Waking the opponents : Anger as underlying mechanism in the moderated mediation dynamic of reversed mobilization

(1)

Waking the opponents

Anger as underlying mechanism in the moderated mediation dynamic of reversed mobilization

Author: Frank Henkus

Student ID: 11431687

Master Thesis

Graduate School of Communication

Master’s programme Communication Science

Supervisor: Joost van Spanje

(2)

2

Abstract

The reversed mobilization dynamic suggests that during referendum campaigns, people with strong negative attitudes towards a referendum proposal will be mobilized after exposure to an article framed positively towards that proposal. So far the only underlying mechanism that researchers have found is perceived risk after this positive frame exposure. This study presents a new moderated mediation framework of the reversed mobilization dynamic, with “anger” as a second mediator next to risk perception, and external political efficacy as a second moderator on the b-path of the mediation. To test this, an experiment (N = 86) was conducted in a Dutch advisory referendum context. We found significant indirect effects for framing on turnout intention via both risk perception and anger, but only for people with both extremely negative attitudes to the proposal and high external efficacy. Our findings confirm the existence of a reversed mobilization dynamic, but the lack finding any direct effects poses new questions to how this dynamic actually works.

Introduction

During election campaigns politicians and their campaign leaders have an obvious interest in getting people to the voting booth in order to bring them a good electoral result. They attempt to strategically frame their message in order influence both the opinion and the voting behavior of citizens. During referendums mobilization and involvement occur less naturally as compared to national elections, which make referendum campaigns unusually influential. Therefore it is of specific interest to understand the dynamics behind mobilization of voters during referendum campaigns.

(3)

3

It is a given that politicians who support a referendum proposal always attempt to convince citizens to vote for their side, by spreading a message in which they emphasize the arguments in favor of the proposal. Striking is a recently developed idea which states that exposure to a positively framed message about a referendum proposal, can in fact mobilize opponents of that proposal. According to this idea, citizens with strong negative attitudes towards such a proposal will not be convinced by that positive message. On the contrary, they are

becoming more aware of the risks that will occur if the proposal passes, and as a consequence they will be extra triggered to go out and vote against it. This dynamic, which is called “reversed mobilization” (Schuck & de Vreese, 2009), can have very important implications for the framing of referendum campaigns. It implies an unintentional effect of spreading a certain campaign message, namely the mobilization of those you do not want to mobilize.

The very limited number of studies that have tested the reversed mobilization dynamic, concluded that this effect can be explained by an increased perception of risk among opponents of a proposal. According to Schuck & de Vreese (2012), risk perception serves as a mediator in the effect of positive referendum proposal framing on the mobilization of opponents. Next to that, these researchers concluded that this opposite framing effect only occurred for people with high external political efficacy.

Schuck and de Vreese advised their colleagues to test emotions as possible mediators in the effect of framing on mobilization. Past studies already demonstrated the mediating role of emotions like anger and enthusiasm in the effect of framing on opinions. (e.g. Lecheler et al., 2015) and behavior (e.g. Bartholomé, 2017) . It is plausible that opponents of a referendum proposal will experience negative emotions after exposure to a message in favor of such a proposal. In turn, such negative emotions could very likely lead to an increased motivation to go to the voting booth, especially for those with higher levels of political efficacy. Where Schuck

(4)

4

and de Vreese (2012) saw a moderating role of efficacy in the a-path of the mediation, which is the effect of a framed message on the mediator, we think it is more likely that this variable will affect the b-path of the mediation, namely the effect of the mediator on the intention to vote.

All in all, negative emotions are an interesting possible second explaining factor in the reversed mobilization dynamic, but in fact this has never been tested. Hence, this research aims to answer the following central research questions:

1. To what extent will a positive referendum proposal message lead to the mobilization of opponents, and what role does external political efficacy play in this dynamic?

2. Can this effect be mediated by either risk perception or negative emotions?

This study will test and expand the reversed mobilization dynamic through an experiment that was developed in the run-up to the Dutch advisory referendum about the possible implementation of a new trawling surveillance law. Substantial mobilization in a Dutch advisory referendum is very important, since a voter turnout of 30 percent is necessary in order for the results to be valid (“Advisory referendum”, n.d.). The new trawling surveillance law expands the power of Dutch security and intelligence agencies, which makes it easier for them to tap information like internet traffic on a large scale. Mols and Janssen (2017) conducted a content analysis about an earlier government surveillance debate in the Netherlands, which followed after the National Security Agency (NSA) leaks of Edward Snowden. They concluded that the most common frame in favor of strong government surveillance was the so-called “End justifies the means frame”, which emphasized the possible threat of terrorism. Our experiment will test if, according to the reversed

(5)

5

mobilization dynamic, exposure to this frame has the effect of bringing opponents of extra government surveillance to the voting booth.

Theoretical Background

Reversed Mobilization

In the run-up to a referendum, politicians strategically frame the political issue at stake in order to influence the attitude and behavior of citizens in their favor. According to Entman (1993), “to frame is to select some aspects of a perceived reality and make them more salient in a

communicating text, in such a way as to promote a particular problem definition, causal

interpretation, moral evaluation, and/or treatment recommendation” (p. 52). If frames contain a certain valence, which means they emphasize either positive or negative consequences of an issue, they have the ability to influence public support for policy measures (De Vreese & Boomgaarden, 2003).

When it comes to the effects of framing, there is some empirical evidence that negatively framed messages are more likely to influence behavior as compared to positively framed

messages. Lau (1985) for instance concluded that exposure to negative information had a bigger impact on the evaluation of presidential candidates than comparable positive information. Moreover, studies show that negative attitudes also have a bigger impact on behavior as compared to positive attitudes. Kernell (1977) found that citizens who disapprove of the president’s job performance are more motivated to vote during midterm elections than citizens who approve. Wattenberg and Brians (1999) conclude that negative presidential campaign advertisements increase political participation. The underlying psychological mechanism of this negativity effect can be found in the prospect theory of Kahneman and Tversky (1979), which

(6)

6

basically states that people are more likely to avoid risks than to seek gains when the same choice is presented in different ways.

It is important to note, however, that not every person sees the same information as negative. Predispositions are of big importance in the perception of messages. According to Sigelman and Kugler (2003) the perception of a campaign tone can vary widely between voters. They state that while most researchers treated negativity fully as a campaign attribute, it is actually important that the citizens whose behavior is being explained actually perceive this negativity. This also means that positive messages can be perceived as negative. In the context of this study, people with a negative pre-existing attitude towards a referendum proposal will perceive a positively framed message about this proposal as negative. In short, this study argues that people with such negative attitudes will be more triggered to go out and vote after

confrontation with a positively framed message. This is what we call the reversed mobilization effect.

The occurrence of reversed mobilization is depending on pre-existing attitudes towards an issue. If these attitudes are strong enough, it can make them resistant to messages that oppose these attitudes. Moreover, exposure to such messages can lead to increased support for their pre-existing attitude. Bizer & Petty (2005) found that people’s preferences are more resistant to persuasion if that preference is conceptualized as opposing something, than as supporting something else. Meffert et. al (2006) discovered that people with a strong candidate preference during a congressional election campaign, were more likely to consume negative information about the preferred candidate, but as a result developed more positive attitudes towards that candidate. They used the negative information to counterargue it, which suggests engagement in motivational information processing. We expect that also in a referendum context, if existing

(7)

7

attitudes are strong enough, counterarguments can lead to a strengthening of these attitudes and subsequently to reversed mobilization.

H1: Confrontation with news framing in favor of a referendum proposal increases turnout

intention for those with higher levels of opposition towards this proposal.

Risk Perception

As previously mentioned, people rather avoid taking risks compared to seeking gains. This risk-aversive attitude is also likely to be found in referendums. Kirchgässner and Schulz (2005) found in a Swiss context that it is easier to mobilize citizens against a referendum proposal than in favor of it. An explanation for this could be that the consequences are normally more uncertain when a proposal would pass, than when it would be rejected and the status quo would be preserved (Hobolt, 2006). The risk for people who oppose a proposal is higher, because a positive outcome of a referendum means an undesired change of the status quo (Schuck & de Vreese, 2009).

The context of the Dutch trawling surveillance law referendum is a bit more complex. This law was already approved by the Dutch parliament, and a referendum was only organized because enough opponents of the law submitted a referendum requests.1 In this light, you could argue that approval of the law after the referendum would mean the maintenance of the status quo, and that rejection of the law would mean an undesired disruption of that status quo for people who support the law. Given the fact that the implementation of the law was postponed until after the referendum, we see the approval of the law as a change of the status quo. This implies the risk of reduced privacy for opponents of the law. If the law will not pass, the status

1_{According to the Dutch Advisory Referendum Act, citizens can organize an advisory referendum to be held on} newly adopted Acts and Treaties. Therefore, at least 300,000 valid requests are needed (“Advisory referendum”, n.d.).

(8)

8

quo remains, which is less risk-inducing. It is therefore more likely that a reversed mobilization effect occurs for citizens who oppose this law than for citizens who support it.

In order for people to react in a risk-aversive way, it is a necessary step to actually perceive a political message as risk-inducing. For this reason it is logical that Schuck & de

Vreese (2012) see the concept of risk perception as the primary explanatory factor of the reversed mobilization effect. According to these researchers, risk perception is a combination of people’s perception of the severity of a risk, combined with their perceived susceptibility to it. In this study, we test risk perception as a mediator between the effect of a positive referendum proposal message on vote intention of people with negative pre-existing attitudes towards this proposal.

However, it is not merely the pre-existing attitude towards a message that can predict how people behave as a consequence of risk perception. In a health communication context, Rimal and Real (2003) found that risk perception only leads to mobilization when their personal efficacy is strong. Efficacious people see the risk as a challenge that they are willing to accept, while inefficacious people are more likely to demotivated, which will result in avoidance. Therefore we will follow Schuck & de Vreese (2012) by expecting that reversed mobilization only will occur among those with higher external political efficacy. While internal efficacy focuses on citizens’ feelings of personal competence to understand and participate in politics, external efficacy is about “citizens’ perceptions of responsiveness of political bodies and actors to citizens’ demands” (Morrell, 2003, p. 590). People who think that political actors respond well to citizens’ demands, should be encouraged to vote. Contrary to Schuck & De Vreese (2012), we expect this extra moderator to affect the b-path of the mediation model. Following Rimal and Real (2003), we hypothesize that the level of efficacy does not affect the level of risk perception, but rather influences how people behave as a consequence of that risk. The above presented arguments can be summarized in the following hypotheses:

(9)

9

H2: Confrontation with news framing in favor of a referendum proposal increases turnout

intention those with both higher levels of opposition towards this proposal and higher levels of

external political efficacy.

H3a: Confrontation with news framing in favor of a referendum proposal increases risk

perception among those with higher levels of opposition towards this proposal.

H3b: Higher levels of risk perception, in turn, lead to increased turnout intention among those

with both higher external political efficacy and higher levels of opposition towards the

referendum proposal.

Emotions

This century a number of communication studies have found that framing can have a substantial effect on emotions of the reader (e.g. Kim et al., 2011; Nabi, 2003). Gross & D’Ambrosio (2004) concluded that this effect is not the same for everybody; it rather depends on individual

predispositions. The researchers conclude that this is in line with appraisal theory, which states that people don’t randomly experience emotional reactions, but that it’s the consequence of their cognitive evaluations of events, based on the beliefs and opinions they hold at that particular time. This theory is relevant for the reversed mobilization dynamic, since this also assumes that the effect of framing depends on pre-existing attitudes and beliefs.

(10)

10

Not only do frames have an impact on emotions, emotions in turn can have an impact on someone’s attitude and behavior. Aarøe (2011) concluded that the effect of episodic and thematic news frames on opinion formation was depending on the intensity of the emotional reactions after reading a news article. Gross (2008) found a mediating role of sympathy and pity in the effect of episodic framing on opinions about minimum sentencing. Other studies see indirect framing effects via emotions on political participation (e.g. Valentino et al., 2011; Bartholomé, 2017).

Which emotions then, could be relevant in the reversed mobilization context? The most basic assumption is that negative frames lead to negative emotions, while positive frames lead to positive emotions (Valentino et al., 2011). However, as we mentioned before, what is positive and what is negative is all about the predispositions and perceptions of the reader. Because this research focuses on a negative perception of a positive referendum proposal, it seems logical to test negative emotions. According to Nabi (2010) it is useful to study discrete emotions in communication science instead of grouping them into a particular valence, because “it better captures the implications of motivational forces on human action and interaction” (p. 158).

What follows is the question which particular negative emotion is most likely to increase turnout intention in a referendum. The negative emotions that have been studied most commonly are fear and anger. Lerner & Keltner (2001) found that fear leads to risk-aversive behavior, while anger leads to more risk-seeking intentions. Since I already concluded that the intention to vote against a referendum proposal is a form of risk-aversive behavior, you could assume that fear would be the right explanatory emotion of reversed mobilization. Risk-averse behavior as a consequence of fear however, is more likely to be a form of retreat and avoidance (Lecheler et al., 2013). Also, fear can lead to a low willingness to process information (Nabi, 1999). These consequences are opposite to mobilization and participation.

(11)

11

Not only does research suggest that message-induced anger leads to more attention and more careful information processing (Turner, 2007), anger also leads to more confrontational behavior (Lecheler et al., 2013). Valentino et al. (2011) tested the effect of different emotions on political participation in three studies, and most importantly concluded that in all contexts anger not only had a positive effect on what the researchers called “cheap participation” variables like attending a rally, but also on “costly participation” variables like donating money to a campaign. In the context of government monitoring, which is very similar to the trawling surveillance law context, Best and Krueger (2011) found that anger was positively associated with political participation, while fear was negatively associated. Since these results suggest that anger mobilizes different forms of participation, we expect that anger can affect mobilization in the referendum context as well. Also regarding the first step in the mediation model, it is plausible that opponents of the referendum proposal are more likely to feel anger than supporters of the proposal after reading a positive frame. This brought me to the following hypotheses:

H4a: Confrontation with news framing in favor of a referendum proposal increases anger among

those with higher levels of opposition towards this proposal.

H4b: Higher levels of anger, in turn, lead to increased turnout intention among those with both

higher external political efficacy and higher levels of opposition towards the referendum

proposal.

(12)

12

All our hypotheses can be summarized in the framework that is presented in Figure 1. The dependent variable of this study is “pro vs. neutral referendum proposal framing”, which affects our dependent variable “turnout intention”, but only when the moderator “attitude towards the proposal” has been taken into account (H1). Moreover, we expect that this reversed mobilization effect is also depending on higher levels of “political efficacy” (H2). A positive frame increases “risk perception” for people with more negative attitudes (H3a), which in turn leads to a higher turnout intention for people with more negative attitudes and higher efficacy (H3b). Finally, we expect anger to serve as a mediator, again with attitude as a moderator on the a-path (H4a), and both attitude and efficacy as moderators on the b-path (H4b).

Figure 1 The moderated mediation model of this study, based on Model 71 of Hayes (2013).

(13)

13

In order to test the reversed mobilizing hypotheses, an online single-factor, posttest only,

between-subjects experimental study was conducted. Participants were randomly assigned to one of two conditions. The first condition concerned a news article framed in favor of the trawling surveillance law. The second condition was a similar news article, but framed neutrally towards this trawling surveillance law.

Sample

A total of 228 participants completed the experiment of this study. The participants were

recruited through both convenience and snowball sampling. One participant was dropped because he or she indicated to be younger than 18 years old. Three participants were dropped from the study, since they declared to be non-Dutch. Because we want to study voting behavior in a Dutch context, both groups were not useful for our experiment since they are ineligible to vote. Given the length of our experiment, 3 respondents who completed the experiment in less than 5 minutes and 3 respondents who needed more than 60 minutes were excluded. Because we are only

interested in opponents of the trawling surveillance law, participants who scored lower than a 5 on our combined seven-point “attitude towards trawling surveillance law” item were excluded.2 This lead to a final sample of 86 respondents. The age of the participants varies from 18 to 77 years (M = 29.37, SD = 12.58), and 50% of the sample is male. 63,9% of the participants have a highest completed education of Bachelor-level or higher. Participation was mostly voluntary, but it is important to note that participants from surveystudent.nl could earn credits so they could publish their own research on this website.

(14)

14 Procedure

All data were collected between December 13, 2017 and January 14, 2018. This was approximately 3 months before the trawling surveillance law referendum was held. The

participants were mostly sampled through a link to the experiment that the researcher posted on his personal Facebook page. Furthermore, the researcher approached his personal network face-to-face to recruit participants. Finally, some respondents were recruited via the website

surveystudent.nl.

For this experiment the online survey program Qualtrics was used. Firstly, the participants had to complete a pretest questionnaire, in which demographic variables and the moderator variables were measured. As a buffer, participants then had to read an article about the possible prohibition of smartphone use on bikes.3 The article was slightly shortened in order to prohibit any fatigue among respondents. Also a few questions about this article had to be answered, which were not used in my analysis. Next, they were exposed to a news article that included one of the two mentioned experimental conditions. Finally, the participants had to complete a posttest questionnaire, in which firstly the manipulation check and mediator variables were measured, and after that also the dependent variable, which is turnout intention. In total, it took the participants between 5 and 43 minutes to complete the experiment (Median = 10.45, M = 14.48, SD = 19.21). The complete experimental survey (in Dutch) can be found in the appendix.

Stimulus

The stimulus of this research consists of a Dutch news article that is framed either neutrally or positively towards the trawling surveillance law. For this I adjusted an online news article from the website of Dutch broadsheet newspaper NRC Handelsblad that was posted on November 1st,

3

The original article was retrieved from the website of the Dutch public service broadcaster NOS: https://nos.nl/artikel/2206406-fietsend-appen-mag-misschien-niet-meer-maar-hoe-gevaarlijk-is-het.html

(15)

15

2017.4 The purpose of the original article was to inform readers about the continuation of the referendum, as a consequence of the necessary 300,000 signatures that were collected by the promoters of the referendum. First of all, I changed the name of the writer of the article, so participants could not recognize him. Also, a number of political expressions that could influence the participants’ view of the issue have been deleted. A sentence that mentioned the minister of Internal Affairs has been left out, as well as paragraphs that discussed the use of the referendum and the necessity of politicians to listen to the outcome. Also the fact that a the host of the popular Dutch satire show Zondag met Lubach promoted the referendum proposal, was left out because of the possible influence of satire on the public (Lee & Kwak, 2014). Because of these adjustments, the focus of the article was limited to the actual content of the law. This way the neutral frame condition was constructed.

The positive frame condition was based on the research of Mols and Janssen (2017). They conducted a content analysis on the Dutch government surveillance debate that followed the National Security Agency (NSA) leaks of Edward Snowden in 2013. In their study they analyzed written content both online and offline, professional and non-journalistic. The researchers

identified six frames in this debate, varying from an End justifies the means frame to an

“Orwellian dystopia” frame. The End justifies the means-frame appeared to be very useful as a positive frame about in the trawling surveillance law referendum debate. Not only is this frame the most clear in her point of view (positive towards the law); this frame was also the most commonly used as the positive point of view according to the mentioned content analysis.

In order to adjust the original article into the End justifies the means frame, I added quotes of a fictional expert to the article. Given the judicial nature of the debate (right to privacy), the

4

The original article was retrieved from https://www.nrc.nl/nieuws/2017/11/01/referendum-over-inlichtingendiensten-gaat-definitief-door-a1579482

(16)

16

expert is a jurist but also a safety expert. In the article the expert brings up the key components of the End justifies the means frame (Mols & Janssen, 2017). First he brings up the problem

definition, which is that 100% privacy and 100% safety do not go together. Also he mentions the cause, namely the threat of terrorism. The consequence this expert states is that citizens have to give up a small amount of their privacy for the safety of society. He appeals to the emotional biases that loss of privacy is a small sacrifice compared to the danger of a terrorist attack, and that ‘it is important to trust the government’. The full articles can be found in the appendix.

Measures

Manipulation check.

The manipulation check for this study consisted of three statements about the article they just read, to which the participants could respond on a seven-point answer scale (1 = strongly

disagree, 7 = strongly agree). The first statement was: “the article showed arguments in favor of the new law”. The second statement was: “the article did not show any arguments about the new law”. The third statement was: “in the article a jurist was quoted”. After reversing the second statement the three items were combined into a single item measure (M = 3.87, SD = 1.79), for which participants that were exposed to the End justifies the means frame need to indicate higher scores in order for the manipulation to be successful. To measure this an independensamples t-test was conducted. Since the mean score on our combined manipulation check variable was significantly higher for participants exposed to the positive frame (M = 4.99, SD = 1.24) than for participants exposed to the neutral frame (M = 2.26, SD = 1.09), t(84) = -10.53, p < .001, 95% CI [-3.25, -2.22], the manipulation was successful .

(17)

17

Attitude towards the trawling surveillance law.

In order to test if we can speak about reversed mobilizing, I measured the attitude towards the trawling surveillance law as a moderator.5 Because I want to remind the participants what the trawling surveillance law means, but also do not want to influence them, minimal information about the law was first presented. The information consists of the following sentences: “The new trawling surveillance law expands the powers of national security and intelligence agencies. The law makes it easier for these services to tap information like internet traffic on a large scale”. After these sentences, the participants were asked two questions. The first question was: “what is your opinion about this law?” This question could be answered on a seven-point scale (1= highly against the law, 7= highly in favor of the law). On the next page, the participants had to answer the second question: ‘how do you feel about the law that expands the powers of intelligence agencies?’ Also this question could be answered on a seven-point answer scale (1 = very positive, 7 = very negative). In order to combine the two items into a single item measure, the first

question was recoded so that higher scores indicate higher levels of opposition. The Cronbach’s alpha of the two items was .84, so the two items were combined into a single item (M = 4.26, SD = 1.54). After filtering out the cases that proved to be not useful (i.e. scores lower than 5), a new five-point item was created (M = 2.50, SD = 1.44). Higher scores on this item indicate higher levels of opposition towards the law.

Risk perception.

Our first moderator variable is risk perception. More specifically, I measure to what extent participants perceive the trawling surveillance law to be a risk. The measure was slightly based on the research of Schuck and de Vreese (2012), and consists of five questions on which

(18)

18

participants had to answer on a seven-point answer scale. These questions both address possible susceptibility and severity. First participants were asked how much they were afraid of the consequences of the trawling surveillance law (1= not afraid at all, 7= very afraid). Secondly, participants were asked: ‘how big do you consider the chance that passing this law would have negative consequences?’ (1 = very small chance, 7 = very big chance). After that, participants had to indicate how much they were worried that the new law would affect the privacy of

consecutively themselves, their family and friends, and innocent citizens (1 = not at all worried, 7 = very worried). Because the five items combined proved to be reliable (α = .84), a single-item measure was created with higher scores indicating a higher risk perception (M = 4.63, SD = 1.26).

External political efficacy.

Our second moderator variable is external political efficacy. Following Morrell (2003), this variable consists of four statements, on which participants had to indicate their level of agreement on a five-point answer scale (1= strongly agree, 5= strongly disagree). The four statements were: ‘In recent years, the government has payed substantial attention to the opinion of the people’, ‘Public officials don't care much what people like me think’, ‘People like me have no influence on decisions made by politicians in the parliament’ and ‘Thanks to elections, public officials pay more attention to what people think’. The first and last item were recoded so for all items higher scores mean more external efficacy. Because the four items combined were reliable (α = .73), the four efficacy items were combined into a single item measure (M = 3.38, SD = .85). Higher scores indicate more external political efficacy.

Anger.

To measure our emotion mediator variable, participants were asked to what extent they feel certain emotions after reading the stimulus article. Based on the research of Lecheler et al.

(19)

19

(2015), I measure the following emotions on a seven-point answer scale (1=not at all, 7= very much): contentment, compassion, enthusiasm, hope, anger, fear, and sadness. Apart from anger (M = 3.40, SD = 1.91), these emotions are used as filler items in this study.

Turnout intention.

The turnout intention in the trawling surveillance law referendum is the dependent variable of this study.6 After answering a few filler items about the trawling surveillance law referendum, participants needed to answer the following question: “thinking about the chance that you would vote, how would this be expressed in percentages?” Participants could answer on an 11-point answer scale (0% - 100%), so high scores indicate a higher turnout intention (M = 72.67, SD = 26.54).

Results

To test our hypotheses, model 71 of PROCESS (Hayes, 2013) was used with 5,000 bootstrap samples to estimate the bias corrected bootstrap confidence intervals (BCBCI). The first hypothesis stated that exposure to a positive referendum proposal frame increases turnout intention as compared to exposure to a neutral frame, and that this effect is stronger for people with higher levels of opposition towards the referendum proposal. Referendum proposal frame exposure was used as independent variable, turnout intention as dependent variable, and attitude as moderator variable. Contrary to what we expected, the interaction effect between frame exposure and attitude was insignificant, t(70) = .19, p > .05, 95% BCBCI [-35.77, 43.19]. Also

6_{Before exposure to this item, an alternative turnout item based on Duff et al. (2007) was also measured. This item} was not chosen in our main analysis because it contained a smaller answer scale. Unless reported differently, the use of this item did not influence the significance of any outcome in this study. The full item can be found in the appendix.

(20)

20

when we controlled this effect for higher and lower values of attitude, no significant result was found (p > .05). As a consequence H1 was rejected.

Hypothesis 2 stated that the expected effect of the first hypothesis was also moderated by external political efficacy, so that a combination of high efficacy and a more negative attitude leads to a higher turnout intention. This three-way interaction effect was not significant, t(70) = -.28, p > .05, 95% BCBCI [-12.85, 9.62]. Also here, no higher or lower values of any of the moderators lead to significant outcomes (p > .05), so therefore also the second hypothesis was rejected.

The next set of hypotheses in this study states that the reversed mobilization dynamic we proposed is mediated by risk perception. Referendum frame exposure was used as our

independent variable, risk perception as our mediator and turnout intention as our dependent variable. Next to that, the attitude variable and the political efficacy variable were used as moderators. Important to note is that we expect efficacy only to affect the b-path of our

(21)

21

moderated mediation model (see Figure 1). Hypothesis 3a states that a positive referendum

proposal frame positively affects risk perception while it is moderated for attitude towards the referendum proposal. According to our model the effect on this a-path is positive and significant,

b = .48, p < .01, 95% BCBCI [.14, .83]. The exact interaction effect is presented in Figure 2. As

we expected, higher levels of opposition towards the trawling surveillance law increases risk perception when exposed to a positive frame. Therefore we can accept H3a.

Figure 2 The effect of referendum proposal framing on risk perception for different attitude levels.

H3b stated that risk perception positively influenced turnout intention while it is moderated for attitude towards the referendum proposal and external political efficacy. This effect turned out to be insignificant (b = .94, p > .05). However, when we look at specific values for both moderators after using model 3 of PROCESS (Hayes, 2013), we find some significant results that are presented in Table 1. We see significant outcomes for highly negative attitudes

(22)

22

towards the referendum proposal with at least average external political efficacy (p < .05). Based on these results we can accept H3b.7

Table 1 The effect of risk perception on turnout intention for different values of the moderator variables.

(Negative) attitude towards

referendum proposal (M) External efficacy (W) b t p

Moderate Low 2.07 .38 .70 Moderate Average .51 .14 .89 Moderate High -1.06 -.21 .83 Average Low 1.99 .60 .55 Average Average 3.75 1.56 .12 Average High 5.50 1.67 .10 High Low 1.92 .50 .62 High Average 6.99 2.22 .03 High High 12.1 2.78 .01

Note: The values of both moderators are the means (average) and one SD above (high) and below (moderate/low) the mean. Significant values are in bold.

Based on the significant outcomes for H3a and H3b, it is likely to find also significant outcomes for the overall mediation model for certain values of the moderators. This is indeed the case; for values one standard deviation above the mean for both moderators we find significant outcomes (b = 7.62, SE = 5.82, 95% BCBCI [.60, 27.71]). Therefore we can accept risk

perception as a mediator in the reversed mobilization dynamic.

Our final set of hypotheses state that next to risk perception, also anger will serve as a mediator between referendum proposal framing and turnout intention. Hypothesis 4a proposed that positive referendum proposal framing has a positive effect on anger for people with more negative attitudes towards this proposal. We found an insignificant effect for this hypothesis (b = .22, p > .05). However, if we use model 1 of PROCESS (Hayes, 2013) to test the moderation, we

7

For our alternative turnout intention item also found significant results for average negative attitudes combined with average external efficacy (b = .39, p = .01).

(23)

23

do find significant outcomes for the mean value for attitude (M = 2.50) (b = 1.19, p <.01) and one standard deviation above the mean value for attitude (3.94) (b = 1.51, p <.01). This indicates that more negative attitudes towards the referendum proposal positively influence anger, which means we can accept H4a.

H4b proposed that anger had a positive effect on turnout intention moderated by attitude and external efficacy. Model 71 showed an insignificant effect (b = 1.72, p > .05). Again when we look at specific moderator values via model 3 of PROCESS, we do find significant outcomes for this b-path. These findings are presented in Table 2. We see positive and significant results if either attitude towards referendum proposal is or external efficacy is high, combined with at least average values of the other moderator (p < .05). Based on this we accept H4b.8

Finally, we also want to test if there is a significant overall mediation effect for certain moderator values. The results are presented in Table 3. Only when high external efficacy was combined with at least mean negative attitudes the indirect effect is positive and significant (p < .05), which means that for these moderator values anger is a mediator in the reversed mobilization dynamic.9

Table 2 The effect of anger on turnout intention for different values of the moderator variables. (Negative) attitude towards

referendum proposal (M) External efficacy (W) b t p

Moderate Low -2.44 -.86 .39 Moderate Average -1.29 -.62 .53 Moderate High -.15 -.04 .97 Average Low -1.27 -.64 .52 Average Average 1.85 1.25 .21 Average High 4.97 2.09 .04 High Low -.10 -.04 .97 High Average 4.99 2.30 .02

8_{It is important to note that for our alternative turnout intention item we did not find significant results for H4b. A} high value for attitude combined with an average value for external efficacy was nearly significant (p = .054) 9_{For our alternative turnout intention item we found no significant overall mediation effects.}

(24)

24

High High 10.09 3.02 .003

Table 3 The indirect effect of positive referendum framing on turnout intention via anger for

To be more sure there we can speak of an actual effect of anger, and not a more broad “emotion effect”, we conducted the same analyses for hope, sadness, enthusiasm, contentment, fear and compassion. After using model 1 of PROCESS for the a-path, we found a significant negative effect of framing on contentment, moderated for attitude (b = -.45, p < .05). For hope we found a positive effect for moderate (b = 1.29, p < .01). and average (b = .92, p < .05) attitude values. Similar findings were found for sadness, enthusiasm, fear and compassion. None of the a-path effects on these emotions showed significant positive results for higher scores on our attitude variable

We conducted model 3 of PROCESS to investigate possible b-path effects. High external efficacy combined with high scores for negative attitude lead to a significant and positive effect for hope on turnout intention (b = 6.47, p < .05), but only for our initial turnout intention item.

BC 5000 BOOT (Negative) attitude towards

referendum proposal (W) External efficacy (V) IND SE LL95 UL95

Moderate Low -2.8 4.84 -19.66 2.02 Moderate Average -.84 3.20 -9.41 3.77 Moderate High 1.15 6.46 -10.70 16.58 Average Low -3.34 4.68 -14.98 4.35 Average Average 1.87 2.56 -1.66 10.02 Average High 7.08 5.02 .63 24.05 High Low -3.58 10.95 -29.90 17.60 High Average 6.18 5.79 -1.78 22.00 High High 15.95 9.65 3.00 49.07

(25)

25

Similar findings are found for enthusiasm (b = 8.41, p < .05). Based on model 71 of PROCESS, none of the emotions lead to a significant overall mediation effect for any values of the

moderators.

Conclusion & discussion

This experiment tested the reversed mobilization dynamic in a Dutch referendum context. We did not find significant direct effects of positive referendum proposal framing on turnout intention for people with more negative attitudes towards this proposal, neither did we find any effects when external political efficacy was taken into account. On the other hand, we did find significant indirect effects for both risk perception and anger, but only for high levels of both external efficacy and negativity towards the referendum proposal.

Our finding that risk perception mediates the effect of framing on turnout intention is consistent with the reversed mobilization dynamic as proposed by Schuck and de Vreese (2012). People with stronger negative attitudes experience a positive frame as a possible risk, an

undesired change of the status quo, which in turn mobilizes them in a referendum. Our significant b-path of the mediation shows that efficacy indeed affects behavior as a consequence of risk perception, in line with Rimal & Real (2003).

This research answered the call of Schuck & de Vreese (2012) to test emotions as possible mediators in the effect of framing on turnout intention. As expected, this study advances current research since our findings suggest that anger is a second underlying mechanism behind this dynamic. Not only does framing encourage multiple emotions, especially the feeling of anger in turn triggered people with negative attitudes and high efficacy to vote. The outcome that only the highest scores on this variable lead to increased turnout intention is in line with the findings of Schuck and de Vreese (2012). These groups are more motivated to resist frames that oppose their

(26)

26 beliefs (Meffert, 2006).

Our finding that anger, and no other of the measured emotions, served as a mediator is in line with research that showed a positive relationship between anger and political participation (Bartholomé, 2017; Best & Krueger, 2011; Valentino et al., 2011). Moreover, we were right by arguing that although mobilization to vote against a referendum proposal is a form of risk-aversive behavior, we should not couple this to a sense of fear. In line with past research (e.g. Lecheler et al., 2013), our findings suggest that fear leads to avoidance, while anger leads to more confrontational behavior.

Our study did not find significant direct effects for framing on turnout intention for people with negative attitudes and high efficacy. This means that, apart from the indirect paths that we tested in this study, there must be another indirect path that operates in the opposite direction (Hayes, 2009). In other words, there is a dynamic that decreases turnout intention for people with negative attitudes and high efficacy, which cancels our two mediators out. It is a challenge for future studies to look at other possible mediators that demobilize people with high external political efficacy.

A possible explanation could be that the positive frame in this case lead to a decrease of opinion strength, which demobilizes (Kiousis & McDevitt, 2008). Schuck and de Vreese (2012) already mentioned that reversed mobilization is more likely to occur in more salient electoral contexts. In such cases, opinions are stronger and individuals will be more likely to resist frame messages that argue against their beliefs, and will lead to further polarization. The trawling surveillance law is a relatively new subject, and public debate about it did not take place until recently. In such a referendum, which Leduc (2002) calls opinion formation, most voter do not have a well-formed opinion yet, and as a consequence volatility is high. It is possible that at least for a part of our sample that was the case. The significant outcomes for hope and enthusiasm on

(27)

27

turnout intention could also explained with this opinion formation concept. After exposure to a positive frame, negative people could in fact experience more hope for positive consequences of the new law, and would vote in favor of it.

We did saw a significant positive relationship of positive referendum proposal framing on almost all emotions when controlled for attitude. We think this has to with the fact that we used a neutral frame as our controlled condition. Neutral frames are very unlikely to appeal to any emotions, so the mean scores on emotions after exposure to a positive frame will easily be significantly higher. If we would have compared the positive frame to a negative frame, positive emotions like enthusiasm would less likely be significant on our a-path.

This study comes with a few limitations. First of all we must conclude that our final sample (N = 86) was relatively small. The combination of this sample with a relatively complex framework with multiple mediators and moderators, make the groups for all the different

conditional effects very small. This has negative consequences for the reliability of the findings. Secondly, our results for anger were inconsistent between our two self-reported turnout intention items. The main turnout intention item in our study was chosen based on the bigger answer scale (11-point vs. seven-point), combined with the fact that this question was asked later in the experiment, which reduces short-term effects. The inconsistency between the two items makes it uncertain if anger indeed is an underlying mechanism in the reversed mobilization dynamic. It is interesting for future studies to test this further, preferably with a multi-item

turnout measure.

Finally, we must note that the setting of this study was highly experimental. Although we distracted participants so they were less likely to find out the purpose, still participants were only exposed to one article about the referendum proposal, while during a campaign there will be many more messages that people can and will be exposed to. Exposure to competitive frames or

(28)

28

to the same frame with different time delays, would give us more information about the strength and persistence of framing effects (Schuck, 2015). Due to limited time and lack of resources, we chose a relatively simple experimental set-up.

These limitations notwithstanding this research contains important implications for framing research. First of all, it expands the reversed mobilization dynamic by adding anger as a second underlying mechanism next to risk perception. More broadly, this study helps us

understand what mobilizes people in a referendum. We found that people need to be triggered, by appealing to an emotion like anger, or make them perceive an undesired outcome as a risk.

Moreover, it shows that if attitudes towards a proposal are strong enough, messages with opposite arguments can lead to more resistance.

The reversed mobilization dynamic is very much underexposed in academic studies, so follow-up studies are more than welcome to see how strong these mechanisms actually are. A challenge for future researches is to compare what we assume to be an opinion formation

referendum, to referendums where attitudes are believed to be stronger (Leduc, 2002). That way we can test if framing can also have a direct reversed mobilization effect. Also, we recommend to test mediators that can have a demobilizing effect for people with negative attitudes, like for instance a decreased government trust.

This study strengthens the idea that emotions are useful mediators in framing effects on behavior, even if the occurrence of this emotion is not intended by the messenger. Especially in a referendum, where turnout is less obvious and people can have a more direct influence on

politics, this unintentional effect of a campaign message can have substantial consequences.

(29)

29

Aarøe, L. (2011). Investigating frame strength: The case of episodic and thematic frames.

Political Communication, 28(2), 207-226.

Advisory referendum. (n.d.). Retrieved from

https://english.kiesraad.nl/elections/advisory-referendum

Bartholomé, G. (2017). Faces of conflict: Interventionism and substantiveness in the conflict

framing process. Amsterdam: (n.p.).

Bizer, G. Y., & Petty, R. E. (2005). How we conceptualize our attitudes matters: The effects of valence framing on the resistance of political attitudes. Political Psychology, 26(4), 553-568.

Best, S. J., & Krueger, B. S. (2011). Government monitoring and political participation in the United States: The distinct roles of anger and anxiety. American Politics Research, 39(1), 85-117.

de Vreese, C., & Boomgaarden, H. (2003). Valenced news frames and public support for the EU. Communications, 28(4), 361-381.

Duff, B., Hanmer, M. J., Park, W. H., & White, I. K. (2007). Good excuses: Understanding who votes with an improved turnout question. Public Opinion Quarterly, 71(1), 67-90.

Entman, R. M. (1993). Framing: Toward clarification of a fractured paradigm. Journal of

Communication, 43(4), 51-58.

Fietsend appen mag misschien niet meer, maar hoe gevaarlijk is het? (2017, December 7). NOS. Retrieved from http://nos.nl

(30)

30

Gross, K. (2008). Framing persuasive appeals: Episodic and thematic framing, emotional response, and policy opinion. Political Psychology, 29(2), 169-192.

Gross, K., & D’Ambrosio, L. (2004). Framing emotional response. Political Psychology, 25(1), 1-29.

Hayes, A. F. (2009). Beyond Baron and Kenny: Statistical mediation analysis in the new millennium. Communication monographs, 76(4), 408-420.

Hayes, A. F. (2013). Model templates for PROCESS for SPSS and SAS. Retrieved from http://www.personal.psu.edu/jxb14/M554/specreg/templates.pdf

Hobolt, S. B. (2006). How parties affect vote choice in European integration referendums. Party

Politics, 12(5), 623-647.

Kahneman, D., and A. Tversky. 1979. Prospect Theory: An Analysis of Decision under Risk.

Econometrics, 47, 111-132.

Kernell, S. (1977). Presidential popularity and negative voting: An alternative explanation of the midterm congressional decline of the president's party. American Political Science

Review, 71(1), 44-66.

Kim, H. J., & Cameron, G. T. (2011). Emotions matter in crisis: The role of anger and sadness in the publics’ response to crisis news framing and corporate crisis

response. Communication Research, 38(6), 826-855.

Kiousis, S., & McDevitt, M. (2008). Agenda setting in civic development: Effects of curricula and issue importance on youth voter turnout. Communication Research, 35(4), 481-502.

(31)

31

Kirchgässner, G., and T. Schulz. 2005. Expected closeness or mobilisation: why do voters go to the polls? Empirical results for Switzerland, 1981—1999. CESifo working paper series No. 1387.

Lau, R. R. (1985). Two explanations for negativity effects in political behavior. American

Journal of Political Science, 29, 119-138.

Lecheler, S., Bos, L., & Vliegenthart, R. (2015). The mediating role of emotions: News framing effects on opinions about immigration. Journalism & Mass Communication

Quarterly, 92(4), 812-83.

Lecheler, S., & de Vreese, C.H. (2013). What a difference a day makes? The effects of repetitive and competitive news framing over time. Communication Research, 40(2), 147-175.

Lecheler, S., Schuck, A. R. T., & de Vreese, C. H. (2013). Dealing with feelings: positive and negative discrete emotions as mediators of news framing effects. Communications: The

European Journal of Communication Research, 38(2), 189-209.

Leduc, L. (2002). Opinion change and voting behaviour in referendums. European Journal of

Political Research, 41(6), 711-732.

Lee, H., & Kwak, N. (2014). The affect effect of political satire: Sarcastic humor, negative emotions, and political participation. Mass Communication and Society, 17(3), 307-328.

Lerner, J. S., & Keltner, D. (2001). Fear, anger, and risk. Journal of Personality and Social

(32)

32

Meffert, M. F., Chung, S., Joiner, A. J., Waks, L., & Garst, J. (2006). The effects of negativity and motivated information processing during a political campaign. Journal of

Morrell, M. E. (2003). Survey and experimental evidence for a reliable and valid measure of internal political efficacy. The Public Opinion Quarterly, 67(4), 589-602.

Mols, A., & Janssen, S. (2017). Not Interesting Enough to be Followed by the NSA: An analysis of Dutch privacy attitudes. Digital Journalism, 5(3), 277-298.

Nabi, R. L. (1999). A cognitive‐ functional model for the effects of discrete negative emotions on information processing, attitude change, and recall. Communication theory, 9(3), 292-320.

Nabi, R. L. (2003). Exploring the framing effects of emotion: Do discrete emotions differentially influence information accessibility, information seeking, and policy

reference?. Communication Research, 30(2), 224-247.

Nabi, R. L. (2010). The case for emphasizing discrete emotions in communication research. Communication Monographs, 77(2), 153-159.

Nelson, T. E., Clawson, R. A., & Oxley, Z. M. (1997). Media framing of a civil liberties conflict and its effect on tolerance. American Political Science Review, 91(3), 567-583.

Rimal, R., & Real, K. (2003). Perceived risk and efficacy beliefs as motivators of change: Use of the risk perception attitude (RPA) framework to understand health behaviours. Human

(33)

33

Schuck, A.R.T., & Feinholdt, A. (2015). News framing effects and emotions:

Research trends and developments, in R. Scott and S.Kosslyn (eds.) Emerging Trends in

the Social and Behavioral Sciences, Hoboken, NJ: John Wiley and Sons

Schuck, A. R.T., & de Vreese, C. H. (2009). Reversed mobilization in referendum campaigns: How positive news framing can mobilize the skeptics. The International Journal of

Press/Politics, 14(1), 40-66.

Schuck, A. R. T., & de Vreese, C. H. (2012). When good news is bad news: Explicating the moderated mediation dynamic behind the reversed mobilization effect. Journal of

Sigelman, L., & Kugler, M. (2003).Why is research on the effects of negative campaigning so inconclusive? Understanding citizens’ perceptions of negativity. The Journal of Politics,

65(1), 142–160.

Turner, M. M. 2007. Using emotion in risk communication: The anger activism model. Public

Relations Review, 33(2), 114–119.

Valentino, N. A., Brader, T., Groenendyk, E. W., Gregorowicz, K., & Hutchings, V. L. (2011). Election night’s alright for fighting: The role of emotions in political participation. The

Journal of Politics, 73(1), 156-170.

van den Dool, P. (2017, November 1). Referendum over sleepwet gaat definitief door. NRC

Handelsblad. Retrieved from http://www.nrc.nl

Wattenberg, M. P., & Brians, C. L. (1999). Negative campaign advertising: Demobilizer or mobilizer? American political science review, 93(4), 891-899.

(34)

34

Appendix

Stimulus Material: neutral frame versus End justifies the means frame (underlined), translated from Dutch.

Referendum about intelligence services will definitively take place

(‘It is important to trust the government’)

On Wednesday morning the electoral council announced that the referendum about the new law for intelligence services will definitively take place. Safety expert Roger Swart is in favor of the law, and emphasizes the urgency of it.

 Gerard van Houten November 1, 2017

Again the Netherlands will go to the ballot box for an advisory referendum, this time about the new Law on the intelligence and security services. On Wednesday morning the electoral council announced that the initiators of the referendum have collected sufficient signatures. According to a sample from the electoral council it turns out that over 384,000 valid signatures have been collected, while 300,000 signatures were required. The referendum will very likely take place on March 21, simultaneously with the municipal elections. Politicians are strongly opposed to each other. Jurist and security expert Roger Swart joins the debate:

“The protection of citizens must be always be the point of focus”.

The new Law on intelligence and security services expands the power of intelligence and security services. The law regulates for instance that the secret services can tap information like internet traffic on a large scale without focus. “A necessary measure”, according to Swart. He states that the recent terrorist attacks in France and Belgium prove that new authorizations are required. “The threat of terrorism is everyday’s business, and it has been shown that the current legislation in West European countries offers inadequate protection. With this law an important step will be taken.”

Critics are afraid that secret services can massively intercept data of citizens, and will store it for too long. Yet there is no other option, Swart thinks. “Citizens have to give up a small amount of their privacy for the safety of society. 100% privacy and 100%

(35)

35 safety do not go together.” Originally the law would go into effect on January 1, but this has been postponed until May 1 because appointing a new review committee takes more time than expected. Swart hopes that the referendum doesn’t throw a spanner in

the works. “It’s also important to trust the government a bit. They ask a small sacrifice of the citizens, and if they don’t give in, the consequences can be enormous. I hope that people realize that.”

Alternative turnout item

This item was measured by following the turnout question of Duff et al. (2007), which proved to minimize voter over-reporting. Respondents first had to read an introductory script: ‘Many people don’t vote during referendums. For instance because they had no possibility, because they didn’t receive a voting ballot, because they were sick, or simply because they had no time’. After that participants were asked the following question: ‘To what extent are you sure if you will vote during the upcoming referendum?’ Participants could answer on a seven-point answer scale (1 = I am sure I will not vote, 7 = I am sure I will vote) (M = 5.37, SD = 1.72). Higher scores indicate a higher turnout intention.

Full experimental survey (in Dutch)

Start of Block: Welkom bij mijn enquete!

Q1 Beste deelnemer,

Welkom, en bedankt dat u mee wilt doen aan mijn onderzoek! Tijdens dit onderzoek zal ik u een aantal vragen stellen over enkele maatschappelijke en politieke onderwerpen, waarbij uw eerlijke beantwoording van grote waarde is voor mij. Bij dit onderzoek zijn verkeerde antwoorden niet mogelijk.

Het invullen van deze vragenlijst duurt niet langer dan 10 minuten. Alvast bedankt voor uw tijd!

(36)

36

Dit onderzoek is uitgevoerd onder de verantwoorlijkheid van de ASCoR, Universiteit van Amsterdam, wat betekent dat de volgende punten gegarandeerd zijn: Uw anonimiteit blijft gewaarborgd, en uw persoonlijke informatie wordt onder geen beding aan derden verstrekt, tenzij u daar nadrukkelijk toestemming voor geeft U kunt deelname aan dit onderzoek weigeren of beperken zonder daarvan een reden te hoeven opgeven. Ook kunt u tot 24 uur na deelname de toestemming om uw antwoorden te gebruiken in dit onderzoek terugtrekken. U zult door uw deelname geen risico lopen om misleid te worden of om blootgesteld te worden aan aanstootgevende beelden of teksten. Indien u dat wenst kunt u tot uiterlijk 5 maanden na het meedoen aan het onderzoek vragen om een rapport dat de resultaten van dit onderzoek uitlegt. Indien u klachten of opmerkingen heeft over het onderzoek kunt u de Ethische Commissie van de ASCoR op het volgende adres bereiken: ASCoR Secretariat, Ethics Committee, University of Amsterdam, Postbus 15793, 1001 NG Amsterdam; +31 20‐ 525 3680; ascor‐ secr-fmg@uva.nl.

Q59 Verklaring akkoord deelname:

o

Ik verklaar dat ik de bovenstaande infomatie gelezen en begrepen heb, en ga akkoord met deelname aan dit onderzoek. (1)

Page Break

End of Block: Welkom bij mijn enquete!

Start of Block: Allereerst enkele algemene vragen

Q11 Allereerst enkele algemene vragen:

Q7 Wat is uw leeftijd?

(37)

37 Q8 Wat is uw geslacht?

o

Man (1)

o

Vrouw (2)

o

Anders (3)

Q10 Wat is uw hoogst afgeronde opleiding?

o

Geen afgeronde opleiding (1)

o

Basisonderwijs / lagere school (2)

o

LBO / VBO / VMBO (kader- en beroepsgerichte leerweg) (3)

o

MAVO / VMBO (theoretische en gemengde leerweg) (4)

o

MBO (5)

o

HAVO / VWO / WO-propedeuse (6)

o

HBO / WO-bachelor (7)

o

WO-doctoraal of master (8) Q9 Heeft u de Nederlandse nationaliteit?

o

Ja (1)

o

Nee (2)

Skip To: End of Survey If Heeft u de Nederlandse nationaliteit? = Nee

End of Block: Allereerst enkele algemene vragen

(38)

38

Q3 Nu volgt een korte serie vragen over enkele maatschappelijke onderwerpen:

Q17

De minister van Infrastructuur is met een nieuw wetsvoorstel gekomen, waarin smartphonegebruik op de fiets vanaf 2020 verboden wordt in Nederland.

Heeft u eerder van dit wetsvoorstel gehoord?

o

Ja (1)

o

Nee (2)

Q18 Wat is uw mening over dit wetsvoorstel?

1 (0) 2 (1) 3 (2) 4 (3) 5 (4) 6 (5) Helemaal mee oneens (1)

o

Helemaal mee eens

Q19 Praat u weleens met anderen over dit onderwerp?

o

Dagelijks (1)

o

Regelmatig (2)

o

Zelden (3)

o

Nooit (4) Page Break

(39)

39

Q56 Wat is uw gevoel bij een wet die smartphonegebruik op de fiets verbiedt? 1 (0) 2 (1) 3 (2) 4 (3) 5 (4) 6 (5) Zeer positief (1)

o

Zeer negatief Page Break

Q16 De nieuwe Wet op de inlichtingen- en veiligheidsdiensten (Wiv), die ook bekend staat als de

'sleepwet', maakt het gemakkelijker voor inlichtingen en veiligheidsdiensten om ongericht op grote schaal informatie als internetverkeer te kunnen aftapppen.

Q13 Heeft u eerder van deze wet gehoord?

o

Ja (1)

o

Nee (2)

Q4 Wat is uw mening over deze wet?

1 (0) 2 (1) 3 (2) 4 (3) 5 (4) 6 (5) Zeer tegen de wet (8)

o

Zeer vóór de wet

(40)

40 Q55 Praat u weleens met anderen over dit onderwerp?

o

Vaak (15)

o

Regelmatig (16)

o

Zelden (17)

o

Nooit (18) Page Break

(41)

41

Q14 Wat is uw gevoel bij de wet die de mogelijkheden van inlichtingendiensten vergroot? 1 (0) 2 (1) 3 (2) 4 (3) 5 (4) 6 (5) Zeer positief (4)

o

Zeer negatief

End of Block: Nu volgt een serie vragen over maatschappelijke onderwerpen Start of Block: Efficacy

Q22 De volgende stellingen gaan over uw mening omtrent de Nederlandse politiek. Lees de stellingen alstublieft goed, en probeer zo eerlijk mogelijk antwoord te geven.

Q23 Ik zie mezelf als ongeschikt om deel te nemen aan de politiek.

o

Helemaal mee eens (1)

o

Enigszins mee eens (2)

o

Noch eens noch oneens (3)

o

Enigszins mee oneens (4)

(42)

42

Q24 Ik heb het gevoel dat ik de belangrijke politieke kwesties in ons land niet goed begrijp.

o

Helemaal mee oneens (17)

Q25 De afgelopen jaren heeft de regering een substantiële hoeveelheid aandacht besteed aan de mening van het volk.

o

Q26 Ik denk niet dat het volksvertegenwoordigers veel uitmaakt wat mensen zoals ik vinden.

o

(43)

43

Q27 Ik heb het gevoel dat als ik volksvertegenwoordiger zou zijn, ik het net zo goed zou doen als de meeste andere mensen.

o

Q28 Mensen zoals ik hebben niks te zeggen over wat de overheid doet.

o

Q30 Ik denk dat ik beter geïnformeerd ben over de politiek dan de meeste andere mensen.

o

Engiszins mee eens (2)

o

(44)

44

Q31 Dankzij verkiezingen besteedt de regering meer aandacht aan de mening van het volk.

o

End of Block: Efficacy

Start of Block: smartphone artikel

Q20

Op de volgende pagina vindt u een kort artikel. Lees het artikel alstublieft zorgvuldig; naderhand zullen vragen worden gesteld over de inhoud. Een "→" knop volgt na een aantal seconden. Klik alstublieft alleen op deze knop als u het artikel helemaal gelezen heeft. U kunt niet terugkeren naar het artikel als u op de "→" knop heeft geklikt.

Klik op "→" om naar het artikel te gaan.

Page Break

Q35 Lees het artikel alstublieft zorgvuldig; naderhand zullen vragen worden gesteld over de inhoud. Een "→" knop volgt na een aantal seconden. Klik alstublieft alleen op deze knop als u het artikel helemaal gelezen heeft. U kunt niet terugkeren naar het artikel als u op de "→" knop heeft geklikt.

(45)

45 Q34

End of Block: smartphone artikel Start of Block: smartphone vragen

(46)

46

Q48 Als u terugdenkt aan het artikel dat u zojuist gelezen heeft, in hoeverre bent u het dan eens met de volgende stellingen:

Noch eens noch oneens (3) Enigszins mee eens (4) Helemaal mee eens (5) De minister is voor een appverbod op de fiets (1)