Keywords

1 Introduction

The field of election campaign communication is concerned with any form of communication by political elites, parties or professional interest groups with the aim of informing, persuading, interacting and mobilizing citizens and, ultimately, influencing the result of a particular election (Scammell 2016). As Esser and Strömbäck (2013) define them, “campaigns are prototypical events of mobilization in which national political communication traditions crystallize, and latent properties come to the fore as if held under a microscope” (Esser and Strömbäck 2013, p. 308). As the authors further argue, electoral campaigns have a clear start and ending point and occur at regular intervals, which make them easily traceable and allow researchers to plan their data analyses well in advance. These qualities make them political events particularly worth-researching, and also equivalent units of analysis to investigate political communication trends across different countries. Against this background, it is no surprise that electoral campaigns are one of the most populated research domains in the field of political communication to date (de Vreese 2017; Graber 2005).

With this chapter we aim to identify main trends and gaps in this popular area of study by relying on a thorough literature review of election campaign communication articles using content analysis published in peer-reviewed journals (n = 29) over the last twenty years.Footnote 1 Studies were selected when they featured a quantitative content analysis and analyzed communication by political actors (politicians, parties, political organizations) during election times. Our review is based on 48 studies that were published between January 2000 and February 2020.

Our literature review shows that the field of election campaign communication has indeed attracted attention from many different disciplines. As an example, studies considered in this chapter have been published in communication science (58%), political science (22%), marketing (9%), sociology (9%), and in recent years in informatics (2%). We also noted that researchers’ interest in election campaigns has intensified in recent years, with two-third of the studies being published in the last ten years. This also goes in line with previous research attesting to an increasing importance of campaigns to explain electoral outcomes, above and beyond more steady and traditional individual determinants of voting (party ID, class) (e.g., Scammell 2016) Electoral campaigns have also been drivers of important innovations over the last years, and their narrowcasting trends hold unprecedented potential to provide fine-grained insight on citizens’ characteristics and political preferences ahead of decision-making processes (Hillygus and Shields 2008; Strömbäck and Kiousis 2014).

Due to the interdisciplinary nature of the field, studies in electoral campaign communication have focused on a broad variety of trends and objects of investigation, on different levels of analysis. To name but a few, we came across studies investigating differences in the communication strategies applied by female and male politicians (e.g., Carlson 2007; Hrbková and Macková 2020; McGregor et al. 2017), cross-cultural differences between individualistic and collectivistic societies (e.g., Chang 2000; Lee and Benoit 2004; Lee et al. 2016; Tak et al. 2007), studies questioning whether new technologies contribute to equalize opportunities to access the public sphere among smaller and bigger parties (e.g., Klinger 2013; Schweitzer 2008) or whether these technologies contribute to a more dialogic relationship between politicians and citizens (e.g., Graham et al. 2013; Graham et al. 2016). Other studies investigate how parties appeal to voters by use of populist communication (e.g., Casero-Ripollés et al. 2017; Jagers and Walgrave 2007) or identify the most successful campaigning strategies for communication via social media (e.g., Bene 2017; Keller and Kleinen-von Königslöw 2018; Staender et al. 2019).

In the following, we outline most frequent research designs and analytical constructs employed by these and further recent studies to content-analyze political messages and identify styles, actors and functions of political communication in election cycles. We finally identify main gaps in studies using content analysis to investigate election campaign communication trends, and provide directions for future research in the field.

2 Frequent Research Designs and Methodological Strategies

Focusing, firstly, on the methodological aspects and most frequently applied research designs, we find that only few studies (n = 6; 13%) apply a multi-method design, for example combining content analysis with other methods such as surveys (Lipsitz 2018), experiments (Druckman et al. 2020), network analyses (Lukamto and Carson 2016), or expert interviews (Magin et al. 2017). Two studies also apply an additional qualitative content analysis (Carlson 2007; McGregor et al. 2017). Cross-sectional studies represent the majority of studies in our sample (n = 42; 88%) and longitudinal research is comparatively scarce (see, e.g., Borah 2016; Druckman et al. 2009; Johnston and Kaid 2002; Lee et al. 2016; Schweitzer 2008; Steffan and Venema 2019).

A majority of studies use manual content analysis (n = 45; 94%). Out of the 48 studies considered, only two studies apply semi-automatic content analysis (Lipsitz 2018; McGregor et al. 2017), three other studies make use of a fully automated content analysis (Lukamto and Carson 2016; Nulty et al. 2016; Vasko and Trilling 2019).

A large majority of the studies looks at communication by individual politicians (n = 34; 70%) and/or political parties (n = 18; 38%). Additionally, one study investigates independent political groups that supported different candidates in the presidential election by sponsoring TV ads (Johnston and Kaid 2002). Studies comparing different political actors (parties, politicians, and/or political interest groups) are rather unusual (Benoit and Airne 2009; Casero-Ripollés et al. 2017; Johnston and Kaid 2002).

In terms of international and comparative perspectives on election campaign communication, we find that a majority of studies are mono-country studies and six studies are two-country studies (Chang 2000; Graham et al. 2016; Lee et al. 2016; Lee and Benoit 2004; Magin et al. 2017; Tak et al. 2007). Only three studies are multi-country studies. Two of them include countries from the European Union (Nulty et al. 2016; Vesnic-Alujevic and van Bauwel 2014) and a single study takes into account a diverse set of countries from different World regions (Ceccobelli 2018). When we look at the geographic focus of the studies, Western democracies stand out. Most frequently, US elections have been the object of analysis, closely followed by studies on elections in European countries. Outside traditional Western democracies, elections from two Asian countries have been subject to investigation: Korea (Lee et al. 2016; Lee and Benoit 2004; Tak et al. 2007) and Taiwan (Chang 2000; Schafferer 2004). Beyond these geographic regions, research into the content of election campaign communication is still scarce, notable exceptions are studies from Australia (Lukamto and Carson 2016), Israel (Marmor-Lavie and Weimann 2008), and Pakistan (Ahmed and Skoric 2015). With regards to the geographic scope of the studies considered, Ceccobelli (2018) presents one of the most extensive studies. Besides elections in the US and Europe, he looks at election campaign communication by politicians in many South American countries, as well as from Australia and New Zealand.

Only looking at the last twenty years, the advent of online communication tools has sparked new research in the field of election campaign communication research with a majority of studies looking into newer online communication channels (n = 24; 50%). At the same time, offline channels remain important for research in this field (n = 21; 47%). Interestingly, only two studies looked at both offline and online election campaign communication with the aim to compare them (Benoit 2000; Vafeiadis et al. 2018). Most specifically, the most prominent content-analyzed communication channel is TV advertising (n = 20; 42%), followed by Facebook and Twitter, with each amounting roughly 20% (n = 9). Content analyses of candidate websites make for 17% of the studies in our sample (n = 8). Newer platforms for online political communication have not received much attention with one notable exception: four studies delve into campaign communication on YouTube (Borah et al. 2018; Erigha and Charles 2012; Shen 2012; Vesnic-Alujevic and van Bauwel 2014). As for previous categories (countries, actors), comparative studies which explicitly compare different communication channels are still an exception (e.g., Baranowski 2015; Lukamto and Carson 2016; McGregor et al. 2017; Schafferer 2004; Vafeiadis et al. 2018).

Finally, we found differences in the type of elections that have been studied with a majority of studies looking at parliamentary elections (n = 32; 66%), followed by presidential elections (n = 13; 28%), as well as gubernatorial (n = 6; 13%) and senatorial (n = 4; 8%) elections. The elections to the EU Parliament and the communication strategies applied therein have also generated a great deal of interest in the research field (e.g., Adam and Maier 2011; Nulty et al. 2016; Vesnic-Alujevic and van Bauwel 2014). While studies that investigate campaign communication in national elections make for the majority of studies, only a few studies have concentrated on regional elections (e.g., Baranowski 2015; Benoit 2000; Lukamto and Carson 2016). An even smaller part of the studies we selected conduct content analyses in multiple election campaigns (Benoit 2000; Benoit and Airne 2009; Lipsitz 2018) or different phases within the election cycle (Ceccobelli 2018; Vasko and Trilling 2019).

3 Main Analytical Constructs

In the following we describe main constructs employed in the studies using content analyses of election campaign communication reviewed in this chapter along four main categories, which we labelled styles, topics, features, and functions. A more comprehensive overview of studies on election campaign communication using these key constructs is presented in Table 1.

Table 1 Most frequently used constructs in content analysis of election campaign communication (own representation)
Full size table

3.1 Styles

Tone. The question of whether political actors communicate positive messages or rather employ a more negative tone against their political opponents in election times has been the subject of a considerable amount of research in election campaign communication (e.g., Borah et al. 2018; Gunsch et al. 2000; Houghton et al. 2017; Kaid and Dimitrova 2005; Vasko and Trilling 2019) Nevertheless, evidence is still mixed, as to whether positive or negative political advertisement is more effective in election contexts, for example (Druckman and Parkin 2005; Norris et al. 1999). This mixed evidence in one of the most populated strands of literature in the field of election campaign communication (political advertising) may be in part rooted in different theoretical and methodological approaches to the study of the tone employed by election campaigners. Whereas a majority of the studies in our sample deal with the general message tone used within election campaign communication (e.g., Houghton et al. 2017; Rijkhoff and Ridout 2019; Vasko and Trilling 2019), a broad variety of related aspects and concepts have been investigated in the past. Most prominently, studies look specifically at the use of negativity in political communication during election times (e.g., Druckman et al. 2009; Johnston and Kaid 2002; Lukamto and Carson 2016). Within the concept negativity, sub-constructs such as negative campaigning (Steffan and Venema 2019) and neighboring concepts such as attack (Benoit 2000; Benoit and Airne 2009; Erigha and Charles 2012), rebuttal (Borah 2016; Lee and Benoit 2004; Schafferer 2004), acclaims (Benoit 2000; Benoit and Airne 2009; Borah 2016), and comparisons (Borah 2016; Torres et al. 2012) have also received substantial scholarly attention (for a detailed overview of how these constructs are conceptualized, see database).

Personalization. An election campaign is deemed to be highly personalized when it revolves around candidates and politicians instead of parties, institutions, or particular policy issues,”as well as when’non-political personality traits’ of electoral contestants are emphasized” (Kriesi 2012, p. 826). As one of the traditional news values (Umbricht and Esser 2016), and with politicians increasingly adapting to the media logic (Strömbäck 2008), personalization becomes one of the most important features of election campaign communication. Together with the study of campaigners’ message tone, personalization is one of the most frequently investigated concepts in election advertising (e.g., Enli and Skogerbø 2013; Hrbková and Macková 2020; McGregor et al. 2017; Metz et al. 2019; Steffan and Venema 2019).

Emotionalization. Emotionalization refers to the strategic use of emotions to target specific voters (Richards 2004). Emotional appeals range from inciting positive emotions like happiness or pride to negative emotions like anger and fear (Kaid and Johnston 1991) other studies looked at how politicians elicit emotions by using e.g. moral language (Lipsitz 2018). Prior research has demonstrated that using emotional appeals has an immediate effect on political attitudes and vote choices (Glaser and Salovey 1998; Way and Masters 1996). Together with personalization and message tone, emotionalization is the third major style addressed in studies on election campaign communication (e.g., Gunsch et al. 2000; Keller and Kleinen-von Königslöw 2018; Vesnic-Alujevic and van Bauwel 2014).

3.2 Topics

Issue. Election campaigns serve three main purposes, namely, introducing the candidate or party (image), raising issues, and stating where the candidate or the party stand on these issues and how they want to address the issue after success in the election (policy) (Nadeau et al. 2008). Research into specific issues raised in election campaign communication has gathered considerable attention, among others economy, social welfare, and environment (e.g., Kaid and Dimitrova 2005; Lukamto and Carson 2016; Vesnic-Alujevic and van Bauwel 2014). Some studies make a specific differentiation between issues on domestic and national politics (Bene 2017; Bühlmann et al. 2016; Vasko and Trilling 2019). More focused, another set of studies addresses the specific positions political contesters hold on different political issues (Druckman et al. 2009, 2020) or whether they address issues their party “own” (e.g., Shen 2012; Staender et al. 2019).

Policy. Election campaigners do not solely communicate, position themselves on, or even “own” certain issues but they also tend to make policy propositions thereof. Accordingly, a substantial amount of studies on election campaign communication also touch upon concrete proposals of how candidates or parties intend to implement specific political ideas or public policies (Lau and Redlawsk 2006). Mostly evaluated as a dichotomous category, studies look at whether campaign communication take up a policy or image focus (e.g., Borah et al. 2018; McGregor et al. 2017; Torres et al. 2012). Another set of studies look specifically at how concrete policy plans were put forward in election advertisements (Johnston and Kaid 2002; Lee and Benoit 2004), whether they adopt a long-term rather than a short term orientation (Lee et al. 2016), or the degree of criticism that specific policies elicit among political opponents (Benoit 2000; Benoit and Airne 2009).

Image. Often considered a dichotomous category, an image focus is treated as the antipode for communication strategies that focus on policies. The image focus displays the candidate and his qualities running for office (Johnston and Kaid 2002). When looking at the topics dealt with in election campaign communication, a significant part of the studies focused on image building for the candidates (e.g., Borah et al. 2018; Erigha and Charles 2012; Tedesco and Dunn 2019). Specifically, when investigating the image of the candidate, studies look into descriptions of the candidate’s character (Benoit 2000; Benoit and Airne 2009; Lee and Benoit 2004), their leadership skills (Druckman et al. 2009, 2020), and references to their party affiliation (Shen 2012).

3.3 Features

Channel. Mostly as a side interest, many studies investigate how a particular campaign message is being delivered and presented. Here, studies differentiate between, for instance, different types of messages (e.g., tweets, replies, or retweet) (e.g., Ahmed and Skoric 2015; Graham et al. 2013), whether the communication includes (hyper-)links to other sites (Carlson 2007), visual, audio or video material (e.g., Chang 2000; Staender et al. 2019). Within this category, Another set of studies deal with the adoption of certain technological features to facilitate campaign communication. This strain of research concentrates mostly on candidate websites (e.g., Baranowski 2015; Jackson and Lilleker 2010; Koc-Michalska et al. 2014).

Reactions. With the advent of election campaign communication through social media, questions of how successful communication strategies have turned out have received increased attention in the field. Through their unique features, social media offers the opportunity to quantify successful strategies by evaluating likes/favors, shares/retweets and comments/replies a post/tweet has received (e.g., Bene 2017; Keller and Kleinen-von Königslöw 2018; Staender et al. 2019).

3.4 Functions

Information. Political actors communicate with their constituency and potential voters for many different purposes. One of such purposes is the provision of information (Nadeau et al. 2008). Usually through one-way communication, political actors inform the public about themselves and their stands on political issues, during the course of an election campaign (Magin et al. 2017). Numerous studies have developed different strategies to measure the function of information within election campaign communication, for example through counting the number of posts published throughout an election campaign or coding whether candidate messages provided news or factual information (e.g., Graham et al. 2013; Klinger 2013; Magin et al. 2017; Shen 2012).

Interaction. Politicians engage in interactions with potential voters in an attempt to gain their support. Unlike when providing information, the strategy of interaction mostly involves a dialogical communication between a political actor and (potential) voters, media representatives or other political actors (Graham et al. 2013). Also termed—among others—discussion, engagement, or (pseudo-) discourse, interaction is mostly examined in studies looking at communication in the context of Web 2.0 (e.g., Graham et al. 2016; Keller and Kleinen-von Königslöw 2018; Klinger 2013; Magin et al. 2017).

Mobilization. The last function of campaign communication we would like to highlight is political mobilization of citizens through e.g. canvassing, donating, spreading the campaign message within their social network (Schweitzer 2008) and ultimately cast their vote. Like the functions of information and interaction, mobilization can be achieved through the use of certain technological features. One of such instance is facilitating the sharing of campaign content or online forms to register as volunteers (Koc-Michalska et al. 2014; Schweitzer 2008), as well as through targeted communication strategies to attract followers to engage in specific (online or offline) specific actions (e.g., Bene 2017; Klinger 2013; Magin et al. 2017).

4 Desiderata and Directions for Future Research

A majority of studies examined in this chapterFootnote 2 suggest that manual content analyses of traditional forms of political communication (political discourses, political ads) in either TV or online platforms from a particular country are the norm in the empirical literature on election campaign communication. However, a number of scholars have pointed at an ever-increasing central role of new forms and specific techniques of online election campaign communication, such as video sharing, interactions through social networks, podcasting, blogging, political web sites or social bots distributing political messages (Dimitrova et al. 2014; Zuiderveen Borgesius et al. 2018), that are relatively uncharted territory in content analyses of campaign communication.

Against this background, the question arises as how online and offline communication modes complement each other, what forms, messages and styles of campaigning are associated to each of them, and which purposes and functions they fulfill. Much research lies ahead to outline and identify the main characteristics, formats and contents of newer forms of election campaign communication, and how they are combined with more traditional means to reach out specific electorates.

As others have highlighted (Blumler and McQuail 2016; Esser 2019), our literature review also evidences a scarcity of studies dealing with the interplay between new and more traditional modes of campaign communication from a comparative perspective. Overall, more cross-national research (across Western and non-Western latitudes) would be needed to account for national, cross-national, and transnational trends in campaign communication. The comparative angle should be extended as to explore patterns in political communication during election and non-election times, as well as long term trends. Future research should also devote its attention to bridging and harmonizing strands of literature on political actors’ campaign communication and trends in media reporting (see chapter on election campaign coverage in this volume) (Esser 2019).

On a methodological note, the Internet provides boundless opportunities to dive into the communication of a broad variety of actors, topics, and platforms across different media and political landscapes, beyond those kernelled in this chapter. New computational approaches to content analyses (see chapter on Automated content analysis in this volume), validated with manual techniques or more qualitative approaches can ease the task of discriminating meaningful analytical constructs and consistent patterns in campaign communication, and represent a promising research avenue for the years to come.

Relatedly, new forms of election campaign communication will also need to be re-visited with an eye on the receivers’ end. So-called political microtargeting, big-data and sophisticated analytical tools are allowing politicians to address constituencies’ specific concerns, identities and tastes with an unprecedented precision virtually everywhere in the Western world (Kosinski et al. 2015). Studies dealing with campaign communication effects need to better account for implications of new modes of political communication consumption, as opportunities to interact with citizens increase. As better tools to grasp the behavior and mood of political information users come to the fore, it is expected that political campaigns will be designed and conceived to fulfill the needs and adapt to new “media multitaskers”, inadvertent political information consumers or people engaging in second-screening by e.g. watching a TV debate while tweeting about it. Researchers using content-analytical tools to investigate communication by political leaders, parties or interest groups will need to consider such new information use patterns and increased interdependencies between the political supply and demand sides when analyzing and making sense of political messages across different platforms.