolerance towards harsh comments. Lastly, partisanship and ideology could influence the way people perceive the comments made in the conditions. To exclude the idea that these pre-existing attitudes influence the effect more than the stimulus, the following variables are taken into account.
Pre-existing political attitudes. To control for the influence of partisanship or
ideology, people were asked to rate themselves on a 1-10 scale with 1 being extreme left and 10 being extreme right. On average the sample was slightly more left oriented M = 4.2 SD =
16 1.58. Respondents were asked how much they liked VVD and D66 together with the other parties in the contemporary Dutch coalition. Lastly, the respondents were asked before exposure to the stimulus if they knew the politicians in the experiment and if yes, whether they liked them.
Moderating variables
Two of the previous mentioned variables, namely political knowledge and tolerance towards political incorrectness could influence the way in which the respondents perceive the
conditions. This is why these variables are used to check for moderating effects.
Political Knowledge. Political knowledge was included as a control variable because the basis of the humorous component was mocking VVD by which ‘getting the joke’ might be of importance for its effectiveness. Therefore, it was suspected that higher knowledge of the Dutch political landscape might result in higher perceived funniness of the humorous
component. Five multiple choice questions were asked. These questions were based on a scale by Snippenburg, Hagemann and Vettehen (2002) but altered towards the contemporary
situation and presenting multiple choice questions opposed to open questions. A ‘don’t know’ option was included. Respondents were asked who the current Prime Minister of The
Netherlands is, the current speaker of the house of representatives, the minister of foreign affairs, the largest party in the house of representatives and the amount of seats in the house.
The sample, on average, was low on political knowledge with a mean and mode of two questions right. This is probably due to the fact that the questions were aimed at the Dutch political landscape, whereas the sample did consists of 62% of respondents who did not spend most of their lives in The Netherlands.
Political correctness. In the pre-test, it became clear that people differ greatly on how funny they thought the statements were. It was apparent that respondents who perceived the
17 comments as very harsh, did not find it funny and vice versa. Although the pre-test was
conducted with a small convenience sample, it indicates individual difference between one’s perception of harshness, which could alter the effect of a harsh or soft effect. To account for this, a measure of tolerance against political incorrectness was included. This measure was taken from research by Straut and Blanton (2015) who created two sub-scales to predict perceived funniness of a joke. The subscale of emotion was used, because it focuses on the way people perceive political correctness, in contrast with the action subscale focusing on taking action against political correctness. For example, respondents were asked to what extent they agreed to the following statement: “I feel angry when a person says something politically incorrect”. The level of concern of political correctness was measured with four items on a Likert scale from 1 “Disagree extremely” to 7 “Agree extremely”. The scale is reliable with a Cohen’d d of d = 0.87. On average the sample was moderate in concern of political correctness with a mean score of M = 3.64 SD = 1.32.
Results
Main effect: evaluation of the sponsor
An independent t-test shows a strong significant backlash effect when attacking an opponent compared to the control group on candidate evaluation and voting intention. Respondents in the attack conditions evaluated the sponsor as significantly less positive (M = 3.96, SD = 1.66) compared to those who were not exposed to an attack (M = 4.62, SD = 1.53) t(119) = 2.16, p = 0.033, d = 0.41, in conjunction respondents were significantly more likely to vote for the sponsor when they were not exposed to an attack (M = 5.18, SD = 1.47) than when they were exposed to an attack (M = 4.3, SD = 1.63) t(119) = 2.97, p = 0.004, d = 0.57.
To test whether the use of a humorous message can mitigate the backlash effect, the effect of condition on sponsor evaluation and voting intention was tested with an independent samples t-test. The test showed no significant differences on intention to vote for the sponsor
18 but did show a moderate near-significant result for positive sponsor evaluation. Respondents in the humorous attack condition evaluated the sponsor as more positive (M = 4.28, SD = 1.67) than those in the non-humorous (M = 3.58, SD = 1.62) t(74) = -1.83 p = 0.071 d = 0.43, controlling for liking of the parties and political ideology. This (non-significant) finding suggests that the use of humor does soften the blow of backlash, this supports hypothesis 1. No results were found for difference in harshness of the attack.
Moderating effect: evaluation of the sponsor and knowledge/political correctness
There were no interaction effects between knowledge and the different conditions on voting intention for the sponsor or sponsor evaluation. Nor were there interaction effects for knowledge or political correctness on candidate evaluations and voting intention within the different conditions.
Main effect: evaluation of the target
The second hypothesis expects an less effective attack – which means higher evaluations for the target of the attack - when respondents were exposed to a humurous condition. An independent t-test showed no significant differences between the humorous and non humorous condition on voting intention and candidate evaluation.
However, when an attack is harsh, the humorous element increases the evaluation of the target significantly. Respondents exposed to an humorous harsh message evaluated the target significantly higher (M = 4.4, SD = 1.82) compared to respondents in the harsh
conditions that were not exposed to an humorous component (M = 3.23, SD = 1.53). t(35) = -2.04, p = 0.049 d = 0.69. This means that an humorous attack is less effective than an non-humorous attack in lowering target evaluations, when that attack is harsh. However, this effect did not remain significant upon including ideology, which implies a moderation effect
19 for ideology.
Comparing the soft humorous attack to the harsh humorous attacks shows a strong increase in target evaluation for the harsh attacks (M = 4.4, SD = 1.82) compared to a soft humorous attack (M = 2.94, SD = 1.58) t(40) = -2.76 p = 0.009 d = 0.86. However, this result did not remain significant when controlling for liking of the parties and ideology b* = .272, t
= 1.78 p = .083, which again implies a moderation effect for ideology. Both possible
moderations for ideology are discussed in the following paragraph.
Moderating effect: evaluation of the target and knowledge/political correctness
There are no moderating effects for political knowledge and political correctness on voting intention and the humorous and non humorous conditions. Results found within the harsh conditions did not remain significant when controlling for ideology. A near significant
interaction effects was found for knowledge and the harsh humorous versus the soft humorous condition b* = 0.448 t = 1.79 p = 0.081, controlling for ideology. This means that being in a harsh humorous attack will result in a higher target evaluation for those with high knowledge. For those with low knowledge, the harsh humorous condition resulted in slightly lower evaluation of the target. This interaction effect can be seen in Figure 1.
Figure 1: Interaction effect between harshness in humor and knowledge on target evaluation
1 2 3 4 5 6 7 8 9 10 Harsh+Humor Soft+Humor Low Knowledge High Knowledge
20 Looking into the moderating effect of ideology, a near significant interaction effect between the use of humor in the harsh conditions and left-right position was found b* = 0.47 t = 1.78 p = 0.084, which means that people exposed to the harsh and humorous condition that were more right-winged rated the target as higher than the left-wing candidates. This shows that the humorous attack is less successful than a non-humorous one amongst right-winged voters, when that attack is harsh. This moderating effect is shown in Figure 2.
Figure 2: Interaction effect between humor/no humor and ideology on target evaluation
No interaction effect between ideology and the soft versus harsh humorous conditions was found.
Hypothesis 2 is partly supported: the results show that the presence of humor can lessen the effect of an attack, or even strongly enhance support for the target. However, this was specifically the case for right-winged voters who were exposed to a harsh attack. In this case it should be taken into consideration that the target of the attack was a politician who belongs to a right-wing party.
1 2 3 4 5 6 7 8 9 10
Harsh+Humor Harsh+No humor
Left Right
21
Main effect: political cynicism
Hypothesis 3 aimed to investigate whether exposure to negative political messages can result in heightened political cynicism. An independent t-test showed no significant results for being exposed to an attack compared to no attack. An independent t-test showed no significant results for being exposed to an attack compared to no attack. The presence or absence of humor did also not yield a significant difference, nor did the harshness of the attack.
Moderating effects on political cynicism
A multiple regression was conducted to investigate whether an interaction between the conditions and political knowledge and political correctness was present. A significant interaction effects was present for political knowledge and being exposed to an attack. The interaction was a moderate predictor of cynicism, b* = 0.39, t = 2.21, p = 0.029, controlling for gender. This means that being exposed to negativity increases political cynicism for those who are high in political knowledge. On the contrary, exposure to negativity with low
knowledge levels resulted in a slight decrease in cynicism. This interaction effect is shown in Figure 3.There were no significant interactions between levels of political correctness and the conditions on political cynicism. Hypothesis 3 is supported, as humor does not matter for increased cynicism for those exposed to an attack.
22 Figure 3: Interaction effect of knowledge and negativity on political cynicism
Conclusion and discussion
The results can be summarized in four trends 1) a backlash effect on sponsor evaluations can be mitigated by using humor in an attack, but 2) humor can lessen the effectiveness of an attack and 3) knowledge matters for the way in which an effect is perceived both in terms of effectiveness and increased political cynicism 4) humor cannot soften the effect of negativity on political cynicism.
Considering these trends. The main conclusion is that the use of humor can indeed mitigate the backlash effect on the sponsor, but could bring along unwanted side-effects. As predicted, the use of humor makes an attack less effective than when that same attack is used without humorous component. However, harshness of the message should be taken into account, although the results could probably be influenced by pre-existing ideas about the politicians at play, making use of harsh attacks can heighten target evaluations. The heightened evaluation of the target could be due to the idea that citizens don’t like harsh attacks since they believe that a marsh message clashes with “social norms that prescribe fairness in interpersonal relations” (Carraro & Castelli, 2010, p. 636). and in reaction to reading such an attack symphonize with the politician being under attack. However, it should
1 2 3 4 5 6 7 No Attack Attack Low Knowledge High Knowledge
23 be taken into account that within this study, the harsh condition was a specific personal attack. It is therefore possible that this found effect has more to do with the personal attack than the harshness per se. This would be in line with Fridkin & Kenney (2008) who found that an uncivil message can still be an effective attack, as long as the message is relevant to the voter. This relevance might also be the reason why knowledge of the political spectrum matters for the effect of harshness and humor. For those with low levels of political knowledge within the Dutch context an harsh and humorous attack is successful in lowering target evaluations, however, the same attack will heighten evaluations for the target for those with high knowledge. This indicates that personal differences amongst voters matter for the way in which an attack is perceived, although in this study no proof was found that tolerance for political incorrectness mattered for the perception of negativity.
It should be taken into account that all the aforementioned effects were on sponsor and target evaluation, in terms of voting intention only the initial backlash effect of attacking an opponent was prevalent. The use of humor could not change these effects. This raises
questions to what extend the effectiveness will also achieve the desired result of earning more votes from the electorate.
In terms of the systematic effects of negativity, knowledge again matters for the effect of attacks. When one is knowledgeable about the political context in which an attack is voiced this negativity will increase political cynicism. However, for those with low knowledge a slight decrease in cynicism is present. This can be seen in line with the findings of (Yoon, Pinkleton & Ko, 2005) who found an increase in cynicism after exposure to negativity to be specifically prevalent for those with high political involvement. The use of a humorous component in the attacks could not alter these effects. It should be taken into account that political cynicism was measured directly after the stimulus, whereas this systemic effect can be seen as long-term attitudes towards politics, which are relatively stable (Banducci & Karp,
24 2003).
In short, this study shows that humor can be effective for politicians who resort in negative campaigning and want to dodge the negative effects on how the electorate perceived them. However, adding an humorous component to ones attack can result in unwanted side-effects such as heightened evaluation of the target, specifically amongst the targets adherents. Moreover, these effects were only prevalent for attitudes and do not account for voting intention. But, most importantly, humor does not work as a shield to prevent negativity from increased political cynicism for those with high levels of political knowledge.
Future research
Since these results are based on a convenient student-sample, the results should not be generalized beyond this demographic. However, Druckman and Kam (2009) show that student samples can still be acceptable in terms of external validity.
This study can be seen as a preliminary study on the moderating effects of humor on the effects negative campaigning. Further research should focus more on the topic of the attack, with previous research indicating that policy versus trait attacks can yield different results (Fridkin & Kenney, 2008) Additionally, an extra dimension of backlash should be incorporated in the studies of negative campaigning in which the heightening of target evaluation is taken into account as has been done by Lau, Sigelman and Rovner (2007) before. Furthermore, this study indicates that individual differences – in this case knowledge – are of importance for the way negativity is perceived. This is in line with recent research suggesting that specific personal traits influence the effect of negativity on political
engagement (Weinschenk & Panagopoulos, 2014). Since this study has found that knowledge matters both for the effectiveness of an attack as well as on political cynicism, political
25 knowledge as well as character characteristics could be incorporated to shed further light on the effects of negative campaigning on specific parts of the electorate.
In terms of measurement, humor remains a difficult variable to implement in research, since what one perceives as funny differs greatly. Within the study, respondents only got to read one sentence that could be perceived as funny whereas a series of humorous components – such as is possible in a campaign movie – is probably more powerful in evoking laughter.
Future research could follow the steps of Baumgartner (2013), using real funny-intended campaign footage, or Boukes et al., (2015) who worked together with a professional satirist to create content specifically for the study. Lastly, since this study was done within a Dutch context, but with student from all over the world, future research could take specific cultural elements in terms of tolerance for negative campaigning and humor into account.
26 References
Agger, R. E., Goldstein, M. N., & Pearl, S. A. (1961). Political cynicism: Measurement and meaning. The Journal of Politics, 23(3), 477-506. https://doi.org/10.2307/2127102
Ansolabehere, S. & Iyengar, S (1995). Going negative: How political advertising shrinks and
polarizes the electorate. New York: Free Press.
Ansolabehere, S., Iyengar, S., Simon, A., & Valentino, N. (1994). Does attack advertising demobilize the electorate?. American political science review, 88(4), 829-838. https://doi.org/10.2307/2082710
Baumgartner, J. C. (2013). Internet Political Ads in 2012: Can Humor Mitigate Unintended Effects of Negative Campaigning?. Social Science Computer Review, 31(5), 601-613. https://doi.org/10.1177/0894439313490399
Banducci, S. A., & Karp, J. A. (2003). How elections change the way citizens view the political system: Campaigns, media effects and electoral outcomes in comparative perspective. British Journal of Political Science, 33(3), 443-467.
https://doi.org/10.1017/S000712340300019X
Becker, A. B. (2012). Comedy types and political campaigns: The differential influence of other-directed hostile humor and self-ridicule on candidate evaluations. Mass
Communication and Society, 15(6), 791-812.
https://doi.org/10.1080/15205436.2011.628431
Bippus, A. (2007). Factors predicting the perceived effectiveness of politicians' use of humor during a debate. International Journal of Humor Research, 20 (2), 105-121.
27 Boukes, M., Boomgaarden, H. G., Moorman, M., & de Vreese, C. H. (2015). At odds:
Laughing and thinking? The appreciation, processing, and persuasiveness of political satire. Journal of Communication, 65(5), 721-744. https://doi.org/10.1111/jcom.12173
Carraro, L., & Castelli, L. (2010). The implicit and explicit effects of negative political campaigns: Is the source really blamed?. Political Psychology, 31(4), 617-645. https://doi.org/10.1111/j.1467-9221.2010.00771.x
Carraro, L., Gawronski, B., & Castelli, L. (2010). Losing on all fronts: The effects of negative versus positive person‐based campaigns on implicit and explicit evaluations of
political candidates. British Journal of Social Psychology, 49(3), 453-470. https://doi.org/10.1348/014466609X468042
Caufield, R. P. (2008). The influence of “infoenterpropagainment”: Exploring the power of political satire as a distinct form of political humor. In J. Baumgartner & J.S. Morris (Eds.) Laughing matters: Humor and American politics in the media age, 3-20. Routledge: New York/London
Costa, P., & Ferreira da Silva, F. (2015). The impact of voter evaluations of leaders’ traits on voting behaviour: Evidence from seven European Countries. West European
Politics, 38(6), 1226-1250. https://doi.org/10.1080/01402382.2015.1004231
Compton, J. (2012). More than laughing? Survey of Political Humor Effects Research. In J. Baumgartner & J.S. Morris (Eds.) Laughing Matters: Humor and American Politics in
the media age 63-88. Routledge: New York/London.
Dardis, F. E., Shen, F., & Edwards, H. H. (2008). Effects of negative political advertising on individuals' cynicism and self-efficacy: The impact of ad type and message
exposures. Mass Communication & Society, 11(1), 24-42. https://doi.org/10.1080/15205430701582512
28 Davis, J. L., Love, T. P., & Killen, G. (2018). Seriously funny: The political work of humor
on social media. New Media & Society, https://doi.org/10.1177/1461444818762602
Druckman, J. N., & Kam, C. D. (2011). Students as experimental participants. A defense of the ‘narrow data base’. In Druckman, J. N., Green, D. P., Kuklinski, J. H., & Lupia, A. (Eds.). Cambridge handbook of experimental political science. 41-57. Camdrige: Cambridge University Press
Duffy, M. E., & Page, J. T. (2013). Does political humor matter? You betcha! Comedy TV's performance of the 2008 Vice Presidential debate. The Journal of Popular
Culture, 46(3), 545-565. https://doi.org/10.1111/jpcu.12038
Fernandes, J. (2013). Effects of negative political advertising and message repetition on candidate evaluation. Mass Communication and Society, 16(2), 268-291.
https://doi.org/10.1080/15205436.2012.672615
Fridkin, K. L., & Kenney, P. J. (2008). The dimensions of negative messages. American
Politics Research, 36(5), 694-723. https://doi.org/10.1177/1532673X08316448
Garramone, G. M. (1985). Effects of negative political advertising: The roles of sponsor and rebuttal. Journal of Broadcasting & Electronic Media, 29(2), 147-159.
https://doi.org/10.1080/08838158509386573
Geer, J. G. (2012). The news media and the rise of negativity in presidential campaigns. PS:
Political Science & Politics, 45(3), 422-427.
https://doi.org/10.1017/S1049096512000492
Harris, M. K. (2009). The Political Application of Humor. Syracuse University Honors
Program Capstone Projects, 497, 1-41. Retrieved from:
29 Innocenti, B., & Miller, E. (2016). The persuasive force of political humor. Journal of
Communication, 66(3), 366-385. https://doi.org/10.1111/jcom.12231
Jasperson, A. E., & Fan, D. P. (2002). An aggregate examination of the backlash effect in political advertising: The case of the 1996 US Senate race in Minnesota. Journal of
Advertising, 31(1), 1-12. https://doi.org/10.1080/00913367.2002.10673656
Kaid, L. L., & Postelnicu, M. (2008). Air amusement versus web wit: Comparing the use of humor in 2004 political advertising on television and the Internet. In J. C Baumgartner & J. S. Morris (Eds.), Lauging matters: Humor and American politics in the media
age,117–130. New York, NY: Routledge.
Krupnikov, Y., & Bauer, N. M. (2014). The relationship between campaign negativity, gender and campaign context. Political Behavior, 36(1), 167-188.
https://doi.org/10.1007/s11109-013-9221-9
LaMarre, H. L., Landreville, K. D., Young, D., & Gilkerson, N. (2014). Humor works in funny ways: Examining satirical tone as a key determinant in political humor message processing. Mass Communication and Society, 17(3), 400-423.
https://doi.org/10.1080/15205436.2014.891137
Lau, R. R., Sigelman, L., & Rovner, I. B. (2007). The effects of negative political campaigns: a meta‐analytic reassessment. Journal of Politics, 69(4), 1176-1209.
https://doi.org/10.1111/j.1468-2508.2007.00618.x
Lau, R. R., & Pomper, G. M. (2004). Negative campaigning: An analysis of US Senate
elections. Oxford: Rowman & Littlefield Publishers.
Meyer, J (2009). Ronald Reagan and Humor: A politician's velvet weapon. Communication
30 Nabi, R. L., Moyer-Gusé, E., & Byrne, S. (2007). All joking aside: A serious investigation
into the persuasive effect of funny social issue messages. Communication
Monographs, 74(1), 29-54. https://doi.org/10.1080/03637750701196896
Nai, A., & Seeberg, H. B. (2018). A series of persuasive events. Sequencing effects of negative and positive messages on party evaluations and perceptions of
negativity. Journal of Marketing Communications, 24(4), 412-432.
Pattyn, S., Van Hiel, A., Dhont, K., & Onraet, E. (2012). Stripping the political cynic: A psychological exploration of the concept of political cynicism. European Journal of
Personality, 26(6), 566-579. https://doi.org/10.1002/per.858
Plasser, F. (2000). American campaign techniques worldwide. Harvard International Journal
of Press/Politics, 5(4), 33-54. https://doi.org/10.1177/1081180X00005004003
Roese, N. J., & Sande, G. N. (1993). Backlash effects in attack politics. Journal of Applied
Social Psychology, 23(8), 632-653. https://doi.org/10.1111/j.1559
1816.1993.tb01106.x
Rozin, P., & Royzman, E. B. (2001). Negativity bias, negativity dominance, and contagion. Personality and Social Psychology Review, 5(4), 296-320. https://doi.org/10.1207/S15327957PSPR0504_2
Shapiro, M. A., & Rieger, R. H. (1992). Comparing positive and negative political advertisin on radio. Journalism Quarterly, 69(1), 135-145.
https://doi.org/10.1177/107769909206900111
Shifman, L., Coleman, S., & Ward, S. (2007). Only joking? Online humour in the 2005 UK general election. Information, Community and Society, 10(4), 465-487.
31 Schenck‐Hamlin, W. J., Procter, D. E., & Rumsey, D. J. (2000). The influence of negative
advertising frames on political cynicism and politician accountability. Human
Communication Research, 26(1), 53-74. https://doi.org/10.1111/j.1468
2958.2000.tb00749.x
Strauts, E., & Blanton, H. (2015). That’s not funny: Instrument validation of the concern for political correctness scale. Personality and Individual Differences, 80, 32-40.
https://doi.org/10.1016/j.paid.2015.02.012
Thorson, E., Ognianova, E., Coyle, J., & Denton, F. (2000). Negative political ads and negative citizen orientations toward politics. Journal of Current Issues & Research in
Advertising, 22(1), 13-40. https://doi.org/10.1080/10641734.2000.10505099
Tsakona, V. (2009). Humor and image politics in parliamentary discourse: A Greek case study. Text & Talk-An Interdisciplinary Journal of Language, Discourse
Communication Studies, 29(2), 219-237. https://doi.org/10.1515/TEXT.2009.010
Van der Meer, T. (2017) Political Trust and the “Crisis of Democracy”. In: Oxford Research
Encyclopedia of Politics. Retrieved from: http://politics.oxfordre.com/
Van Snippenburg, L.B. Hagemann, C.P.M, Hendriks Vettehen, P.G.J. (2002) Politieke kennis en mediagebruik in Nederland. Mens en Maatschappij, 77(1), 65-89. Retrieved from: http://repository.ubn.ru.nl/bitstream/handle/2066/62307/62307pub.pdf
Walter, A. S., & Nai, A. (2015). Explaining the use of attack behaviour in the electoral battlefield: A literature overview. In Nai, A. & Walter (Eds.) New perspectives on
negative campaigning: Why attack politics matters. 115-128. Colchester: ECPRS
32 Weinschenk, A. C., & Panagopoulos, C. (2014). Personality, negativity, and political
participation. Journal of Social and Political Psychology, 2(1), 164-182. https://doi.org/10.5964/jspp.v2i1.280
Yoon, K., Pinkleton, B. E., & Ko, W. (2005). Effects of negative political advertising on voting intention: An exploration of the roles of involvement and source credibility in the development of voter cynicism. Journal of marketing communications, 11(2), 95-112. https://doi.org/10.1080/1352726042000315423
33 Appendix A. Treatments English
B.1: No Attack (control)
34
A.3: Humorous Soft
35
36 Appendix B Treatments Dutch
B.1: No Attack (control)
37
B.3 Humorous Soft
38
