Comparative Digital Political Communication: Comparisons Across Countries,Platforms,and Time

Abstract

Comparative communication research needs to catch up to other disciplines. In this special issue and the associated International Communication Association preconference, we focus on comparative work related to digital political communication. This introduction argues that comparative digital political communication needs to consider comparisons across various dimensions, including countries, platforms, and time, whereas existing comparative communication research focuses on country or territorial comparison. We highlight the six submissions’ approaches to comparative work. Each submission provides at least one of these three dimensions of contrast. We conclude with a discussion of enduring gaps in this field of research, such as the lack of studies using time as a dimension of comparison. Time is crucial for understanding ever-changing digital media platforms. We also conclude by discussing some ongoing challenges in political communication research.

Introduction

Political communication examines “the flow of information and the exchange of messages among political actors, citizens, and the media” (Esser & Pfetsch, 2020, p. 337). For over a decade, in several countries and contexts, digital communication technologies such as social media have grown increasingly important in political processes like elections, public consultations, and political advocacy work. In addition, these technologies are important for facilitating and documenting everyday political interactions among citizens across the globe and between political actors/groups and their supporters. Indeed, our current “fourth age” (Bennett & Pfetsch, 2018; Magin et al., 2016) or “fourth era” (Klinger & Koc-Michalska, 2022; Roemmele & Gibson, 2020) of political communication is often defined by the characteristics of various digital tools and approaches. Research on these themes has flourished, giving rise to different theoretical viewpoints, methodological perspectives, and empirical starting points.

While the diversity of digital political communication research is largely beneficial, recent work has pointed out a lack of comparative approaches in dealing with issues like online political campaigning and the online political participation of citizens (Boulianne, 2020; Esser & Pfetsch, 2020; Jacobs et al., 2020; Kreiss et al., 2017), mirroring tendencies discernible in the broader field of media and communication studies (Liu et al., 2020; Matassi & Boczkowski, 2021, 2023). Esser and Vliegenthart (2017, p. 1) claim that “comparative communication research is lagging” in quantity and quality compared to sociology and political science. They define comparative research “as the contrast among different macro-level units, such as world regions, countries, sub-national regions, social milieus, language areas and cultural thickenings, at one point or more points in time” (p. 2).

In response to a perceived lack of comparative work in communication and media studies, we organized a preconference attached to the International Communication Association annual meeting in May 2023 in Toronto, Canada. The preconference was affiliated with the Political Communication section of ICA. As part of our call for papers, we received almost 100 abstracts, far more than we expected. The high number of submissions reflects the great interest in comparative digital political communication. From this list of submissions, we invited more than 20 papers to be presented at the preconference. From the presented papers, we invited submissions for this special issue; most of the invited submissions made it through the peer review process.

Theme: Comparisons Across Countries, Platforms, and Time

Much of the comparative communication work focuses on country comparisons. Zhao and Liu (2020) searched communication journals to identify comparative studies about Internet use/digital media. They found 45 articles and within these studies, more than half of studies conducted comparative work that was “cross-territorial.” However, Esser and Vliegenthart (2017) identified other dimensions for comparison, including language groups and multiple points in time. Our attention to these other units for comparison reflects our own research related to political communication, which includes a comparison of different countries as well as media and political systems (Boulianne, 2019, 2020, 2023; Boulianne & Humprecht, 2023; Larsson, 2020, 2021; Larsson et al., 2024; Larsson & Kalsnes, 2014).

Within the preconference and special issue, we hoped to revive time as a basis of comparison. Esser and Vliegenthart (2017, p. 2) wrote, “spatial (cross-territorial) comparisons should be supplemented wherever possible by a longitudinal (cross-temporal) dimension to account for the fact that systems and cultures are not frozen in time; rather, they are constantly changing under the influence of transformation processes, such as Americanization, Europeanization, globalization, modernization, or commercialization.” While they see time as important to understanding systems and cultures, we argue that time can be a basis for comparative communication. In particular, comparisons across time are critical to communication scholarship focused on digital media. These technologies are changing rapidly and as such, time has become a critical variable for analysis, perhaps more important than cross-national differences (Boulianne, 2020). For example, in their submission, Russmann et al. (2024) consider Facebook user engagement in 2014 (which was restricted to the thumbs-up/like button) versus 2019 (which includes a range of responses). Matassi and Boczkowski (2023) also encourage historical comparisons in cross-platform research (see Chapter 5).

In our call for papers, we included “time” in the title to encourage submissions that considered political communication over time, but the large majority of papers were about comparisons related to countries and platforms. Indeed, for the preconference program, we could not populate a session on the topic of time. Instead, the time dimension was considered alongside cross-national or platform differences. Since we had both published articles considering time as an important dimension of comparison (Boulianne, 2020; Larsson, 2023 and more recently, Boulianne & Larsson, 2024), we were surprised by the lack of submissions on this dimension of comparison. However, perhaps we should not have been surprised since Zhao and Liu (2020) had documented that only 9% of comparative studies included a temporal comparison. Furthermore, Downey (2020) noted that the field of communication had yet to address how to “qualitatively and systematically explain change over time” (p. 5794), but argued that comparison across time is “essential to understand and explain the process of change (or indeed stasis)” (p. 5793).

Expanding on Esser and Vliegenthart’s (2017) definition, we add a new element to comparative communication research: platform comparisons reflecting our collaborative work comparing social media platforms (Boulianne & Larsson, 2023). Our approach fits with other work within communication. In 2018, Political Communication published a special issue on cross-media comparisons (Bode & Vraga, 2018). Beyond political communication, Matassi and Boczkowski (2023) published an aptly titled book, To Know Is to Compare: Studying Social Media across Nations, Media, and Platforms.

Platforms span countries, language groups, and cultures, raising questions about the ongoing relevance of these other points of comparison. Instead, platforms can be thought of as user communities with different cultural norms for social interaction. These patterns of social interaction are governed by platform affordances (Evans et al., 2017) or digital architectures (Bossetta, 2018), which simultaneously constrain and enable the actions possible to users. Indeed, Matassi and Boczkowski (2023) describe the importance of platform comparison alongside cross-national work, as they write, “Platforms are deployed and used all over the world, with information flows connecting accounts located in multiple countries via reactions, comments, and shares” (p. 3).

Indeed, platform comparisons can benefit from the legacy of cross-national comparative research, which tries to grapple with whether differences are explained by culture or institutions. For example, in comparative political behavior (Boulianne & Lee, 2023), studies point out how the French are more likely to protest, but the question is whether this finding reflects the structure of the French government (macro), the strength of unions (meso), or micro-level explanations such as personality, discontent, or the perceived efficacy of this tactic. Alternatively, the explanation may relate to culturally specific explanations, such as norms, or historical tendencies dating back to the French revolution.

We need to ask those questions about platforms: Is Twitter discourse more uncivil because of affordances (potential for anonymity), because of the groups and organizations on the platform, or because of the user-generated cultural norms on the platform (see Farjam & Dutceac Segesten, 2024; Schmidt et al., 2024). If the explanation is the platform’s affordances, will these patterns replicate across different countries? Matassi and Boczkowski (2023) address this point. They explain that

a popular interpretive frame in cross-national and regional comparative studies of social media is making sense of the findings in terms of either divergence or convergence of phenomena under examination. On the one hand, there is the notion that under certain circumstances the culture associated with the national territory effectively shapes the use of platforms and ends up producing significant variations. On the other hand, there is the idea that despite the differences among countries, there are major points in common in the use of social media (Matassi & Boczkowski, 2023, p. 35).

As such, cross-national platform comparison studies need to reconcile these two possible outcomes.

In the conclusion of their essay, Esser and Vliegenthart (2017, p. 17) mention “transnational Internet platforms, such as Facebook or YouTube.” These transnational media provide comparativists with opportunities to study “how those developments play out differently in different contexts” (Esser & Vliegenthart, 2017, p. 17). Alternatively, the study of these transnational platforms might reveal how they produce similar outcomes in different countries, contributing to more global, less country-specific political communication experiences.

Special Issue

This special issue seeks to address critical gaps in scholarship focusing on digital platforms that can offer transnational experiences. We hoped to revive the role of “time” as a rarely discussed but critical dimension of comparative research, especially when considering digital media. Finally, we have narrowed the scope to “political communication” rather than the entire field of communication and media studies. We offer a new set of studies that address comparisons across countries, platforms, and time; many submissions consider more than one of these dimensions.

This special issue highlights different approaches to comparison. For example, Vliegenthart et al.’s submission considers 25 countries, whereas Russmann et al.’s submission considers 12 countries and Farjam and Dutceac Segesten considers six language groups. Schatto-Eckrodt et al. and Ridout et al.’s contributions look at four platforms, whereas Schmidt et al.’s submission considered Facebook, Twitter, and legacy media. The number of cases for comparison varies and thus necessitates a different approach to contextualizing the study’s findings. A small number of cases allow for side-by-side comparisons of results, whereas a large number of cases do not allow for such a presentation. Instead, the results are discussed in relation to macro-level measures (e.g., Vliegenthart et al., 2024). Within political communication research, there is a call to contextualize findings describing the country/historical context (Rojas & Valenzuela, 2019). This contextualization is manageable in an article-manuscript when there are a handful of cases for comparison but can be extremely difficult when there are a large number of cases or the comparison is made across two or more dimensions (e.g., countries and time). The section below offers more details about how these studies have addressed the comparative dimension.

Country and Time Comparisons

Vliegenthart et al.’s (2024) submission considers public opinion about data-driven campaigning using a study conducted in 25 countries. They examine whether there are differences based on the regulatory context. In particular, public opinion may be similar in countries governed by the same data protection legislation, such as Europe (EU) and the General Data Protection Regulation (GDPR). They point out the difficulty of isolating these “different regimes of data protection” because the legislation has implications for other jurisdictions. Other jurisdictions are emulating GDPR principles. Furthermore, the legislation applies to EU citizens’ data, meaning that the legislation would apply to any platform, business, or researcher using data from EU citizens. This means that major social media platforms, such as Facebook and Twitter, must also comply with GDPR. As such, the legislation has far-reaching implications. The study is important for demonstrating the role of “transnational” policies in political communication; as such, it makes an important contribution to comparative political communication.

Russmann et al.'s (2024) submission studies personalization strategies on Facebook across 12 EU member states during two elections to the European parliament—2014 and 2019. Employing a double comparative approach—comparing across election periods and countries—the authors differentiate between centralized personalization, emotional personalization, and private personalization. Specifically, these aspects or varieties of personalization are studied using quantitative content analysis, coding for (among other variables) the presence or absence of personalization indicators. Using metadata such as user engagement metrics (likes, comments, and shares), the authors study not only the ways in which personalization was used by the selected political actors but also how these choices appeared to have influenced the engagement expressed by their Facebook followers and audiences. Using more than one way of comparison can be tricky—not the least in terms of fitting all the necessary background and methodological information within the typical length constraints of a research article.

Moving beyond elections, Farjam and Dutceac Segesten (2024) look at the legacy and new media coverage of migration, an important political issue. They look at six different languages and examine differences in an exploratory way, rather than outlining a theory of differences in language groups. This is another innovative approach to comparative communication, as the focus is not on countries and their different immigration policies. Instead, the study focuses on Twitter and the circulation of different media sources across the six language groups (see Matassi & Boczkowski, 2023, see Chapter 6 about languages). As noted, Esser and Vliegenthart (2017) discuss language groups as a point of comparison, but in practice, the comparative scholarship focuses on territorial/country differences (Zhao & Liu, 2020). As such, this piece makes an important contribution to scholarship. Farjam and Dutceac Segesten’s piece is also distinctive in picking up a line of comparative scholarship comparing legacy media to social media (see Matassi & Boczkowski, 2023, see Chapter 3). In addition, the piece also considers changes over time, that is, the rise and fall of attention to migration in the six language groups, offering another angle for the comparative perspective.

Platform and Time Comparisons

Continuing with the legacy news media as a point of comparison, Schmidt et al.'s (2024) contribution explores the degree of incivility across different platforms. In particular, they use legacy media as the baseline and then compare Facebook and Twitter using web tracking data in Germany. They find that online communication is more uncivil on Twitter, which they explain in terms of the pseudonymized usernames, rewards (e.g., likes) attached to this type of content, different types of networks (esp. compared to Facebook’s friends and family focus), and limited length of posts. In addition to platform comparisons, they compare over time. They examine the pre- and post-election periods, expecting that pre-election would involve higher levels of incivility and instead finding the reverse. Of all the contributions in this special issue, this paper is distinctive in providing explicit expectations of differences, including preregistered hypotheses.

Schatto-Eckrodt et al.’s (2024) contribution also considers legacy media but with a different twist. They looked at four platforms and assess the degree to which posts link to established media versus new media sources. Their platforms are Gab, 4chan, Twitter, and Reddit. They explain their selection of platforms as alternating between two types of social media (microblogs and discussion boards) and between established versus fringe platforms. The piece is also distinctive in connecting legacy media to social media, similar to Farjam and Dutceac Segesten’s submission (also see Matassi & Boczkowski, 2023, Chapter 3). Like Farjam and Dutceac Segesten (2024) as well as Schmidt et al. (2024) and Schatto-Eckrodt et al. (2024) include a time dimension as a point of comparison, considering how the links to different sources change over time. They expect that once established media cover a critical event, this will reduce uncertainty and reduce the flow of conspiracy theories circulating; they found support for this theory.

Ridout et al. (2024) consider advertisements in the 2020 U.S. presidential elections on YouTube, Google Search, Instagram, and Facebook. They consider similarities and differences but frame these in terms of convergence and divergence. Specifically, they examine the extent to which messaging is similar or different across the four platforms, focusing on three content features—the use of partisan language, the tone of the advertising, and the goals of the advertisements. They find that the two Meta platforms converge in advertising with their high rates of partisan language, which they explain in terms of the platform’s ability to target ads based on political characteristics. YouTube differs from the other sources (divergence), instead resembling television. The resemblance to television is not in terms of video content but rather the strategy of avoiding strong partisan appeals. They consider a variety of explanations for divergence, including audience, digital architecture, and genre.

Limitations and Future Opportunities

With the ICA preconference and the subsequent special issue, our goal was to connect cross-national comparative work with platform and time-period research. Further efforts are required to integrate the time dimension into comparative studies about digital political communication.

We hoped to include research that spanned the globe and that the ICA connection would help us achieve that goal. However, in the final stages, we ended up with six papers focused on the Global North, replicating a pattern in the broader scholarship about communication and media studies (Braun et al., 2023; Matassi & Boczkowski, 2023). This outcome is partially explained by our choice to have an in-person ICA event instead of online. As a consequence, some accepted participants did not receive their visas in time to participate, despite our best efforts to support their visa applications with letters of support and confirmation of paper acceptance. Our choice of modalities likely deterred those who could not afford to travel to Toronto for a single-day conference. To ease the financial burden and equalize opportunities to participate, we offered this preconference (with associated meals and refreshments) for free. We acquired research funding, so we did not need to collect a registration fee from presenters or other participants.

Finally, our call for papers reflected ICA values around equity and diversity (Gardner, 2023). We achieved this goal for the preconference program. This balance was quite an accomplishment since the Political Communication division at ICA tends to have a lower percentage of female authors than other divisions; Computational Methods is the division with the lowest percentage of female authors, which we note, since most of our submissions and the papers in this special issue are based on computational methods (Braun et al., 2023). Our final set of papers involved changing authors’ roles, which left us with a special issue with mostly male first-authors. While we recognize this bias, we also point out that this bias was an outcome that we could not control (e.g., authorship order changes). Again, these are limitations of our special issue; further research might consider whether these biases reflect the broader field of political communication. The U.S. bias in political communication is well-documented (Rojas & Valenzuela, 2019); the gender bias in political communication has received less attention.

In conclusion, we sought to move comparative digital political communication research forward to explore country, platform, and time differences. As noted, we ended up with submissions that considered time with other dimensions of comparison. While this could be viewed as a limitation, we also note that the overlapping comparative dimensions fits with Matassi and Boczkowski’s (2023) advice to consider multiple dimensions of comparison. While they focused on overlaps between countries, media, and platforms, we consider overlaps in country, platforms, and time. We also wanted to note the practical challenge of tackling multiple dimensions of comparison within a single 8000-word article. We agree that this contextualization is important (Rojas & Valenzuela, 2019) but indeed a challenge for comparisons across multiple dimensions within a single article-length manuscript.

Lastly, we would like to thank everyone who submitted to and attended our preconference as well as our co-organizers in Canada: Mireille Lalancette and Thomas Galipeau. We would also like to thank the authors and reviewers for helping to ensure the special issue was completed on time.