Negative Campaign Advertising. Which effect does New Media have?

Table of Content

1. Introduction
1.1 Overview
1.2 Online Advertising in U.S. Presidential Elections
1.3 Advertising in the 2016 Presidential Election
1.4 Regulation of Political Advertising on TV and the Internet
1.5 Research Question
1.6 Structure

2. Theoretical Background
2.1 Political Campaigning and Advertising
2.1.1 Functions of Campaign Messages
2.1.2 Negativity in Political Advertising
2.1.3 Effects of Negative Campaigning
2.1.4 Online Campaigning
2.1.5 Content and Distribution of Online Ads
2.1.6 Image vs. Issue Ads
2.1.7 Negativity in Issue and Image Ads
2.1.8 Emotional Appeals in Political Ads
2.2 Framing Theory
2.2.1 Frames and Types of Framing
2.2.2 Framing Effects
2.2.3 Framing and Emotions
2.2.4 Framing and Distinct Emotional Responses
2.2.5 Implications on Political Decision-Making
2.3 Hypotheses

3. Methodological Approach
3.1 Data Collection and Sampling
3.2 Analysis of the Transcripts
3.2.1 Text analysis
3.2.2 Sentiment analysis
3.3 Regression Analysis
3.3.1 Linear Regression
3.3.2 Logistic Regression
3.4 Operationalization
3.4.1 Preprocessing of Transcript Data
3.4.2 Variables of Interest
3.4.3 Control Variables
3.5 Methodological Challenges

4. Descriptive Results
4.1 Campaign’s Advertising Strategies in the 2016 Election
4.2 Tone of Advertising
4.2.1 The Ten Most Frequent Negative and Positive Words
4.2.2 Positive Campaign Ads
4.2.3 Negative Campaign Ads
4.3 Fear and Anger Appeals
4.3.1 The Ten Most Frequent Words Evoking Fear
4.3.2 Fear Evoking Ad
4.3.3 The Most Frequent Words Evoking Anger
4.3.4 Anger Evoking Ad

5. Empirical Assessment of Theoretical Hypotheses
5.1 Tone of TV and YouTube Ads
5.1.1 Tests using the BING Sentiment score
5.1.2 Tests using the SentimentR score
5.2. Tone of Issue and Image Ads
5.2.1 Tests using the BING Sentiment score
5.2.2 Tests using the SentimentR score
5.3 Issue and Image Ads on TV and YouTube
5.4 Emotional Appeals in Issue and Image Ads
5.4.1 Tests using Percentage of Anger words
5.4.2 Tests using Percentage of Fear words
5.5 Emotional Appeals on TV and YouTube
5.5.1 Tests using Anger
5.5.2 Tests using Fear

6. Discussion
6.1 Overall Tone of Advertising
6.2 Issue and Image Advertising
6.3 Fear and Anger Appeals

7. Conclusion and Research Perspectives

8. References

9. Appendix
9.1 Frequency Tables
9.2 t-test Tables
9.3 Regression Tables

List of Tables

Abbildung in dieser Leseprobe nicht enthalten

List of Figures

Abbildung in dieser Leseprobe nicht enthalten

1. Introduction

1.1 Overview

Dating back to 1952, when the Eisenhower Answers America ad ushered political advertising as a feature of presidential campaign communication, candidates have embraced political advertising in the United States (Dunn & Tedesco, 2017). Ever since, those ads are one of the main ways by which presidential candidates communicate with voters (McNair, 2011). Prior to candidate webpages, email and social media, televised political ads presented a rare form of communication, allowing for a candidate-controlled message by the campaign (Dunn & Tedesco, 2017). Some scholars even argue that these messages presented by the candidates leading up to the election can be seen as a cornerstone of functioning democracies. Campaigns are one of the few moments in large-scale democracies when competing candidates make mass appeals to wide segments of the population. While campaigning, politicians present themselves and their records to citizens. These messages undergird the legitimacy that comes with winning elections. Victorious candidate, for the most part, make governing decisions that resonate broadly with the messages and promises of their campaign and their success/failure is judged in accordance to them (Fridkin & Kenney, 2012).

Depending on the chosen strategy, these messages can both be presented positively, i.e. promoting a candidate or negatively, allowing them to attack the image and issue stances of opponents or respond to attacks made by the opponent (Dunn & Tedesco, 2017). In recent decades, most political ads have chosen the negative approach. Thereby, Lyndon Johnson’s ad entitled The Daisy Girl of the 1964 presidential election remains one of the most memorable. It showed a young girl playing he loves me, he loves me not with a daisy and when the last petal was plucked, a voice counted down to a nuclear explosion. The tagline because the stakes are too high for you to stay at home created an even higher sense of urgency (Suggett, 2016). The implication was that a nuclear war was imminent if you voted for Barry Goldwater, Johnson’s Republican opponent, yet Goldwater was never mentioned by name (Fowler et al, 2016). Despite criticism at the time, Johnson’s subsequent electoral win can be seen as proof for the efficacy of negative campaigning and the reach of television (Suggett, 2016).

Traditionally, broadcast media has largely been the sole channel to broadcasted campaign ads. More recently, advertising has taken place across a variety of platforms, from ads in cinema, radio, television and online (de Boer et al., 2012). While traditional communication networks rely on potentially expensive, frequently mediated forms of mass messaging, social media platforms such as Twitter and YouTube offer inexpensive, and substantively different tools for candidates (Bode et al., 2016).

Overall, there is little discussion about the fact that the Internet has transformed the way politicians communicate with citizens. However, the speedy profusion of the Internet complicates a determination of the exact nature of ramifications for political advertising (Druckman et al. 2010). In an era of a transforming media culture across nations, political advertising has become more widely accessible, monitored, and crucial to understand (McNair, 2011). Additionally, most ads broadcast on television or radio are more hierarchically structured and marked by stricter gatekeepers and regulations. Online ads, however, are being spread more rapidly and through less centralized media flows (de Boer et al., 2012). Moreover, even though candidates are now able to use a broader set of options when addressing their voters, evidence of the importance of political advertising remains (Dunn & Tedesco, 2017).

1.2 Online Advertising in U.S. Presidential Elections

The rise of new media has unquestionably transformed campaign communications. Campaigns have proceeded from dabbling with websites in the 1990s, to sending emails and launching professionally managed sites, over to sharing content on social media (Druckman et al., 2014). The 2008 campaign was an important turning point for candidates in the process of applying new communications tools to reach a growing and increasingly engaged audience (Smith, 2009). Since Ryan Lizza of the New York Times, referred to the 2006 Midterm elections of the US Congress as The YouTube Election, YouTube has become a central election channel in the USA.

Particularly Barack Obama’s 2008 campaign is often regarded as revolutionary with regards to the use of new media. The campaign was notably successful in giving supporters the feeling that they were part of the campaign (Kraski, 2012). Although he also used traditional media outlets and ran negative ads, his campaign was based on a positive message, Hope and Change. Internet blogs and message boards carried the message of Hope across the nation (Suggett, 2016). The Obama campaign also took advantage of the possibility to upload lengthier videos on YouTube (Fowler et al., 2016). For instance, his 37-minute YouTube video of his speech on racial discrimination in the pre-election campaign received over six million clicks (Salmond, 2012).

In his re-election campaign in 2012, the Obama campaign expanded their online strategy, due to the altering media landscape (Think with Google, 2013). For their first general election push, they did not turn to a 30-second ad but a 17-minute Hollywood-style online documentary, narrated by Tom Hanks. At the same time, viewers were able to post campaign content to their Facebook pages, volunteer and donate, all without having to leave Obama’s YouTube page. As Darrell West, of the Center for Technology Innovation at the Brookings Institution put it, It’s hard to be persuasive through a direct ads. But if you can get people to share videos, it adds a degree of credibility because a friend is endorsing it (Peters, 2012).

Over the course of the re-election campaign, Obama’s team leveraged over 471,000+ subscribers and 289,000,000+ video views on their YouTube channel (Think with Google, 2013). According to Pew Research (2012), the Obama campaign also produced more than twice as many YouTube videos compared to his competitor Mitt Romney. As Nate Lubin, Director of Digital Marketing for Obama for America stated, We knew that voters were increasingly hard to reach on TV, and that a large chunk of voters that we needed could not be reached at all through TV […]. YouTube helped provide us the scale, reach, and targeting to speak to our audience in a way few other platforms could (Think with Google, 2013).

1.3 Advertising in the 2016 Presidential Election

In terms of advertising, the 2016 presidential election stands out specifically through the fact that advertising spending for television did not exceed spending in 2012. The trend during the last few election cycles has consistently indicated spending to far outpace the prior cycle in terms of dollars spent by the major-party candidates (Dunn & Tedesco, 2017). Clinton’s final total was far less than Obama’s budget in the past two elections. Her spending was almost cut in half when compared to Obama’s 404 million U.S. Dollar budget in 2012. When comparing the two campaigns, the Clinton campaign vastly outspent the Trump campaign in terms of TV ad buys. Although the Trump campaign made up ground in the final weeks of the campaign, it did not nearly reach the level of the Clinton budget in terms of overall spending. By the 2nd November 2016, Clinton had already spent 211.4 million US Dollar on TV ads, while Trump had spent only 74 million US Dollar, about of third of Clinton’s budget (Parry-Giles, 2017).

But not only the spending was found to have decreased, also the volume of broadcasted advertising had lowered compared to 2012. An Analysis by the Wesleyan Media Project reports numbers of ads aired by candidates, parties and outside groups, shows that between the 16th of September, 2016, and the 13th of October, 2016, the number of ads aired has more than halved compared to the same period during the 2012 presidential election (Dunn & Tedesco, 2017).

While those reports show that advertising spending and advertising volume decreased for Clinton when compared to Obama, the numbers are even more significant when Trump is compared to Romney. While most easily recognized that Trump ran an unconventional campaign, it seemed puzzling to observers that Trump espoused his lack of need for political advertising (Dunn & Tedesco, 2017).

One reason for the drop in spending and volume in TV ads could be that the 2016 candidates focused more energy on online advertising and social media compared to TV advertising (Parry-Giles, 2017). In January of 2016, three political ads ranked among YouTube's ten most-watched ads for the first time in its history and thereby, delivering millions more views to campaigns than ads in the best timeslots on any TV channel. Also, YouTube is particularly popular among younger viewers, with more 18-to-49-year-olds watching its videos than tuning into any cable network in the U.S. But, according to internal Google surveys, more than half of baby boomers and seniors are watching online videos as well (Harwell, 2016). Most importantly, overall digital spending reached an unprecedented high. According to Borrell Associates, a market research firm, digital spending for 2016 was estimated at 1.6 billion US Dollar, an 576 percent increase since 2012. Despite this increase in ad spending on social media, TV remained the most dominant platform for political ads with a 70 percent share of ad revenue (Parry-Giles, 2017).

With regards to the spread of their YouTube content, Hillary Clinton was far ahead of Donald Trump. Her political ads have drawn 26.2 million views as compared to his 19.9 million. In addition, Clinton’s main channel Hillary Clinton reached about 135’000 subscribers, while Trump’s campaign channel Donald J. Trump for President had 104’000 subscribers (Schill & Hendricks, 2017). This also could have resulted from the fact that the Clinton campaign was more active on their primary YouTube channel, releasing over 70 videos starting July 2016. Besides containing YouTube exclusive content, the channel also included every televised ad and her social media accounts re-circulated the ads for an even broader reach (Parry-Giles, 2017). However, it should also be noted that, YouTube is the only social media outlet where Clinton prevailed against her Republican opponent (Schill & Hendricks, 2017). Trump’s effective use of social media has given him high media visibility. By August 2016, Trump had over 22.7 million likes and followers on Facebook, Twitter and Instagram while Hillary Clinton only reached 15 million combined (Parry-Giles, 2017).

1.4 Regulation of Political Advertising on TV and the Internet

Regulations related to political advertising in the United States have been codified in four main federal statutes: the Communications Act of 1934, the Federal Election Campaign Act of 1971, the creation of the FEC in 1975, and the Bipartisan Campaign Act of 2002. Two federal agencies are in charge with overseeing and enforcing regulations: (1) the FCC is tasked with regulating broadcast stations that run political ads and (2) the FEC regulates the campaigns that produce them. The FCC’s responsibilities include requiring all broadcast stations to announce who sponsored any ad that appears on their station. The commission also monitors incidents of libel and slander, maintains the federal licensing system required for all broadcast stations and can issue fines or revoke licenses if rules are broken (Dube, 2009).

In contrast, the FEC is tasked with monitoring campaigns and their production of ads, in hope of increasing accountability by directly linking ads to candidates (Dube, 2009). The regulation that is probably most familiar to voters is aimed at negative advertising. This rule requires a disclaimer notice to be shown in all ads (Just & Crigler, 2017). The FEC defines a disclaimer notice as a statement placed on a public communication that identifies the person or persons who paid for the communication and which person or persons authorized the communication (FEC). These requirements are also related to the height of the disclaimer and the duration of time it needs to appear on screen (Dube, 2009). Furthermore, televised communications must include an oral disclaimer spoken by the candidate with whom the candidate identifies himself or herself and states that he or she has approved the communication (FEC). This is accomplished by the inclusion of an I approved this message statement. Here, a candidate appears on screen during the ad, usually at the beginning or at the end, and states I am (name of candidate) and I approve this message (Dube, 2009). Such a disclaimer must show up on any electioneering communication and on any public communication by any person that expressly advocates the election or defeat of a clearly identified candidate or solicits funds in connection with a federal election (FEC).

For these FEC regulations electioneering communication is defined as a broadcast, cable or satellite communication that fulfils the following conditions: (1) must refer to a clearly identified federal candidate; (2) is publicly distributed within 30 days before a primary election or within 60 days before a general election; and (3) in the case of Congressional candidates only, is targeted to the relevant electorate (can be received by 50,000 or more persons in the district or state the candidate seeks to represent). As defined in FEC regulations, the term public communication includes broadcast, cable or satellite transmission, yet does not include Internet ads, except for communications placed for a fee on another person’s web site (FEC). This addition was made in 2005 following the U.S. District Court decision in Shays and Meehan vs. FEC, which ruled that the FEC’s previous definition of public communication did not include Internet communication. However, the current FEC definition of public communication does not include communications placed for free on advertising supported sites such as YouTube (Dube, 2009).

With regards to the message contained in ads, the United States can be considered to be at the libertarian end of the spectrum of free speech. Consequently, political speech is generously protected even with regards to discriminatory language and false information. Even though the Supreme Court does not protect obscenity, fighting in words or words that are a clear and present danger, hate speech and symbolic speech, e.g. flag burning, have long been protected even in advertising. For instance, the Supreme Court rejected the idea to press criminal charges if a candidate lies about his or her time in spent military (Just & Crigler, 2017).

Further decisions by the Supreme Court have related to the distinction of certain types of advertising. In its decision of 1976, Buckley vs. Valeo, the Court defines general rules for the distinction of two broad categories, which they referred to as express advocacy and issue advocacy. According to their ruling, issue ads are only supposed to discuss broad political issues rather than specific candidates. While the central characteristic of issue advocacy ads are ideally policy issues, express advocacy explicitly encourages citizens to vote for or against a specific candidate. This distinction is instrumental for judges’ assessment if ads are subject to federal campaign regulations. The consequences of these regulations affect organizations sponsoring express advocacy which are restricted in the size of individual donations, are limited to non-corporate and non-union contributions and required to publicly disclose their contributors. On the contrary, issue advocacy ads should aim to educate the broad public and are protected by the First Amendment, which guarantees freedom of speech. Consequently, issue advocacy ads are not regulated by federal law and the sponsors of issue ads are not required to publicly disclose the sources of their funding. In addition, there is no limit on the amount of money any individual, union or corporation can contribute to an issue ad campaign. As possible guidelines, the Court offers footnote issued that any advertising containing such clear expressions of support or opposition as vote for, elect, support, vote against, defeat, or reject should be classified as express advocacy. Despite fierce legal and political debate, lower courts have been reluctant to categorize ads that do not contain any of the above listed words as express advocacy even if showing clear disdain or support for a candidate (PBS).

Measures and rulings like these were also designed because of concern about negative and misleading content. It was assumed that requiring candidates to approve the messages they paid for would also reduce the occurrence of negativity and inaccuracies in political ads by forcing candidates to take responsibility for the messages they pay for, while still protecting the First Amendment (Dube, 2009). Past research has indicated that this effort has not been successful, as the current regulatory framework allows for a considerable amount of misleading and negative information in political messages (Ansolabehere et. al., 1994).

Given this situation, the opening up of a new advertising outlet through the Internet with less regulation as in the case of YouTube could allow for even more negative, inaccurate and unaccountable content (Dube, 2009). In the aftermath of the 2016 election, a bipartisan bill known as the Honest Ads Act was introduced, marking it the first major attempt to regulate online platforms that sell ads to apply the same rules as more traditional advertising on TV, radio and in print. The attempt was justified by the alleged influence of Russia in the 2016 presidential election by buying and placing misleading political ads on platforms such as Facebook, Twitter and Google.

As the bill’s announcement states (Hatmaker, 2017):

The content and purchaser(s) of those online advertisements are a mystery to the public because of outdated laws that have failed to keep up with evolving technology. The Honest Ads Act would prevent foreign actors from influencing our elections by ensuring that political ads sold online are covered by the same rules as ads sold on TV, radio, and satellite.

Disclosures for ad financing would apply to any entity that purchases more than 500 U.S. dollar in ads cumulatively across a platform, a fairly low threshold for disclosures (Hatmaker, 2017). To that end, Internet companies would be required to maintain a public file of all political ads they run, similar to what broadcasting companies are currently obligated to do. Meanwhile, some companies have responded by introducing more transparency checks of their own. Facebook, Google and Twitter each pledged they would make political ads easier to spot while providing more information about the audiences those ads target (Romm, 2017).

The urge for new regulations revived an old debate at the FEC as how detailed political ads have to disclose about their efforts to sway voters online. The agency began this debate in 2011, but never issued any disclosure rules for tech companies, in part due to lobbying efforts by companies like Facebook (Romm, 2017).

However, in November 2017, Google urged the FEC to consider more explicit rules around online political ads, perhaps even a ban on foreign entities from purchasing election ads focused on some issues, not just candidates. In its filing with the FEC, Google sought to emphasize its special position among social media companies. Its platform allows political ads in search, on websites of publishers that participate in Google’s ad networks or on platforms owned by Google, such as YouTube. But Google also told the agency it had to modernize its disclaimer rule so that political committees and other organizations have clear notice regarding the disclaimers they are required to include with their internet communications. Google further appeared to be calling on the FEC to wade into whether that ban includes issue-focused ads that may not express advocacy for or against a particular candidate. In many cases, the ads purchased by Russians during the 2016 presidential race on Facebook, for example, didn’t mention Donald Trump or Hillary Clinton explicitly, but instead focused on divisive topics such as Black Lives Matter and gay rights (Romm, 2017).

1.5 Research Question

Scholars regularly disagree on the exact implications of how the Internet has impacted political campaigning (Druckman et al., 2010) and whether online and offline campaigns convey the same messages to the same audience or not (Ceron & d’Adda, 2016). Supporters of the innovation hypothesis have long vindicated the point of view that the media-specific features of the Internet, such as interactivity, multimedia and information capacity, fundamentally affect the nature politics is presented to the public (Schweitzer, 2008). Along these lines, online campaigning follows different patterns when compared to traditional campaigning (Ceron & d’Adda, 2016). Particularly with the rise of social media, candidates can potentially develop personalized and targeted messages, different from the party’s official campaign (Vergeer et al., 2013). In addition, the digital forum is expected to have equalized or democratized the access to the political discourse (Ridout et al., 2010).

The proponents of the normalization hypothesis take a contesting standpoint. This standpoint argues that the Internet mirrors real societal features and therefore, also reinforces traditional campaign tactics in online campaigning (Schweitzer, 2008). For political campaigns, the Internet is, therefore, nothing more than an extended tool to distribute the same information used in offline campaigning (Vergeer et al., 2013, p. 482).

In this master thesis, these two contradictory assumptions regarding political ads on television and on YouTube are examined. In the past few election cycles, presidential campaigns have begun to integrate YouTube into existing messaging strategies, which traditionally used to rely heavily on television advertising. The goal is to determine whether campaign messages on YouTube should be considered as a complement to traditional means of communication or a surrogate for them. In particular, the differing channels are analyzed concerning their content and tone. Thus, it should be determined whether the tone of YouTube messages occupy a unique place in modern campaigns distinct from that of television ads. Understanding how politicians use YouTube as compared to older forms of communication is crucial in order to comprehend the full campaign repertoire used by candidates and the possible differences in the types of messages, i.e. more positive or negative, to which citizens are exposed (Bode et al., 2016).

Hence, the research question reads as follows:

How does the content of political ads on YouTube and on TV differ with regards to their negativity?

This question is studied by focusing on negative campaigning in the context of the U.S. presidential election of 2016.

There are at least six reasons to compare to the content of ads on TV and YouTube in the U.S. case: (1) broadcast advertising has essentially been the central focus of research on campaign messaging, reaching beyond any other element of campaign communication (Ridout et al., 2015). As a result, hypotheses and findings can be better justified by relying on the extensive literature that already exists, frequently with a specific focus on the United States. (2) TV advertising has long been considered one of the most important channels through which campaigns communicate with voters, signaled in part by the amount of money candidates spend on it (Bode et al., 2016). Thereby, 2016 marks a particularly interesting example, considering the fact that Hillary Clinton’s spending on TV advertising largely exceeded Donald Trump’s, yet this failed to result in an electoral victory. In addition, for the first time in recent history, advertising spending on TV has decreased compared to the previous cycle, implying a change in importance of different channels (Dunn & Tedesco, 2017). (3) Both TV and YouTube ads occur in discrete units, often multiple times per day. This allows for a more coherent comparison of the content during an election cycle, as compared to more static forms such as campaign websites (Bode et al., 2016). (4) Even though Facebook and Twitter also offer the option to include campaign videos on their sides, YouTube remains the frontrunner as the second most visited website globally and with almost 5 billion videos watched per day (Fortunelords). In addition, videos shared on other social media sites often simply link directly to a YouTube video. (5) TV and YouTube advertising are currently subject to different rules under the FEC. Ads uploaded exclusively on YouTube are exepted to differ because these ads are placed for free on advertiser supported sites which are not included in the regulations (Dube, 2009). Hence, the difference in content regulation could also impact the tone and the message spread in these ads. (6) Campaigns in the United States have traditionally been seen as trendsetters in political campaigning. Found trends and implications may thus, serve as an example for future campaign strategies in elections worldwide.

1.6 Structure

In order to assess the research question, the next chapter 2 will define a number of core concepts used throughout the thesis and lay out the theoretical arguments for the hypotheses. In chapter 3, I will describe the sampling of the advertising data, the underlying methodological approaches and the operationalization of the measures relevant to examine the hypotheses. Chapter 4 provides descriptive results of the tone and content in the advertising data. The tests of the hypotheses and the decision about their confirmation or rejection will follow in chapter 5. In chapter 6, I will discuss the obtained results in connection with the general political and media environment surrounding the 2016 campaign. Lastly, chapter 7 concludes the thesis and addresses future research perspectives. Additionally, the topics of specific subsections will be described individually and in further detail in the beginning of each chapter.

2. Theoretical Background

In this chapter, I will outline the theoretical concepts and previous research that form the underlying arguments for the hypotheses described in section 2.3. The first part will be devoted to the general characteristics and functions of political advertising, different types of messaging and their emotional appeals. In the second part, I will take a closer look at the theory of framing, the cognitive processes they can provoke and subsequently, how this affects the political behavior of voters. Lastly, I will base my main hypotheses on these arguments.

2.1 Political Campaigning and Advertising

2.1.1 Functions of Campaign Messages

Election campaigns give voters the chance to familiarize themselves with eligible candidates, their stands on issues and the party they represent. Next to general media exposure, the production of political ads serves as a central tool in political campaigning (Johnston & Kaid, 2002). An electoral victory can be achieved through both positive and negative campaign messages. While the former positively express the political achievements and intentions of one's own candidate or party, negative campaigning aims to portraying bad aspects of the political competition (Schmücking, 2015).

Thereby, a primary goal for lesser-known candidates and those campaigning outside of their previous constituency is simply to raise awareness of their name. Voters must have heard of a candidate before they can be expected to have any image of or attitude about that candidate. In this sense, the goal of political advertising is not unlike the one of advertising for a new product on the market (Harris & Sanborn, 2013). As in commercial advertising, a candidate or product is tried to be sold (McNair, 2011).

Additionally, researchers have focused on the persuasive advertising efforts of campaigns. Probably the most essential goal of campaigning is to persuade the electorate though political messages and lead them to adopt attitudes. McGuire (1985) developed a model to illustrate which political messages are able to persuade citizens to change attitudes, and thereby shape their behavior. For example, a candidate ad might be considered effective when it (1) leads individuals to increase their favorability toward the preferential candidate and (2) encourages them to vote for that candidate on Election Day. Another model by Ridout and Franz (2011) conceptualizes a distinction between direct and indirect forms of persuasion. Direct persuasion, which they classify as vote choice effects, occurs when citizens alter their vote choice after encountering persuasive messaging. Indirect persuasion, which they label mobilization effects, relates to the alteration of citizens’ likelihood of heading to the polls on election day after being exposed to ads.

Other functions of political campaigning such as the portrayal of a certain image, the familiarization with a candidate’s policy stances and mobilization will be discussed in more detail in later sections.

2.1.2 Negativity in Political Advertising

Numerous terms have been used synonymously to describe negative campaigns: Negative Campaigning, Assault Election, Attack Politics and Negative Advertising among others (Schmücking, 2015). Negative content in political campaigns can be defined as an act of speech or communication that contains one or more attacks that are personally, voluntarily and directly addressed toward political opponents (Ansolabehere et al. 1994). Through negative advertising, the competing candidates or parties should be portrayed as unreliable, unpredictable, a threat to the general public and consequently, a positive image of candidates be permanently destroyed (Falter & Römmele, 2002). In particularly hard-fought races, so- called dirt attacks aimed at the political opponent often go beyond what the general public perceives as good taste (Schmücking, 2015). Ads used in this way also have the goal of dissuading individuals from showing up to vote for the opposing candidate. Since it is difficult to convince strong supporters of an opposing candidate to rethink their choice, the target audience of such ads tends to be undecided voters (Belt, 2017).

Holtz-Bacha (2001) distinguishes between three different types of negative advertising, namely: (1) direct attacks, (2) direct comparison and (3) implied comparison. Direct attacks are aimed directly at the political opponent. The direct comparison makes one candidate appear in a more positive light, while trying to discredit the other. Lastly, in the implicit comparison, the opponent is not named by name and the interpretation of the ad is left to the spectator. These attacks can therefore only be characterized as negative through interpretation.

Despite the fact that negative ads seem to have played a crucial role in most elections in the United States since the invention of TV, negative ads became more and more frequent in the past few cycles (Kaid, 2004). Theyh have dominated the airwaves in recent elections comprising 65 percent of all congressional ads aired between the 01 September and election day in 2006 and 66 percent in 2008 during the same period. Negativity reached its peak in the 2010 congressional races with 75 percent of all general election airings. Although it has decreased slightly in the following cycles, it nevertheless remained steadily high. In the same timeframe, presidential campaigns were marked by even more negativity. Nearly three quarters of general election airings in the 2008 contest between Barack Obama and John McCain were negative and that figure grew to nearly 87 percent four years later in the race between Obama and Mitt Romney. While parties and interest groups can be seen as the main drivers of negative advertising, candidate committees also do not refrain from going negativity in ads. In fact, candidate-sponsored ads comprise roughly half of all negative ads on air (Motta & Fowler, 2016).

As the use of negativity has grown, so has the criticisms, e.g. Franz et al. (2008) observe that negative campaigning is often perceived as serving to corrupt and debase democratic discourse, to mislead and confuse citizens, to shrink and polarize the electorate and to constrain elected representatives in their efforts to promote constructive policies. Yet, the right dose has been found to be the most determining factor for the success or failure of negative campaigns. Otherwise a so-called adverse reaction or boomerang effect might occur. Recipients might perceive certain ads as dishonest and unfair and will then develop negative feelings towards the aggressor (Strohmeier, 2013). How these effects of negative advertising will influence citizens’ attitudes and behavior, will be addressed in the next section.

2.1.3 Effects of Negative Campaigning

While the literature is relatively consistent on the extent and the evolution of negative campaigning, a general consensus on its effects is missing (Nai, 2012). A great deal of attention has focused on whether campaign advertising mobilizes or demobilizes political participation. A set of controlled experiments run by Ansolabehere et al. (1994) tested the demobilization hypothesis and found negative advertising to be strongly lowering voter turnout, especially among non-partisans. Over the years, many studies reached contradictory conclusions, with results proofing both a mobilizing and a demobilizing effect (Valentino & Nardis, 2013). For example, Goldstein & Freedman (2002) found negativity to enhance engagement and participation, perhaps because negative information tends to be more memorable.

More recently, Krupnikov (2011) argues that both sets of findings may be entirely correct: The impact of negativity during campaigns may depend on when it appears. For undecided voters, negativity can boost participation by providing information that can help people differentiate between two candidates. However, after making up one’s mind, negativity has the tendency to be demobilizing by possibly making people question the potential benefits of their candidate choice. Thus, for decided voters the media environment may have an influence on the motivation to act on that choice, yet not on rethinking the choice itself.

Results on evaluating the impact of negative ads on other democratic values seem to be more consistent. Citizen’s political efficacy – the degree to which they feel that their actions have an impact on politics – declines as a result the use of negative ads (Belt, 2017). Furthermore, negative campaigning has been positively associated with higher individual cynicism toward political elites (Valentino et al., 2001) and lower trust in government in general (Brader, 2005). Although these effects may also partially be explained due more to coverage of negative ads in the media rather than the impact of the ads themselves (Geer, 2006).

2.1.4 Online Campaigning

Not only in the U.S., but in countries all over the world, there is a growing appreciation among politicians and parties of the importance of the Internet in election campaigns. As well as being a channel for parties and politicians to distribute their election programs, it offers a new way for them to interact with potential voters and to mobilize their followers. Through online tools, it becomes possible to bypass traditional media channels and address citizens directly in a more targeted and personalized manner (Marcinkowski & Metag, 2014). Additionally, using online tools is generally cheaper than buying ads on TV and candidates can quickly address or respond to certain issues or events happening during the campaign (Fowler et al., 2016).

Although political pundits espouse the game-changing nature of social media for politics, research on how campaigns are integrating these media into their broader communication strategies and how they compare with more traditional media outreach is rare. Nonetheless, candidates’ have increasingly employed social media strategies and research shows that online content has been influencing voters and, thus, potentially election outcomes for quite some time (Scheufele & Tewksbury, 2007).

In accordance with online strategies, campaigns tend to target undecided voters primarily through persuasive appeals in TV advertising, while their mobilization efforts to convince partisans to show up at the polls are more concentrated online (Fowler et al., 2016). This differing focus can be rationalized by the fact that partisans are more likely to share mobilizing messages on social media, whereas swing voters can be harder to reach. However, the lines of this traditional division are blurring as addressable TV becomes more of a reality and campaigns invest millions on individual-level data to also form persuasive efforts online (Motta & Fowler, 2016). Moreover, TV ads often target older voters, yet those ads are also uploaded to YouTube and other social networks that are predominantly used by younger audiences (Perry-Giles, 2017).

When focusing on voters who receive the message, Prior (2007) provides an example of how changing media technology can interact with mass-level variation in psychological dimensions such as interest and attention. He examines the impact of changing media institutions on knowledge of public affairs and electoral participation. When three major broadcast networks dominated US television, i.e. NBC, ABC and CBS, a heterogeneous audience was exposed to homogeneous content, mostly in form of the national nightly news broadcasts. This equal basis reduced the disparities in knowledge and turnout between citizens, regardless of their political knowledge and interest. With the proliferation of information sources on the Internet choice was boosted. This allowed politically motivated citizens to seek out even more news than they had done previously, while others could avoid news entirely. Media choice then widens the gap between the least and most interested citizens. Moreover, although access to the Internet seems to be a matter of course nowadays, many citizens still use it primarily for entertainment purposes (Kalmus et al., 2011). In contrast, other studies found that social media sites were more likely to expose politically uninterested citizens to candidates’ pages than traditional websites. In broader terms, studies have shown beneficial effects on citizens’ sense of efficacy and knowledge of candidate issue stances (Tedesco, 2007; Warnick et al., 2005).

To summarize, because the Internet has enlarged the possibility of political acts and exchange, providing a theoretical and empirical background for them often poses a challenge for scholars. For instance, the simple act of accessing political information differs greatly from the production and distribution of political messages on YouTube, Facebook and other outlets. Despite the message itself, the messenger further determines, which theory seems most suitable. To understand how news media makes a decision on which campaign event to cover, gate-keeping or agenda-setting theories can be applied. If citizens decide to watch the same campaign event on YouTube, the theoretical frame shifts (Bimber, 2012). With regards to YouTube in particular, different angles have been explored. Some papers considered YouTube as a top-down communication channel used by electoral candidates, while others examined participatory content sharing such as viral videos and user generated content related to elections (Eveland et al., 2012).

2.1.5 Content and Distribution of Online Ads

Studies on the content and distribution of online ads are limited thus far. With regards to the tone of the ads in comparison to the content on TV, previous research has shown contradictory results. According to a study of the US presidential and Senate elections between 2004 and 2010 by Clark-Pingley (2011), YouTube campaigns are generally more negative than TV campaigns. He explains this by the fact that the targeted voters on YouTube are relatively strong backers of the candidate and thus, there is a smaller risk of the feared boomerang effect to negative ads (Clark-Pingley, 2011).

However, a 2012 study by the Brookings Institution, examining the tone, content, and popularity of 3,118 YouTube videos in 12 countries, found that the content of YouTube videos is generally more positive than of TV ads (Salmond, 2012). Similar findings were presented in a study by the Wesleyan Media Project. In an analysis of ads during the 2012 U.S Presidential election, only 43 percent of online ads had a negative connotation compared to 56 percent of TV ads (Fowler et al., 2016). One possible explanation for these findings might involve that undecided voters and weak supporters, which could still be persuaded, are generally the target audience of negative advertising. This audience group is more likely to randomly watch a candidate’s ad on TV than to search explicitly for YouTube videos (Salmond, 2012). The goals of positive advertising, on the other hand, is to encourage existing supporters to get to the polls or donate money. In the same study, they further conclude of that nearly three out four TV ads focused on policy issues, while only half of online ads did (Fowler et al., 2016).

Other studies have laid their focus on how online videos are being used in an electoral context. The general conjecture has been that campaigns take advantage of the expanded possibilities YouTube provides in contrast to TV ads. Instead of creating videos that resembled ads seen on TV or just uploading TV ads, campaigns can take advantage of the possibility to upload lengthier videos on YouTube and create more innovative content (Fowler et al., 2016). However, studies examining this matter have found conflicting results. Klotz (2010) conducted a systematic investigation of the most popular videos having to do with the 2006 and 2008 U.S. Senate races. The most popular videos were produced by the candidates and political parties and not by actors outside of the usual political players. Moreover, Klotz found little innovation regarding the formats as some assumed would be provided by YouTube. For instance, three-fourths of the most popular videos were 30 seconds in length, the same length as traditional TV ads. Another study, however, finds a little more evidence that the Internet is altering the format of political advertising. Carlson and Strandberg’s (2008) study of the 2007 elections in Finland found a few examples of ads designed for TV migrating to YouTube. Moreover, 8 of the 10 most popular videos were produced by candidates from small political parties, providing them with the opportunity to reach a wider audience. However, it is questionable if these findings are generalizable to other countries due to differing electoral systems and Internet cultures (Ridout et al., 2015).

Yet, it should be pointed out, that when online advertising is genuinely new with regards to its format or content, it may create substantial buzz. Subsequently, traditional news media might see itself forced to pay attention and give unknown candidates the possibility to reach a wider audience. This media coverage, in turn, may lead to additional views, thus increasing the impact of the ad (Ridout et al., 2015). This phenomenon is also referred to as ad amplification or simply free advertising or earned media (Fowler et al., 2016).

When it comes to the impact of individual ads, it seems that characteristics of the ads themselves drive viewership and shares. For instance, negative ads received substantially greater viewership in the case of the 2010 U.S. Senate races (Ridout et al., 2015), as well as the 2012 Presidential races (Fowler et al., 2016). People also seemed to view and share ads that focus on either policy stances or personal characteristics to the most extent. Thereby, ads talking about personal characteristics were surprisingly popular compared to their actual occurrence online (Fowler et al., 2016).

2.1.6 Image vs. Issue Ads

Among the most essential functions of political advertising, two of the most important lie in the construction of a particular image of a candidate and the positioning and explanation of policy issues (Johnston & Kaid, 2002). Shyles (1986) defines image commercials as those that are concerned with the candidate's received or projected personality traits and character attributes, without advocating policy positions. Issue advertising, on the contrary, are concerned with current topics linked to the national interest.

A main goal of image advertising is to present the candidate as a likeable human being and a man of the people. This focus on modesty might seem ironic, considering that most presidential candidates in the U.S. are members of a distinct elite group with Ivy League credentials (Iyengar & McGrady, 2007). Based on a definition by Johnson-Cartee and Copeland (1991), the most common elements of image ads deal with credibility, competency, and honesty. Political advertising also seeks to reinforce or alter an image of a candidate and redefine an existing one. One effective way to communicate an image is through eliciting emotional responses in the viewer (Harris & Sanborn, 2013).

Issue advertising fall into two broad classes: on one hand, they focus on a candidate’s past experience and proven accomplishments as a public servant or in business and on the other hand, they serve as a tool to summarize the candidate’s preferences on policy issues. In order to fit into a TV ad, these appeals have to be framed simply, so that candidates can convincingly communicate their position (Just et al., 1990). Generally, issue ads are perceived to be more substantive (Iyengar & McGrady, 2007). Furthermore, political advertising also sets the agenda on issues by conveying to voters what issues they should be particularly concerned about (Schleuder et al., 1991).

Due to its importance, a large body of research on political advertising focuses on the characteristics of each content. Analyses of issue and image content date back to the work conducted by Duane Tucker in 1959, who compared radio and TV ads from the 1956 Oregon State race. His results suggest that a central aspect of a candidate’s winning strategy was his advantageous capability in establishing a positive image. Influential studies have since been conducted by Leonard Shyles, who analyzed political ads of presidential primaries in the 1980s. In Issue-related ads, he assessed a strong emphasis on defense and foreign policy during 1984 primaries (Shyles, 1984). With regards to image ads, he discovered that typical image-type content included competence and other special qualities, honesty, leadership and strength (Hacker, 2004). In general, the relative importance of image and issue advertising largely depends on the national geopolitical dynamics. In peaceful and economically stable times, image ads tend to become more frequent. When voters are concerned about the state of the country, both incumbents and challengers turn to issue ads (Iyengar & McGrady, 2007).

In recent years, criticism about the assumed increase of image advertising grew louder. Its central argument concerns the presumption that by focusing more and more on a candidate’s personality, the political discourse is trivialized. Nonetheless, those concerns seem to have little basis in empirical reality, since research has shown that most political advertising on every channel, concentrate more often on issues than on the image of a candidate (Kaid, 2004). For instance, an analysis by the Wesleyan Media Project followed trends in advertising content, including negativity and policy from 1998 till 2012. The majority of political ads that were aired in each election were concerned solely with policy issues, with the exception of 2002 where still roughly half of the ads were focused on issues. Moreover, there are comparatively few ad airings that are solely concerned with the candidates’ image (Ridout et al., 2014). Similar results were found during House and Senate races in 2014 where 60 percent of all ads focused on policy matters for the most part. At the same time about 85 percent contained at least some mention of a policy issue (Fowler et al., 2016).

Even though this research contradicts the often uttered concern that political ads are dominated by image over issue content, it should be emphasized that issue advertising can still lack substance. For instance, Payne et al. (1989) reach the conclusion that issues raised in ads rarely go beyond vague policy preferences and that instead, ads are dominated by emotionally provoking images and symbols. Moreover, a clear distinction between issue and image campaigns is not always easily achieved. As Rudd (1986) notes, issue ads are often used to bolster aspects of a candidate’s image and thus, it becomes difficult to be separate them (Kaid, 2004).

Besides that, the importance of mainstream media and more recently online media on the perception of election campaigns should not be underestimated. When TV news began covering ads, they gave free media time to the most defamatory ads, which made for a compelling viewership (Just & Crigler, 2017). Nowadays, for almost all types of media, the qualifications of presidential candidates are more amplified than issues (Brettschneider, 2002).

2.1.7 Negativity in Issue and Image Ads

The distinction between issue and image ads is further relevant in discussions about negative ads about the opponent. A frequently raised argument claims that by negatively talking about the adversary's personality or the image of the party, issue-oriented election campaigns are deliberately avoided. The aim of such attacks is the same as in negative advertising overall, to create a public image of the competition that is as detrimental as possible (Schmücking, 2015).

Most research studying this matter propose that negative ads tend to be more issue-oriented than positive ads. Kaid and Johnston (1991) reached this conclusion from a study of over 800 presidential ads aired between 1960 and 1988 and confirmed it in later analyses that included presidential ads in 1992, 1996, and 2000 (Johnston & Kaid, 2002).

Additionally, other studies have explicitly focused on the perception of negativity on either the image or the issues stances of the opposing candidate. With regards to negative ads, Johnson and Copeland (1991) identified 10 types of negative advertising that they categorized as either political or personal issue appeals. Political topics concerned the political record, issue stands, and voting record. Personal topics concerned personal life, current or past marriage, criminal activities, family members, religion, medical history and sex life. They use the term negative issue appeals because whether a negative ad uses a negative appeal directed at a traditional political issue or whether it uses a personal characteristic issue, the negative appeal ultimately becomes a campaign issue. The type of negative appeal used may also respond to questions about fairness and unfairness of political attacks and thus, determine the effectiveness of an ad. Out of the ten topics introduced by Johnson and Copeland, the topics that respondents felt were fair areas of attack were stands on issues (93 percent), political activities (83 percent), and voting record (84 percent). The only topic among the personal topics considered fair to attack was criminal activities (88 percent). In general, respondents viewed negative political issue appeals in a positive light, while they viewed negative personal characteristic issue appeals unfavorably.

2.1.8 Emotional Appeals in Political Ads

An important component of a candidate’s video style concerns the types of appeals or arguments used in the ads (Johnston & Kaid, 2002). A large body of research literature emphasizes the impact of emotional messages on political attitudes and voting behavior (Marmor-Lavie & Weimann, 2005).

One possible way to distinguish emotional appeals is by contrasting them to rational appeals. Unlike emotional appeals, rational appeals attempt to allow a rational comparison between the campaign’s arguments by using messages containing facts and information (Rosselli et al., 1995). Regarding the processing of such information, research separates an experiential from a rational mode. The rational mode, which refers to rationality and logic, is analytic, while the experiential mode is controlled by emotions and is holistic. The dual system theory by Epstein’s (1994) describes how these two systems incline to lead to different, often unrelated responses that may be inconsistent. Consistent with this theory, studies of consumers’ choices demonstrate instances of consumers using either type of judgment. On one hand, they use analytical judgment by assessing and combining information into an overall evaluation. On the other hand, they can use an affective judgment by considering their emotions towards the target (Saqib, 2014).

Campaign professionals have long understood the importance of speaking to potential voters at an emotional level and not exclusively at a rational one. Emotional appeals are central to the work of political consultants and are especially important in televised political advertising. As one communications consultant states, Emotion is at the center of American politics. You have to have people excited if you want to win (Warren, 2016). Advertising tries to manipulate emotions, especially enthusiasm for the sponsor and fear, anger and disgust with the opponent (Brader, 2006).

While older research in political science used to be limited to the distinction between positive and negative emotions, recent work specifies various types of negative emotions (Ridout & Searles, 2011). For example, Brader (2006) provides findings suggesting that the use of specific emotional appeals in political advertising is common overall. He reports on coding of the emotional appeals contained in over 1,400 political ads aired in 1999 and 2000. Overall, enthusiasm is the most common emotional appeal, appearing in 73 percent of ads. Appeals to pride were found in 54 percent of ads, anger appeals were present in 46 percent of ads and fear appeals in 41 percent of the ads. Less than one percent of the ads contained no appeal to emotion. Ridout and Searles (2011) similarly coded emotional appeals made in 628 political ads of the U.S. Senate races in 2004. Their findings were fairly consistent with Brader’s earlier findings. Pride was the most common emotional appeal contained in 85 percent of ads. 84 percent of ads contained an appeal to enthusiasm, 48 percent contained an appeal to anger, and 24 percent employed a fear appeal.

When it comes to emotional appeals, image and issue ads differ in which appeals are prevalent. Johnston and Kaid (2002) found in their analysis of 1’213 ads from 13 U.S. presidential elections that fear appeals have been significantly more used in issue ads in order to evoke fear among their constituency about a certain issue rather than about the other candidate’s personality. On the other hand, the emotional appeals in image ads are dominated by positive emotions, such as the credibility and trustworthiness of the individual candidate (Johnston & Kaid, 2002).

In the subsequent analysis, special attention is given to negative emotions, particularly fear and anger. Both emotions can be aroused in the context of many social and political issues and have been discussed extensively in the literature (Nabi, 2003). Put generally, the primary personal fears experienced by constituents concern one’s standard of living, economic well- being, and physical health. Collective fears relate to concerns for the nation which include peace and war, economic stability and instability, national unity and disunity, and prosperity and lack of prosperity. As with negative advertising in general, fear-inducing tactics suggest if the opponent is elected, he or she would do things that the voter would judge to be disastrous for the collective or for the voter’s immediate environment (Johnson-Cartee & Copeland, 1991). The emotion of anger is elicited, often latently, as a result of economic and social inequalities. Many people in all societies believe that compensation differences do not accord with their notion of justice, such as the remuneration according to effort, merit or community value. Political parties use anger as campaigning tool in need to maintain their base against poaching from other parties. This requires creating an emotional connection with supporters and thus, the establishment of an us vs. them narrative. This can be maintained best when the other is cast negatively as an object of aversion (Ost, 2004).

Finally, it is important to note, however, that just because a political message contains an appeal to a particular emotion does not mean viewers will also experience said emotion (Schnur, 2007). Furthermore, emotional appeals within political messages are conveyed not only by the spoken words, but also by the visuals and music (Brader, 2006). However, these means are not considered in this thesis.

2.2 Framing Theory

2.2.1 Frames and Types of Framing

When analyzing the content of political messages, a large emphasis has been put on how those messages are framed. Gamson and Modiglini (1989, p. 143) characterize a frame as a central organizing idea or story line that provides meaning to an unfolding strip of events weaving a connection among them. The frame suggests what the controversy is about, the essence of the issue.

The theory on framing argues that the way in which information is presented or the perspective taken in a message, influences the responses individuals will have to the respective issue. According to Entman (1993, p. 52), 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. Thus, it could be posited that through framing attitudes the way a message prioritizes selected pieces of information over others is impacted. Ergo, frames may have the power to influence individuals’ perception of problems and their chosen solutions with regards to said issue (Nabi, 2003).

Frames can further be categorized into different types. Framing studies typically employ either equivalency or emphasis frames. Equivalency frames represent logically equivalent alternatives portrayed in different ways, whereas emphasis frames simplify reality by only focusing on a subset of relevant aspects of a situation or issue (Druckman, 2001).

Besides that, Issue-specific frames pertain to a specific topic, while generic frames are applicable to a wide range of topics (Nelson et al., 2012). Typically, issue frames are further characterized by a specific valence. This domain touches on one of the most fundamental characteristics of political discourse, namely that elites attempt to affect support for or rejection of an issue by emphasizing the positive or negative aspects of it (de Vreese & Schefeler, 2012). Valence frames are often used in studies aimed at competitive framing effects, that is, in studies that test how exposure to frames that hold opposing views of an issue influence opinion formation (Lecheler et al., 2013). Generally, valence is operationalized as two or more of the following four categories: positive, negative, neutral, and mixed or poignant (both positive and negative) (Lang et al., 1995).

The observation that negative and positive information are processed differently has been supported with regards to a broader psychological phenomena. Previous research has repeatedly shown that negative information has more impact on attitudes and evaluations than positive appeals (Lau, 1985). People pay more attention to negative information and rely more heavily on it in subsequent evaluations (Fiske, 1980). Most prominently, valence is central to the original study of framing in Kahneman and Tversky’s (1981) Asian disease problem. The study examines how individuals’ preferences shift depending on whether equivalent outcomes are described in terms of the number of lives saved out of 600 (e.g. 200 are saved) as opposed to the number of lives lost (e.g. 400 are lost). People generally preferred the absolute certainty inherent in a positive framing effect, which guarantees gains. Yet, when options were framed as a likely gain, risk-averse choices become predominant. When options are framed as likely losses and decision-makers adopt a more risk-seeking behavior, leading to negative framing effect.

2.2.2 Framing Effects

Research on framing has both focused on frames as a dependent or independent variable. When understood as a dependent variable, frames are contained within a message and thus, the result of production processes in for example the news media or political ads. When conceived as independent variables, the reception and impact of the frame is analyzed or in other words its effect (Igartua et al., 2011). One of the main goals of studies in this field is to describe the psychological processes that underlie framing effects. Thereby, the most common approach has been to specify multiple ways of cognitively processing frame information (Nelson et al., 2012).

For example, Chong and Druckman (2007) argue that a consideration must first be available to an individual’s memory. Based on these models of memory-based judgment formation, individuals rely on information that can be accessed in long-term memory when they form a judgment (Hastie & Park, 1986).

Secondly, the consideration must be accessible, or more specifically it must first be known and ready for use (Nelson et al., 2012). In this view, frames influence which cognitions the recipient can access. By increasing the salience of particular aspects of an issue, frames activate corresponding beliefs and increase their accessibility in memory (Iyengar, 1991).

Third, depending on the context, different considerations may be consciously evaluated, as a person decides about the applicability of his or her accessible interpretations. (Nelson et al., 2012). Frames generally combine various aspects of issues and thus, applicability effects refer to the process through which the connection of multiple aspects in a frame is adopted by the recipient (Price & Tewksbury, 1997). Because frames influence which relations between objects the recipient regards as plausible, they affect the beliefs, which influence recipients in subsequent judgment making (Kühne, 2012).

Another model combines these three processes with a fourth by adding new beliefs to an individual’s belief content (Slothuus, 2008). Frames can present new information and in doing so, change prior held beliefs (Lecheler & de Vreese, 2012). Therefore, it can be noted that in contrast to the importance hypothesis, which assumes that frames influence how existing beliefs are weighted in judgment processes, this perspective states that frames influence opinions by conveying new beliefs (Kühne, 2012).

2.2.3 Framing and Emotions

As discussed in the previous section, the great majority of studies on framing effects have concentrated on determining the cognitive effects of frames. However, this type of effect says little about whether frames can influence how and what citizens feel about politics. As Gross and Brewer note (2007), emotions should not be neglected, given that emotions play a fundamental role in political processes, for example, by affecting the appraisal of candidates in an electoral confrontation. It can therefore be expected that frames may influence not only opinion but also the emotional responses that people report (Gross & D’Ambrosio, 2004, p. 19).

In general, emotions can be defined as internal, mental state representing evaluative, valenced reactions to events, agents, or objects of differing intensity (Clore et al., 1994). They are generally intense and of short duration (Fiske & Taylor, 1991). Emotional processes can be distinguished from cognitive processes. According to LeDoux (1999, p. 69), emotion and cognition are best thought of as separate but interacting mental functions mediated by separate but interacting brain systems. On one hand, emotions can be seen as the product of cognitive evaluations and on the other hand, have substantial effects on cognitive processing and judgments (Kühne, 2012).

[...]