Political Communication PDF

The Oxford Handbook of Political Psychology (3rd edn) Leonie Huddy (ed.) et al. https://doi.org/10.1093/oxfordhb/9780197541302.001.0001 Published: 2023 Online ISBN: 9780197541333 Print ISBN: 9780197541302...

The Oxford Handbook of Political Psychology (3rd edn) Leonie Huddy (ed.) et al. https://doi.org/10.1093/oxfordhb/9780197541302.001.0001 Published: 2023 Online ISBN: 9780197541333 Print ISBN: 9780197541302 Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 Search in this book CHAPTER 16 Political Communication  Dannagal G. Young, Joanne M. Miller https://doi.org/10.1093/oxfordhb/9780197541302.013.15 Pages 555–600 Published: 18 September 2023 Abstract This chapter argues that the conventional theories and processes at the core of the discipline of political communication are rooted in assumptions that no longer hold and contexts that no longer exist. Today’s media users experience decentralized, interpersonal, horizontal, networked politically relevant communication every day. Users experience socially contextualized messaging within a networked system that is predicated on the economics and logics of media fragmentation and micro- segmentation. Message producers, political campaigns, and social media platforms integrate users’ social psychological characteristics into messaging content, aesthetics, and dissemination strategies. This chapter outlines how these qualitative changes necessitate a reimagining of the eld of political communication. The shift away from traditional mass media models to networked, decentralized media systems through digital technologies has crucial implications for: (a) the scope of what constitutes political communication and (b) the integration of political psychology into the study of political communication. Keywords: media fragmentation, political polarization, entertainment, news, digital technologies, political discussion, political campaigns, political advertising, social media, media e ects Subject: Political Behaviour, Politics Series: Oxford Handbooks Collection: Oxford Handbooks Online All disciplines evolve and expand over time. With advances in analytical procedures and data availability, as well as paradigm shifts prompted by younger scholars and increased interdisciplinarity, such evolutions are to be expected. But when a discipline’s theories and processes are rooted in assumptions that no longer hold and contexts that no longer exist—it is high time for a fundamental reconsideration and reimagination. It is in that spirit that we approach our discussion of the eld of political communication.  In 2005, after having studied political communication for 40 years, Doris Graber, founding editor of the eld’s agship journal, Political Communication, proposed a de nition of political communication framed in terms of impact. Graber and Smith wrote, the eld of political communication “encompasses the construction, sending, receiving, and processing of messages that potentially have a signi cant direct or indirect impact on politics. The message senders or message receivers may be politicians, journalists, members of interest groups, or private, unorganized citizens” (Graber & Smith, 2005, p. 479). In this chapter, we embrace and expand upon Graber and Smith’s (2005) already-broad de nition and conceptualize political communication to encompass intentional, strategic, campaign-oriented discourse, Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 but all forms of communications that “potentially have a signi cant direct or indirect impact on politics” (p. 479). Political communication is a discipline that is inextricably linked to context. As such, the logic, structure, and action-potential of media technologies (also known as “media a ordances”; see Faraj & Azad, 2012) are central to any understanding of the ways in which the media shape and are shaped by public opinion and political behavior. We propose that the shift away from traditional mass media models to networked, decentralized media systems through digital technologies has crucial implications for: (a) the scope of what constitutes political communication and (b) the integration of political psychology into the study of political communication. The organization of this chapter is as follows. First, we brie y review the accepted narrative of the eld. Next, we discuss the ways in which the media landscape has changed, focusing on two broad and overlapping dynamics: (1) the decentralized, interpersonal, networked capacity of digital technologies, and (2) the economics of micro-segmentation. This is followed by a discussion of the implications of these dynamics for the study of the political psychology of communication processes—in terms of both what we study and how we study it—and thoughts about implications and future research. p. 556 1. Established History of Political Communication As readers of comprehensive reviews of any research domain as vast as political communication are aware, authors must impose a structure on the literature, lest the text becomes a cacophony of theories, methods, ndings, and implications. Although the speci cs of these structures can be idiosyncratic, based on the author’s and intended audience’s intellectual emphasis and expertise, the broad contours of these organizational structures can become emblematic of a eld. In political communication, the origin story is typically constructed in terms of three phases of media e ects—hypodermic (media e ects are powerful and direct), limited (media e ects are rare and contingent), and subtle/moderated (whereas direct persuasion via the media may not be all that common, other types of media e ects such as agenda-setting, priming, and framing can be consequential, especially under certain conditions and for certain types of people). For example, in their chapter for the second edition of this volume, Valentino and Nardis (2013) begin with a brief review of the three phases, then provide a comprehensive review of recent research, organizing their chapter in terms of four domains of in uence: attention, learning, attitude change, and action. Other recent treatments are similarly structured and comprehensive (e.g., Iyengar, 2017; Jamieson, 2017; and Pooley, 2006). In this section, we brie y describe the history of the three phases (we point readers to the references cited above for more detailed treatments) to set the stage for our reimagining of the political psychology of political communication. The birth of the formal study of media e ects dates to the explosion of mass media o erings in the rst half of the 20th century. First newspapers and magazines, then lm and broadcast radio, followed by television, all garnered the interest—and concern—of sociologists and psychologists. Scholars feared that such far- reaching media with such vast audiences might have powerful “hypodermic” e ects on the public’s political attitudes and behaviors (Lasswell, 1927, 1930, 1935). “In the early days of communication study, the audience was considered relatively passive and defenseless, and communication could shoot something  In a mass society, people are seen as easily influenced by mass media and political movements because they lack strong local or communal relationships to ground them. This makes them more likely to join mass movements or accept propaganda, often seen as a driving factor behind authoritarian regimes and mass culture. into them” (Schramm, 1971, p. 9). This perspective was rooted in post-war (both WWI and WWII) fears about propaganda, apocryphal (and exaggerated) accounts of the “mass hysteria” caused by H. G. Wells’ War of the Worlds radio broadcast (Cantril et al., 1940), studies of the impact of WWII war bond drives (Merton et al., 1946), as well as the dominant social science theories of the day—mass society theory in sociology (Blumer, 1951; Durkheim, 1897; Kornhauser, 1959; Mills, 1956) and stimulus-response theory in psychology (e.g., Guthrie, 1935; Hull, 1943; Thorndike, 1898; see Bineham, 1988, for a critical review of the received wisdom about this rst phase of media e ects). Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 However, early media e ects studies were not bearing out the fears of a passive and defenseless audience. The People’s Choice Study (Lazarsfeld et al., 1948) and the American Soldier Studies (Merton & Lazarsfeld, 1950) pointed to limited direct e ects of media messages on attitudes and behaviors, instead pointing to processes like selective exposure, selective attention, and individual di erences in media e ects (Klapper, 1960; Lazarsfeld et al., 1948; Schramm & Carter, 1959; Sears & Freedman, 1967). People sought out media content that supported their preexisting views, avoided content that contradicted those views, and varied in their responses to media based on factors like education and religion, all of which suggested that media p. 557 were far better at reinforcement than persuasion (Lazarsfeld & Merton, 1948; Klapper, 1960). As Valentino and Nardis (2013, p. 2) summarize, “citizens, it seemed, didn’t pay much attention to politics, and when they did receive new information, it was primarily from friends and relatives they considered authorities on the subject” (Katz & Lazarsfeld, 1955). This “Two-Step Flow” notion of media in uence— where political media content is ltered through “opinion leaders” in a social network and then relayed to less interested members of the group—helped to explain why mass opinion change seemed so rare (see Druckman et al., 2018, for a recent review and empirical con rmation of the two-step communication ow theory). The cognitive revolution in media e ects research in the 1980s–1990s, with its focus on indirect e ects of mediated messages on construct salience and judgment formation, meant that scholars could still assert that politically important media e ects existed, even while acknowledging that strong, direct persuasion e ects often did not. The agenda-setting e ect of mass media (e.g., Iyengar & Kinder, 1987; Miller, 2007; McCombs, 2004; McCombs & Shaw, 1972; McCombs & Shaw, 1993) posited that mass media might not be in uential in directly persuading audiences or telling them what to think, but were quite e ective in telling the public what to think about. This increased the relevance of construct salience, mental models, and schema theory to the study of political communication and news e ects. Soon the concepts of media priming (e.g., Iyengar & Kinder, 1987, Miller & Krosnick, 1997) and framing (e.g., Entman, 1993; Gross, 2008; Iyengar, 1991; Nelson & Oxley, 1999; Nelson et al., 1997; Scheufele, 1993; Scheufele & Tewksbury, 2007), as well as the interconnections between agenda-setting, priming, and framing began to garner (and still receive) signi cant attention (e.g., McCombs & Ghanem, 2001; McCombs et al., 1997; Roskos-Ewoldsen et al., 2002). The cognitive revolution also highlighted the logical pairing between the elds of political communication and political psychology. Such work framed media e ects in terms of schema theory and mental models, and brought with it a renewed focus on individual di erences in media e ects. As audience “needs,” “traits,” and “evaluative dispositions” were found to moderate the impact of media on individuals, and the very mechanisms underlying the e ects (e.g., Domke & Wackman, 1998; Miller, 2007; Miller & Krosnick, 2000; Oliver, 2002; Valentino & Nardis, 2013), human psychology became vital to the study of political media e ects.  2. Phases of Media E ects and the Changing Media Landscape This established history of the eld of political communication, from hypodermic needle theory to limited e ects to subtle/moderated e ects (agenda-setting, priming, framing, and their nuances) is characterized by shifts in theoretical perspectives resulting from changes in the media context and methodological advances. Such advances often necessitated revisions of previous conventional wisdom. The eld of propaganda studies (e.g., Lasswell 1927), from which the hypodermic needle theory was born, was de ned Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 by the logic of broadcast media, through which centralized, powerful sources of information could deliver homogenous content to disconnected masses. With few opportunities to “talk back” to powerful message senders, few options for diverse content, few incentives for media outlets to produce diverse content, and no media infrastructure connecting members of the mass to one another, scholars feared that media audiences would be susceptible to large-scale top-down propaganda campaigns. p. 558 With advances in survey and experimental methodologies and statistics, as well as new theories generated by the social-cognitive revolution in psychology (e.g. Barone et al., 1997), scholars found little empirical support for the hypodermic needle theory, concluding that the media had very limited e ects (especially in the context of persuasion). Changes in the media landscape also played a role in the shift from phase one to phase two—greater media options (in both form and content) in the mid-20th century gave individuals more choice. For the most part, individuals chose content that did not challenge their pre-existing beliefs, attitudes, and behaviors. As media continued to fragment and diversify in the late 20th century, scholars pursued forms of media e ects beyond direct persuasion, including subtle cognitive e ects like agenda- setting, priming, and framing reviewed above. The eld also started to respond to the diversi cation of politically relevant content in early 2000s, integrating these cognitive theories into the study of political entertainment including talk shows, late-night comedy programming, and satire. (Moy et al., 2006; Young, 2004, 2008). Expanding the study of media e ects to sources not traditionally de ned as “news” or “public a airs” has shown that just about any politically-adjacent content can provide information, set agendas, prime issues and traits, frame stories, and encourage political discussion and participation. Starting with the earliest day of media e ects research, scholars had de ned “mass media communication” in terms of broadcast media such as newspapers, radio, and television—media that “enabled one or a few individuals to reach an audience of many” and through which “feedback [from message receiver to message sender] was minimal” (Reardon & Rogers, 1988). Such “linear and unidirectional” (Reardon & Rogers, 1988) conceptualizations of communication such as Lasswell’s (1948) “who said what to whom with what e ect” made sense given the largely top-down, unidirectional nature of media technologies at the time. And throughout most of the 20th century, mass media messages were produced and distributed by powerful message producers—who controlled the dissemination of messages to large, un-networked mass audiences. They were sustained either by the government, or (most notably in the US context) by media corporations selling access to a large mass audience to advertisers. But advances in digital technologies and social media have transformed both the experience of media users and the economics of media industries. Today’s media users experience decentralized, interpersonal, horizontal, networked politically relevant communication every day (Klinger & Svensson, 2015). And they experience this socially-contextualized messaging within a system that relies on the economics and logics of micro-segmentation. Both of these changes make it necessary to position political psychology more centrally within the eld of political communication. In sum, shifts in the dominant theoretical approaches to the study of political communication throughout the 20th century were dictated, in part, by the changing media landscape. We submit that despite numerous theoretical and conceptual advances, the “third phase” of the eld of political communication research is predicated on the existence of a mass media system that—by and large—no longer exists (Cha ee & Metzget, 2001). It is high time for a rethinking of the eld of political communication. To be sure, we are not  the rst to argue that the qualitative di erences between the media landscape of today and of yesterday necessitate such a rethinking. For example, Robison (2019) proposes an approach that centers on propaganda studies. And in their introduction to a special issue of Public Opinion Quarterly focusing on new approaches to political communication research in a changing media environment, Edgerly and Thorson p. 559 (2020) highlight three themes that emerge from the articles in the special issue, prompted by a recognition of the fundamental changes to the media environment: (1) motivations to create and share political content (self-expression), (2) a renewed focus on digital inequities, and (3) methodological innovations. In this chapter, we focus on how the shift away from traditional mass media models to Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 networked, decentralized media systems through digital technologies has crucial implications for: (1) the scope of what constitutes political communication and (2) the integration of political psychology into the study of political communication. 3. Political Communication as Decentralized, Networked, and Interpersonal Digital technologies have become decentralized by design, rendering information redundant across the network and empowering individual users to create, share, and publicly react to information (Abbate, 2000). This “network media logic” (Klinger & Svensson, 2015) means that individuals are both audience members and content producers; both recipients and participants (Ritzer & Jurgenson, 2010). In the early 2000s, when platforms like Wikipedia, YouTube, Facebook, and Twitter reimagined the internet, they prioritized interactivity, user-generated content, and horizontal interpersonal networks (O’Reilly, 2009). Web 2.0 took the decentralized control of the internet and placed it within overlapping socially networked communities of friends, family members, colleagues, and neighbors. No longer was the audience a disconnected mass. No longer was the audience a mass at all (Cha ee & Metzger, 2001). And as the lines between producer and receiver of information became increasingly blurred, no longer was the audience “an audience,” but perhaps best described as “the people formerly known as the audience” (Rosen, 2008). By embedding political communication within a social context, digital technologies and social media require that we increasingly think of all political communication in terms of interpersonal communication (Holbert & Geidner, 2009; Shah et al., 2017). Because of political communication’s roots in rhetoric and top-down propaganda research, and because of the challenge of studying informal interpersonal communication, most historical considerations of political communication have not included examinations of discourse between and among regular people (McNair, 2011). And although political communication scholars have long studied interpersonal political discussion and deliberation (Eveland et al., 2011; Gastil et al., 2008; Stromer-Galley, 2003; Wojcieszak & Mutz, 2009), these examinations have been considered separate from the study of strategic political communication, a distinction that is increasingly obsolete. Recall that in the early days of media e ects research, as strong persuasive e ects of media messages were proving to be elusive, researchers did nd indirect in uence of media—in the form of communication between people, through social networks and trusted opinion leaders (Druckman et al., 2018; Lazarsfeld et al., 1948). Today, with interpersonal communication embedded within digital media systems, constructs and processes relating to how people perceive, interact with, and respond to one another are increasingly relevant (Perlo , 2015). Cappella et al. (2015) describe, for example, the role of “impression and relationship-management” motives as increasingly important to the study of media e ects in our socially p. 560 networked media environment (Cappella et al., 2015, p. 11). Characteristics and dispositional traits like con ict orientation (Sydnor, 2019), need for uniqueness (Lantian et al., 2017), empathy (Simas et al., 2020) and status-seeking (Petersen et al., 2021) all help bridge the gap between political psychology and the social context in which political communication is increasingly embedded. Walther (2017) recommends scholars make use of meta-constructs that bridge the media-interpersonal divide, including social meta-constructs  like relationships, interactivity, mutual in uence, and social goals. Such approaches necessitate conceptual shifts from “factors to actors” (Cappella, 2017) and require scholars to recognize the interdependency of individuals within a social system and to consider how individual-level variables interact with explicitly social media contexts in studies of processes and e ects. 4. Political Communication as Fragmented and Micro-targeted Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 Accompanying these transformations that compel us to recenter interpersonal communication within the eld of political communication, are changes in media logics (see Altheide & Snow, 1979) and economics that encourage micro-targeting (Klinger & Svensson, 2015). As the era of analog and broadcast gave way to cable and digital through the 1980s–1990s, media fragmentation—or the explosion in the number of media o erings—di used audiences across hundreds of outlets, shrinking each individual audience and giving way to an era of media specialization and deliberate audience segmentation (Turow, 1997). Political communication scholars expressed concerns about the potential for media fragmentation to create “distinct issue publics” (Gurevitch et al., 2009, p. 170) and blur the lines of what constitutes professional journalism in ways that reduce public scrutiny and polarize the public (Mancini, 2013). The eld produced empirical evidence of various forms of fragmentation-related polarization; ideological polarization stemming from exposure to like-minded content (Baum & Groeling, 2008; Stroud, 2011); and political interest polarization, whereby individuals who preferred entertainment over information could drop out of the political media ecosystem altogether (Prior, 2007). Specialized cable channels and the Internet of the early 2000s were indeed fragmented, but the era of Web 2.0 born after 2004 (O’Reilly, 2009) transformed specialized media to “micro-targeted” media— communications that utilize user data to develop highly tailored media experiences driven by user socio- demographics, psychological characteristics, geographic data, and online consumer and social behaviors. Such connectivity data (Van Dijck, 2013) inform algorithms designed to maximize site visits and time on site in a hyper-competitive media environment. The economics of media competition have always encouraged the creation of public a airs content that is decontextualized, personalized, emotional, con ict-driven and dramatic (Bennett, 2016; Spencer & Spencer, 2007). Such norms lead to news content that contributes to various cognitive e ects; from horse-race and strategy coverage that fuels political cynicism (Bennett, 2016; Cappella & Jamieson, 1997; Patterson, 1993) to episodically framed personalized news (see Iyengar, 1991) that increases the audience’s perceptions that individuals (rather than institutions or government) are responsible for their own political, social, or economic fate (Gross, 2008). Scholars warn that the p. 561 increasing consolidation of news industries and reduced quality of journalism, coupled with media fragmentation and polarization, will have detrimental e ects on democracies around the globe (Van Aelst et al., 2017). Moreover, these dynamics are playing out in the midst of a vast sea of user data—data that increasingly inform the behaviors of media producers, regarding media content, argumentation, tone, aesthetics, and distribution (Deibert, 2019; Jamieson, 2013). Whereas psychology used to enter the media e ects conversation in the form of cognitive processing mechanisms and “individual di erence variables” (see Oliver, 2002), today psychological processes and traits are anticipated by content producers and distributors. Audience psychology is “baked into” media products and media strategies from the start, to attract, arouse, and solidify connections with individual users and viewers. This is perhaps most obvious in the context of psychologically micro-targeted political advertising through social media platforms (Krotzek, 2019; Kruikmeier et al., 2016; Zarouali et al., 2022). But the splintering of the information ecosystem has also increased the ability (and need) of message producers across media forms to create emotionally evocative, psychologically compatible (to their intended audience) content—in terms of both substance and aesthetics. Psychological traits such as tolerance for ambiguity which correlate with political ideology (Jost et al., 2003) and aesthetic preferences They not only tailor what they say (the substance) but also how it looks (the aesthetics) to appeal to their audience’s psychological preferences.  (Cleridou & Furnham, 2014) may account for the distinct look and feel of genres preferred by liberals and conservatives (Young, 2019). In other words, media content is strategically created to attract speci c audiences, based on their psychological make-up, preferences, and identities. Such “micro-segmentation” is dominating political media decisions and contributing to the production of content that will persuade connected individuals to watch, to use, to read, to engage, and to return. We assert that these drastic changes to the media landscape require even greater integration of political psychology into the study of political communication—especially the psychology of aesthetics (Jacobsen, 2006; Pels, 2003), political ideology (Jost et al., 2003), social/political identity (Huddy, 2001; Huddy et al., 2015), and emotion Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 (Albertson & Gadarian, 2015; Webster, 2020; Young, 2019) in ways that eschew any vestiges of the notion that political communication consists of a distinct non-strategic content creator, on the one hand, and a passive content receiver, on the other. Both transformations stemming from the digital media ecosystem—the decentralized, networked, interpersonal logic on the one hand and the fragmented, micro-segmented logics and economics on the other—are interrelated and mutually reinforcing. In the following pages, we re ect on the implications of these transformations for various political communication theories and processes and discuss their signi cance for the role of political psychology within political communication. Harkening back to Lasswell’s (1948) model of media communication, we rst discuss the ways in which these media transformations have fundamentally revolutionized the production of political communication (the “who”). We then review research that reconceptualizes the audience as active participants in the political communication process (the “whom”). Next, we discuss some of the ways in which the decentralized/networked/interpersonal and micro-segmented media landscape a ects media content (the “what”), focusing speci cally on emotional content and the production and spread of misinformation. Finally, we discuss how these processes a ect message processing, our shared (or not) “reality,” and political polarization (the “with what e ect” of Lasswell’s framework). Our choice to organize this reimagining of the eld of political communication in terms of the who, the p. 562 whom, the what, and the e ects is not without its drawbacks. In making this choice, we are conscious of the fact that we are adopting a framework borne out of a media context that no longer exists. After much deliberation, it seemed to us that Lasswell’s (1948) framework actually spotlights, rather than obscures, the stark di erences between the media landscape of yesterday and the media landscape of today. 5. The “Who” We begin with the who. The conventional conceptualization of the sources of political communication focuses on elites—primarily news executives, professional journalists, politicians, and pundits—as the sole broadcasters of information, and of professional media organizations as the conduits of that information. In this section, we explore recent theoretical and empirical work that helps us reconceptualize and broaden our understanding of the “who.” 5.1. From Gatekeepers to Curated Information Flows With the advent of decentralized control over the ow of information and the shift from a vertical and disconnected to a horizontal and networked experience of media users, the role of information “gatekeepers” has fundamentally changed (Shoemaker et al., 2001). Whereas news executives and organizations used to have the unique power to “winnow, shape, and prod potential news messages…into those few that are actually transmitted by the news media” (Shoemaker et al., 2001, p. 233), that gatekeeping power has now been di used across individuals, politicians, political organizations, social media platforms, and even rogue political actors.  Individual citizens, in the role of information producers and sharers, play the role of “citizen journalists” (Wall, 2015) or “participatory journalists” (Fröhlich et al., 2012), motivated by dissatisfaction with traditional news, a desire to engage/challenge dominant narratives, and to connect with others to acquire ideas and experiences (Fröhlich et al., 2012; Kim & Lowrey, 2015). On the recipient side, citizen journalism is viewed as most credible by those who are already cynical and skeptical about media in general (Carr et al., 2014). Videos of hostile encounters between citizens and government o cials such as the military or police have complicated dominant elite narratives (Ali & Fahmy, 2013; Meraz, 2009), negatively a ected viewers’ perceptions of law enforcement (Boivin et al., 2017; Parry et al., 2019), and contributed to the growth in Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 social justice movements (Anthony & Thomas, 2010). Recognizing the value—and appeal—of citizen journalism, some news organizations have integrated more user generated content (Thurman, 2008) and non-elite sources into their reporting (Hermida et al., 2014). Such shifts in the de nitions of “journalism” and “objectivity” are illustrated in scholars’ calls for professional journalists to reject objectivity, tap into their authentic moral voices, and draw upon their experiences, communities, and identities as they engage with events and issues (Carlson et al., 2021). While these changes are operating at the level of media production, they can a ect how individuals select and process political information. Kelly (2019), for example, illustrates how perceptions of objectivity and credibility of news stories from unknown sources are p. 563 driven by pre-existing beliefs, leading to higher perceptions of bias when encountering balanced news than belief-discon rming news. The declining formal role of elite journalistic gatekeepers also means more opportunities for direct strategic communication from political elites to citizens, although empirical examinations reveal these instances of “direct-to-voter” social media communications are rare (Kleis Nielsen & Vaccari, 2013). In The Rhetorical Presidency, Tulis (1987) introduced the concept of “going public” to refer to presidents bypassing interactions with Congress by directly addressing the American people on matters of policy. Scacco and Coe (2021) expand on this with the concept of the “Ubiquitous Presidency” whereby digital technologies allow certain elites (namely, President Trump) to communicate directly with followers and avoid journalistic inquiry and accountability, while still relying on message ampli cation and legitimacy from news outlets. Direct communications from public o cials to supporters increases opportunities for “para-social” interactions, through which individuals feel that they have intimate relationships with people they do not actually know in real life (Horton & Wohl, 1956). These perceptions can shape both cognitive and a ective aspects of message processing (Lee & Shin, 2012), including increasing a nity for like-minded politicians through the para-social bond created through exposure to their Twitter feeds (Paravati et al., 2020). Meanwhile, news organizations have become increasingly dependent on search engines and social media for the distribution and reach of their content (Kleis Nielsen & Ganter; 2018), leading some to call these “digital intermediaries” our new gatekeepers (Kleis Nielsen, 2016; Powers, 2017). Yet, digital intermediaries are not acting alone in curating the ow of information. User behavior and preferences inform the criteria that intermediaries use to shape information curation and distribution (Pariser, 2011). Wallace (2018) urges that we think about digital gatekeeping in terms of individual actors, algorithms, and platform incentives; or: “who is selecting which information according to what selection mechanism, and how is the news item framed before reaching the public?” (p. 288). Thorson and Wells (2016) conceptualize these processes not as gatekeeping, but as “curated information ows” through which journalists, strategic communicators, individual media users (personal curators), social contacts, and algorithmic lters all shape an individual’s mediated experience.  5.2. Entertainment TV and Beyond as Conduits for Political Communication Not only does today’s qualitatively di erent media environment require scholars to rede ne the “who” as being everyone, it also requires us to think more broadly about the channels of that communication. Media fragmentation and accompanying changes in the economics of media fueled an evolution in political media forms and genres through the 1990s to today. The distinction between news and entertainment that had been embedded into network television broke down as cable and the internet allowed—and encouraged— the creation of new hybrid genres of information (Baym, 2017; Chadwick, 2017; Williams & Delli Carpini, Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 2011; Young & Gray, 2013). Political communication scholars began studying the political content and e ects p. 564 of late-night talk shows (Niven et al., 2003; Young, 2004), daytime talk shows (Baum & Jamison, 2006; Glynn et al., 2007), historical reenactment lms (LaMarre & Landreville, 2009), and narrative documentaries (Whiteman, 2004). Rather than remaining handcu ed to the false distinction between “political communication” and “entertainment,” theory-driven analyses have revealed important political consequences of exposure to programming designed to entertain. In the context of late-night comedy research alone, researchers have found evidence of political and public policy knowledge acquisition (Becker & Bode, 2018; Cao, 2008; Hardy et al., 2014), increased political discussion and participation among viewers (Feldman & Young, 2008; Landreville et al., 2010), increased attention to and awareness of politics and public policy (Brewer et al., 2018; Feldman et al., 2011), and increased information seeking from—and attention to—more traditional political content (Feldman & Young, 2008; Xenos & Becker, 2009). Importantly, many of these e ects of exposure are conditional upon the individual characteristics of the audience—including political knowledge (Young, 2004), political interest (Xenos & Becker, 2009), age (Cao, 2008), and ideology (Lamarre et al., 2009). Research has also begun to illuminate the ways in which campaign communications are a ected by digital and social media a ordances, described by Perlo (2015) as “the quintessential media features that facilitate actions or trigger psychological heuristics” (p. 549). According to Faraj and Azad (2012), media technologies shape the potential for users to engage in certain actions by virtue of their structures and internal logics. These “action-potentials”—or social media a ordances—are especially implicated in political campaign communications (Bossetta, 2018), as evidenced by the fact that digital technology companies promote their platforms to candidates and work with campaigns to help them take advantage of platform characteristics to target likely voters (Kreiss & McGregor, 2018; see also Towner & Baumgartner, eds., 2021, for new research on how candidates and campaigns leveraged new media technology to reach potential voters during the 2020 elections). Whereas political communication textbooks through the mid-1990s typically limited the concept of political communication to news, political advertising, and political debates, contemporary scholars explore the content, psychology, and e ects of any mediated form of information that might have political implications. Recent work points to the relationship between videogame play reduced support for democratic and prosocial values (Bacovsky, 2021), increased postfeminist attitudes among viewers of YouTube makeup tutorials (Chae, 2021), the civic role of online meme culture (Penney, 2020), and the playful exploration of contentious issues related to gender and power on the social media platform, TikTok (Vijay & Gekker, 2021). Social scientists have also begun to explore complex phenomena such as “Black Twitter” as a place where people of color construct (Harlow & Benbrook, 2019) and perform (Florini, 2014) racial identity and advance counternarratives (Graham & Smith, 2016). Political scientists have also studied the intersection of sports, media, and politics, including how perceptions of NFL’s Colin Kaepernick’s public protests of police brutality are shaped by political and racial attitudes (Stepp & Castle, 2021; Towler et al., 2020).  6. The “Whom” The reimagining of the “who” discussed above blurs the lines between who and whom to such a degree that p. 565 we risk creating a distinction without a di erence. Not only are the traditional receivers of political communication now also producers of that content, but the new media landscape provides them with more opportunities to be active and e ective seekers of information, while also incentivizing the strategic curation and packaging of content by producers (at both the elite and mass level) seeking to attract their Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 intended audience. The notion of agency—who has it and what it even means—has been upended. 6.1. Uses and Gratifications Uses and Grati cations (U&G) approaches to the study of media, popularized in the 1970s (Katz et al., 1973; Rubin, 1983), o er a valuable framework to consider both why people engage with media and how they process and are shaped by media experiences as a result. When originally developed, U&G’s assumption of an active audience that “used media” to meet certain needs and grati cations contrasted sharply with that of the old passive-audience-centered media e ects paradigm. When integrated with e ects mechanisms, researchers found the impact of media exposure on outcomes like learning and message engagement was contingent on the needs and grati cations that guided media behaviors in the rst place (Rubin, 1983). In the digital media environment, the networked, interactive, and customizable “media a ordances” (Sundar, 2008; Sundar & Limperos, 2013) ought to shape not only media e ects mechanisms, but the very needs and grati cations users seek to satisfy through media use (see Perlo , 2015). Perlo (2015), for example, suggests that future U&G approaches integrate the a ordances of digital technologies with unique psychological characteristics of the user to posit novel sets of uses and grati cations that individuals might seek to satisfy through media use. For instance, he points to Bergman et al.’s (2011) nding that although narcissism does not predict time spent using social media among millennials, it does predict the reasons why they use social media—to increase their number of friends, share their activities, and project a positive image. Since motivations for media use shape a ective and cognitive processes associated with that usage, motivational distinctions ought to be factored into political media e ects mechanisms. Perlo also urges the integration of neuroscienti c approaches to better understand neuropsychological processes that correlate with individuals’ media orientations and anticipated rewards. Consistent with the e cacy of such approaches, Sherry (2001) found signi cant correlations between television viewing motivations and various traits that share neural substrates, namely: mood, task orientation, and cognitive rigidity. Early integrations of U&G approaches helped explicate how media user orientations and motivations shape cognition, attitude, and behavior e ects (McLeod & McDonald, 1985). More recent applications have explored how motivations to use mobile messaging apps a ect political discussion and e cacy (Pang, 2018), how audience motivations rooted in social utility shape political participation through social media use (Chen & Chan, 2017), how surveillance and escapism grati cations predict news consumption among college students (Diddi & LaRose, 2006), and how social media platform adoption is associated with motivations including hedonism, self-esteem, belonging, and reciprocity (Pai & Arnott, 2013). Illustrating the integration of interpersonal communication theory with U&G approaches in the context of political p. 566 communication, Pennington and Winfrey (2021) nd that individuals who are more social and interpersonally-focused in their online behaviors are less likely to discuss politics on Facebook. the U&G framework highlights the active role users play in selecting media based on their needs and motivations, which influences their experiences and outcomes in various contexts, including politics.  6.2. The Importance of Social Identity There is increasing recognition in the eld of political communication that the fragmented and interpersonally networked information landscape makes identity-driven motivations increasingly relevant to the processes outlined above (see, for example, Slater, 2007). Tajfel and Turner’s Social Identity Theory (SIT; see Tajfel & Turner, 1979) and Turner and Oakes’ concept of self-categorization (1986) o er useful frameworks for considering the ways in which identity shapes media preferences and e ects. Even before the birth of social media, Blumer & Kavanagh (1999) discussed the intersection of media fragmentation and Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 identity, noting the “increased social di erentiation and specialization, [and the] fragmenting [of] social organization, interests, and identities” through media. In his 1998 Ithiel De Sola Pool lecture, Bennett (1998) highlighted declining con dence in institutions and the rise in “lifestyle politics” resulting from economically driven social fragmentation. He also characterized the splintering of the media system as both a response—and contributor—to these trends. In 2008, Bennett and Iyengar revisited the concept of identity, arguing that among young people in particular, we see “shifting and far more exible identity formations that require considerable self-re exivity and identity management.” (p. 716), including through their media behaviors. But while political communication scholars were recognizing the increasing salience of identity-based motivations in political life, political psychologists were pointing to the limits of SIT in political contexts (Huddy, 2001). Yet, even while urging caution, Huddy (2001) acknowledged the capacity of “category salience” to shape identity in a given context with potentially important consequences. We propose that interpersonally networked and micro-segmented digital political media are tailor made (quite literally) to be just this: a context designed to activate and prime the salience of our identity categories. The microtargeting of symbols and texts through digital media, combined with the incentives to “perform identity” in an interpersonally networked social media context (Vaidhyanathan, 2018), speak not only to group “boundaries” but also to group “meaning” consistent with self-categorization and social identity theories (Haslam et al., 1992). Recent work points to the intersection of user psychology and social media logics in incentivizing fast, inattentive—and often identity-based—motivations, over accuracy-based motivations (see Fazio, 2020; Pennycook et al., 2021). In the US context, socio-demographic, political, and psychological characteristics are increasingly enmeshed and overlapping (Mason, Chapter 24, this volume; Mason, 2018). This partisan separation along racial, religious, and geographic lines is accompanied by an epistemic separation on the left and the right (Oliver & Wood, 2018). This “social” and “epistemic” sorting, combined with the economics and logics of fragmented and digital media, means that identity categories are used, reinforced, and cultivated through the process of political communication itself, with implications for political polarization and a ective polarization (see Iyengar et al., 2012; Young, 2023). These dynamics are the essence of Slater’s (2007, 2015) Reinforcing Media Spirals (RMS), which proposes that identity motivates media behaviors in ways that p. 567 reinforce that identity. Importantly, according to RMS, when one’s identity is perceived to be under threat, these patterns are exacerbated, such that media use not only reinforces identities, but can facilitate their polarization (see Long et al., 2019).  6.3. The Psychology of Media Selectivity: Beyond Political Attitudes Whereas media fragmentation facilitates selective exposure and avoidance behaviors based on political ideology and political attitudes, it also invites people to make media selections based on content preferences and psychological traits. Sydnor (2019), for instance, illustrates that individuals who are con ict averse intentionally avoid uncivil political media, hence contributing to gaps in political engagement based on gender and race. Gerber et al. (2011) nd that individuals high in openness, extraversion, and emotional stability are most likely to consume political media and that these traits shape preferences for speci c Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 genres of political information as well. People who prefer entertainment over information can selectively avoid political content altogether, leading to a decline in political knowledge and voter turnout among those least politically inclined (Prior, 2007; To & Kalogeropoulos, 2020). But just as social media might allow incidental exposure to ideologically diverse political information through friends and family, so too might social networks diversify media experiences in general, encouraging, for example, news consumption among younger and less politically interested users (Fletcher & Nielsen, 2018). Integrating the study of aesthetic preferences—style or taste—into political communication research presents a rich opportunity for political psychologists studying media selectivity patterns and e ects. Pels (2003) explains that “political style or taste refers to the fact that judgments of persons…normally have a holistic and unarticulated character…” and proposes that bringing aesthetics and “style” into the study of politics “usefully bridges the divide between form and content, detail and essence, presentation and principle, sentiment and reason, and hence relativises a one-sided cerebral or rationalist approach to political behavior” (p. 48). Integrating the concepts of media “style,” “packaging,” or “aesthetics” allows us to move beyond classic message features (like political topics, argument quality, message source, or medium), to consideration of look, feel, tone, and style of political symbols and texts. Understanding how such characteristics resonate with user psychology may help explicate selective exposure, selective processing, and even selective production (see Cui et al., 2020). This is especially true in a socially and epistemically sorted and yet networked political media ecosystem in which media organizations are economically incentivized to tailor their content to the presumed psychological characteristics, identities, and aesthetic preferences of their intended audience. One example of the integration of aesthetics into political psychological approaches to communication is found in the study of political entertainment. As reviewed above, the boundaries between entertainment and news have eroded (Williams & Delli Carpini, 2011; Young & Gray, 2013), resulting in media content and genres that are increasingly hybrid; unruly combinations of forms, categories, and structures (Baym, 2017; Chadwick, 2017). These combinations, part entertainment and part news—part serious and part play— create fertile ground for political psychologists studying political communication, as individuals’ tolerance p. 568 for ambiguity likely a ects their comfort with such blendings (Young, 2019). Work on the psychology of aesthetic preferences indicates that tolerance for ambiguity and need for cognition are correlated with preferences for artistic (Wiersema et al., 2012) and musical styles (Rawlings et al., 2000). Young (2019) thus proposes that ambiguous ironic satire is compatible with the psychological correlates of social and cultural liberals—high in tolerance for ambiguity and need for cognition (see Jost et al., 2003), hence accounting for the left-leaning “bias” of much political satire, as well as the right leaning “bias” in didactic, literal, morally serious “outrage programming” (see Berry & Sobieraj, 2013). Such considerations are important as scholars study how individuals engage with politically relevant hybrid media forms including historical reenactment lms (LaMarre & Landreville, 2009), political talk shows (Roth et al., 2014), and even journalists’ Twitter pro le pages (Ottovordemgentschenfelde, 2017). Another fruitful application of the concept of “political media as aesthetic” comes from recent work on the psychology and appeal of populism. Its thin ideology, appeal to “traditional popular values,” and antagonistic divide between “the people” and “elites” (see Mudde, 2004; Stanley, 2008) have led political communication scholars to describe populism itself as a “communication style” rather than a political  ideology (Bracciale et al., 2017; De Vreese et al., 2018; Jagers & Walgrave, 2007). Populism as an aesthetic is thus emotionally evocative, intimate anti-establishment messaging focused on regular people that draws a contrast between the monolith of good people and corrupt elites (Bracciale & Martella, 2017). Such aesthetics are likely to interact with audience psychology in ways that appeal to and mobilize certain kinds of individuals more than others, such as those who score lower in agreeableness (Bakker et al., 2021), higher in need for closure and certainty (Kruglanski, Molinario & Sensales, 2021), and higher in narcissism and psychopathy (Nai & Martinez i Coma, 2019; for a review of the literature on political language, see Hopkins, Chapter 9, this volume). Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 6.4. Echo Chambers Another implication of the increasingly interpersonal nature of media experiences and the micro- segmentation of audiences is the potential formation of echo chambers and lter bubbles. As media choice increases, individuals can increasingly “opt in” or “out” of media experiences, facilitating selective exposure to and avoidance of ideologically consonant—or even politically relevant (Prior, 2007)—media content. In their paradigm-challenging essay, Bennett and Iyengar (2008) suggested that selectivity behaviors would be so great in this new media landscape, that media persuasion e ects would all but disappear, even as other e ects like agenda-setting or priming might continue. “As media audiences devolve into smaller, like-minded subsets of the electorate,” they wrote, “it becomes less likely that media messages will do anything other than reinforce prior predispositions. Most media users will rarely nd themselves in the path of attitude-discrepant information” (p. 724). Such “echo chambers” or “ lter bubbles” became a growing concern among political scientists and political communication scholars from 2000 to the 2010s (Sunstein, 2001), as empirical examinations pointed to a public engaging in selective exposure and avoidance of politically dissonant mediated information in ways that could fuel political polarization (Stroud, 2011; see below). Yet, while echo chambers and lter bubbles are certainly made possible through media fragmentation, research has failed to produce substantial p. 569 evidence to support the notion that most people restrict themselves to only like-minded media content (see Guess et al., 2018; Guess, 2021). Although selective exposure to like-minded content does occur, and moral-emotional content from elites spreads e ciently within their ideological social networks (especially on the right) (Brady et al., 2019), evidence of selective-avoidance of belief-discon rming information is scarce (Garrett, 2009). Barbera et al. (2015) conclude that “homophilic tendencies in online interaction do not imply that information about current events is necessarily constrained by the walls of an echo chamber” (p. 1539). Users tend to have diverse media diets and vary in the extent to which they pay attention to political or current events information (Guess, 2018), and those users who do pay attention to like-minded political content pay more, not less, attention to belief-dissonant programming (Garrett et al., 2011, Nelson & Webster, 2017). Even on social media, where algorithms and users themselves increase the ideological homogeneity of their newsfeeds (Bakshy et al., 2015), users are still exposed to some belief-discon rming information. In fact, in the most fragmented context of all—the internet—selective exposure and avoidance are complicated by social networks, where social media posts of friends and family can serve as heuristics that guide ideologically diverse information consumption (Masip et al., 2018; Messing & Westwood, 2014; Flaxman et al., 2016; Zuiderveen et al., 2016), even trumping partisan selective exposure (Anspach, 2017). These results are supported by Dylko et al. (2018), who nd that although customized media experiences increase selective exposure dynamics in ways that increase political polarization, this e ect is limited to customization processes that are automatically embedded within a technology. When users have the ability to customize media experiences, these relationships shrink, again highlighting how user agency often diversi es—rather than homogenizes—media diets. Such ndings certainly call into question Bennett and Iyengar’s (2008) pronouncement of the “demise of the inadvertent audience” (p. 717). Whereas media  fragmentation facilitates selective exposure and avoidance, the interpersonal networked nature of the social media experience does not. Yet attending to and being receptive to belief-discon rming content are two di erent things that may be motivated by di erent goals. Knobloch-Westerwick and Klienman (2012), for example, nd that people engage in less selective avoidance of belief-discon rming information when they expect their side is going to lose, pointing to anticipated “informational utility” as individuals anticipate that exposure to belief- discon rming content “can aid individuals in making future decisions” (p. 171). Valentino et al.’s (2009) Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 work is consistent with this explanatory mechanism, with individuals seeking out belief-discon rming content as a way of monitoring their information environments. In this work, anxiety was found to fuel information seeking behaviors, thereby increasing exposure to belief-discon rming content, illustrating that these consumption behaviors are shaped by emotional needs of the audience. Finally, calls to relax the fears of echo chambers and lter bubbles are focused on the modal cases—the most likely outcomes for most people. This raises questions about whether we should be more concerned about the outcomes of users at the margins—culturally and politically. Boutyline and Willer (2017) for example, found that social networks (on Twitter) were most homogenous among those users with the most extreme political views. Evidence about the moderating role of political engagement is mixed; Guess (2021) nds a positive relationship between political engagement and media diet homogeneity, whereas Dubois p. 570 and Blank (2018) nd the reverse (see also Eady et al., 2019). Emotions such as anger may fuel the kind of dynamics that lead to concerning echo-chambers, while fear and anxiety may mitigate them (Wollebaek et al., 2019). Recent work by Stier et al. (2020b) indicates that populist attitudes interact with contextual factors to fuel selective exposure patterns, with populist attitudes contributing to lower (but not zero) exposure to traditional news and greater exposure to hyper-partisan outlets. The intersection of misinformation and echo-chambers has given rise to renewed interest in how misinformation may thrive and spread through online networks (Törnberg, 2018). This has led some to consider how mere exposure to belief-discordant information within a vocal and ideologically homogeneous social networks might still facilitate anti-democratic and polarizing outcomes. In Garrett’s (2017) condemnation of “the echo chamber distraction,” he highlights the potential dangers of “engagement echo chambers” or networks that have the capacity to promote falsehoods merely by “consistently a rming one view” (p. 371). 7. The “What” Both the networked/interpersonal nature of the media landscape today, as well as the expansion of media outlets and corresponding economic incentives to create micro-targeted content that will garner the largest audience possible, have had profound e ects on the content of political communication—the “what.” A comprehensive review of the myriad ways in which the new media landscape has reshaped content production is beyond the scope of this chapter. Here, we brie y review scholarship that highlights two such transformations—hyper-emotional content and the spread of falsehoods and conspiracy theories.  7.1. Emotional Content Since the early 2000s there has been a growing recognition of the importance of emotions in political communication (e.g., Brader and Gadarian, Chapter 6, this volume; Marcus et al., 2000; Neuman et al. 2007). For many years, this research was primarily focused on the e ects of political ads—most notably, the impact of negative ads on candidate evaluations and voter turnout (Brader, 2006; Chang, 2001; Geer, 2006; Kahn & Kenney, 1999; Mattes & Redlawsk, 2014; Tedesco, 2002, Valentino et al., 2011). The questions guiding this fruitful line of research center around the role of emotions in the political persuasion process Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 (e.g. Albertson & Gadarian, 2015). The decentralized, fragmented, and networked nature of today’s media environment has opened the door for a new set of questions about the ways in which media organizations and algorithms strategically push emotional content aimed at attracting clicks, likes, and shares, as well as the e ects of such content. These new lines of research have documented di erences in the positive versus negative emotionality of news content in traditional media outlets and on social media (e.g., Soroka et al. 2018), as well as whether information that is positive or negative in tone is more likely to be shared. Evidence on this second point is mixed, with some scholars nding that negative information is more likely to be shared than positive information (Godes et al., 2005; Hasell & Weeks, 2016), and others nding that p. 571 positive information is more likely to be shared (Berger & Milkman 2012; Kraft et al. 2020). Aaroe and Peterson’s (2020) work suggests that episodically framed news stories that resonate with our pre-existing cognitive biases are more likely to be shared and to shape our attitudes when they are experienced and processed through interpersonal networks. Meanwhile, Brady et al. (2017) nd that content that includes moral-emotional language is more likely to be shared than less emotional content. In her book, Disrespectful Democracy: The Psychology of Political Incivility, Sydnor (2019) demonstrates the central role that the psychological construct of con ict orientation plays in our understanding of the e ects of emotionally arousing media content. Through experimental and survey research, she demonstrates how people who are more “con ict avoidant” are more likely to experience negative emotions (anxiety, disgust, and anger) in response to televised incivility, whereas those who are “con ict approaching” are more likely to experience positive emotions (like being entertained and amused) in response to that same content. Those who are con ict averse are then turned o by political media and more likely to opt out of political content altogether. Building on the work of Mutz and Reeves (2005), which explored how televised incivility can increase attention while disrupting other forms of cognitive processing, Kosmides and Theocaris (2020) nd that social media incivility can actually induce positive emotions (enthusiasm), especially among those for whom the content is attitude-consistent. Borah (2013) points to the anonymity a orded by social media as one of the drivers of incivility and nds that uncivil news frames are a double-edged sword, increasing the credibility of a news article, on the one hand, while also decreasing political trust and e cacy, on the other.  7.2. Misinformation and Conspiracy Theories In 1944, former head of “Rumor Control” for the Massachusetts Committee of Public Safety Robert Knapp de ned “rumor” in Public Opinion Quarterly as “a proposition for belief of topical reference disseminated without o cial veri cation” (p. 22). He described rumors as providing detailed “information” spread by word of mouth that satisfy some emotional or psychological need of a community. It is the “spread by word of mouth” element that imbues the decades-old concept of “rumors” with increasing relevance in today’s social media ecosystem. In their prescient 2000 article, Kuklinski et al. argue that the conventional Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 “informed vs. uninformed” dichotomy con ates two very di erent kinds of uninformed people—those who do not know the facts, on the one hand, and those who believe false information (the “misinformed”), on the other. They called for more research on the causes and consequences of misinformation (“information that is false, but the person who is disseminating it believes that it is true”; Wardle & Derakhshan, 2018, p. 44), pointing to the potential deleterious e ects of a misinformed public that is likely to resist facts that contradict their misinformed beliefs. Vraga and Bode (2020) call for a di erentiation between information that is false and information that is merely unsubstantiated, such as operationalizing misinformation as “information considered incorrect based on the best available evidence from relevant experts at the time” (p. 138) for example. Such distinctions would take into account the uid nature of expert consensus, especially from such domains as science and law. Disinformation, in contrast, is “information that is false, and the person who is disseminating it knows it is p. 572 false” (Wardle & Derakhshan, 2018). E orts to spread information that is knowingly false violate the free will and transparency assumptions at the heart of persuasion theory (see Perlo , 2020). Disinformation is thus more akin to propaganda and deception. Propaganda and persuasion scholars have long recognized that mass media propaganda, absent supplementary interpersonal messaging, was unlikely to persuade audiences (Lazarsfeld & Merton, 1948). Because decentralized social media embed media messaging and user-generated content within interpersonal networks, disinformation and other forms of propaganda that are spread through these channels may be at a persuasive advantage (Young & McGregor, 2020). Jamieson (2020) chronicles how Russian propaganda operatives—especially online “trolls”—capitalized on the decentralized and networked logics of Facebook to exploit preexisting cultural cleavages during the 2016 US presidential election. She also illustrates how news organizations, driven by biases toward personalized, con ict-oriented news, legitimized and ampli ed these disinformation e orts. Munger (2020) argues that economic incentives that drive clickbait media lead to “credibility cascades” that enable misinformation to ourish. Meanwhile Garrett (2017) argues that the threat posed by deliberate politically strategic disinformation campaigns is far greater than the threat posed by the logics of the technologies themselves. He also explores how citizens might use the a ordances of social media platforms to increase information accuracy, accountability, and civility. Yet research on how elite-driven disinformation interacts with media a ordances, logics, and economics paints a far clearer picture of democratic harm than of democratic health. Rhetoric scholars have detailed how political elites such as Donald Trump (Mercieca 2018) make strategic use of the a ordances of the new media environment to spread falsehoods directly to the public. Using the metaphor of a polluted landscape, Philips and Milner (2021) explore how polluted information such as White supremacist ideology and conspiracy theories have permeated the media landscape and point to the interconnected, intertwined, and deeply networked nature of mis- and disinformation as being particularly pernicious and di cult to combat. Rosenblum and Muirhead (2020) identify a new form of conspiracism borne out of the new media landscape, which they de ne as conspiracy without a theory. “There is no punctilious demand for proofs, no exhaustive amassing of evidence, no dots revealed to form a pattern, no close examination of the operators plotting in the shadows. The new conspiracism dispenses with the burden of explanation. Instead, we have innuendo and verbal gesture…What validates the new conspiracism is not evidence but repetition…The e ect of conspiracist thinking once it ceases to function as any sort of explanation is delegitimation” (p. 3).  Domestic propaganda e orts abound in the new media landscape (Shin et al., 2018), sometimes originating with inaccurate online rumors that resurface repeatedly to an extent that accurate rumors do not. On Twitter, such misinformation is often repackaged and brought back to life strategically by partisan news organizations, hence transforming misinformation into disinformation (Shin et al., 2018). Moreover, misinformation can spread “farther, faster, deeper, and more broadly than the truth” (Vosoughi et al., 2018, p. 1146; but see Guess et al., 2019 for work that shows that the sharing of misinformation during the 2016 election was rare). Recognition of the potential negative consequences of belief in misinformation and conspiracy theories (Douglas et al., 2019; Kuklinski et al., 2000) has also spurred research that explores the Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 e ectiveness of content aimed at combating misinformation, through either inoculation or “prebunking” (Lewandowsky & van der Linden, 2021), refutations from unlikely sources (Berinsky, 2015), or through p. 573 correcting misinformation once it has taken hold (e.g., Chan et al., 2017; Garrett et al., 2013; Nyhan et al., 2020; Ophir et al., 2020; Trujillo et al., 2021; Walter et al., 2020; Young et al., 2018) 8. The “E ects” Lastly, we review scholarship that demonstrates how the changing media landscape necessitates a reimagining of the e ects of political communication, focusing on message processing, cognitive theories of media e ects (speci cally, agenda-setting), and political polarization. 8.1. Message Processing Dual-process models of information processing like the Elaboration Likelihood Model (Petty & Cacioppo, 1986) and the Heuristic Systematic Model (Chaiken, 1980) continue to dominate cognitive psychological approaches to political communication processing and e ects (for a review of this literature see Jerit & Kam, Chapter 17, this volume). Both models conceptualize individual motivation (see Chen et al., 1999) and ability as the factors that determine whether individuals are thoughtful and analytical as they process information, or if they are guided by heuristic cues, like source credibility (Sundar, 2008), attractiveness, or aesthetic features of the message. The notion that information processing is contingent on an individual’s motivation and ability to engage thoughtfully with a message is crucial when considering the e ects of political messages embedded within decentralized networked ecosystems that are also fragmented and micro- segmented. Writing in 1995, Negroponte introduced the concepts of “push” and “pull” media to capture the distinction between old broadcast media that were “pushed” out to people versus digital media that people can select and “pull” toward themselves. “Being digital will change the nature of mass media from a process of pushing bits at people to one of allowing people (or their computers) to pull at them,” he wrote(1995, p. 84). Building on the concepts of “push” and “pull” media, Holbert et al. (2010) explicate the implications of digital “pull” media for cognitive processing. They propose that the logics of old mass “push” media were likely to be met by audience members comparatively lower in motivation to process media content (because there was less opportunity for content selection), hence resulting in a greater likelihood of peripheral and heuristic message processing. But with fragmentation, choice, and interpersonally contextualized media experiences, users are guided by a motivation and ability to “pull” relevant, interesting, socially bene cial content toward them. Holbert et al. (2010) argue that this means media content is likely to be engaged with thoughtfully, or centrally, hence resulting in attitudes that are resistant and longer-lasting, or—if scrutiny results in counterargumentaion—less attitude change and more message resistance. These dynamics also complicate the historical distinction between media exposure and media attention (Cha ee & Schleuder, 1986; Semetko et al., 1992), as the decision to select and “pull” content toward oneself in a high choice media environment might itself be indicative of interest in and attention to that content. p. 574 Since political messaging and news are now embedded within social networks, interpersonal dynamics ought to shape motivation and ability to thoughtfully engage with information content as well. In the context of public opinion toward an unfamiliar issue (nanotechnologies), Ho, Scheufele, and Corley (2013) found that people are mainly guided by heuristic cues when processing news, yet interpersonal communication changes these dynamics. The link between cognitive elaboration and public opinion toward nanotechnologies was contingent on interpersonal communication on the topic. In an examination of how people process and learn from information obtained through news versus through social contacts, Carlson (2019) found people learned more through news—except when the social contact was like-minded and Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 knowledgeable, in which case individuals learned as much through the social contact as they did through news. Looking at the e ects of online comments on news story processing and engagement, message relevance (related to processing motivation) matters. User comments focused on a speci c dimension (regionalism) of a news story shape audience recall and persuasion, especially for individuals for whom that dimension (regionalism) is especially salient (e.g., those highest in regional self-identity; Lee et al., 2017). Users’ perceptions of the credibility of news stories are also a ected by the valence of user comments, especially for users to whom the content is relevant (Winter et al., 2015). Yet, while social media users consider their friends more credible than news organizations in general, they still rate speci c news articles as more credible when shared by a news organization than by a friend (Tandoc, 2019). When it comes to political expression and information sharing decisions in online contexts, experimental work indicates that motivation to thoughtfully process a message might be overshadowed by reputational and social motivations (Winter, 2020). Winter (2020) found that users who anticipated writing on social media about a given news article were more likely to be motivated by impression management than by the quality of the arguments in the article itself. Whereas people express themselves politically online for social and persuasion-related reasons (see Eveland et al., 2011), the e ects of political expression on cognitive elaboration and subsequent political participation is contingent on which motivations are guiding them, with persuasion motivations a bigger driver of elaboration and participation (compared to social motivations), mediated by news-seeking behaviors (Yoo et al., 2017). Finally, the notion that exposure to “pull” media might re ect increased motivation to engage in cognitive elaboration and argument scrutiny may itself be complicated by factors related to the physical technologies themselves. Many users, especially younger users, engage in “second-screen” or “dual-screen” behaviors (see Gil de Zuniga & Liu, 2017), typically involving the use of a smartphone or tablet while (or immediately following) consuming other media, such as television or streaming content (see Nee & Barker, 2020). Such dual-screen activities during news or political debate viewing can increase cognitive load and reduce news processing, recall, and comprehension (Van Cauwenberge et al., 2014). However, research also points to emotional and social bene ts of second-screen behaviors, including building feelings of community and learning the opinions of others (Nee & Barker, 2020), as well as political bene ts, including seeking additional information, sharing information, and expanding knowledge in ways that fuel further political engagement (Chadwick et al., 2017; Vaccari et al., 2015). Increased reliance on smartphones, tablets, and computers for news (see Shearer, 2021) has spurred p. 575 scholars to study the impact of screen size and mobile news features on cognitive processing, attention, and learning, with disappointing results. Dunaway et al. (2018) demonstrate that although shifts in news norms and production routines for mobile devices might increase access to news, they may also erode regular in-depth engagement with news stories. Their eye-tracking studies illustrate that even when users perceive that they are paying attention to news content on mobile devices, that attention is lower than it is on computers. Additional lab studies reveal that video news, when consumed on mobile devices, elicits lower psychophysiological responses than when consumed on larger screens, hence indicating less cognitive engagement with—and arousal to—news programming on mobile (Dunaway & Soroka, 2021).  8.2. Agenda-Setting Revisited. Again. And Again and Again The decentralized control of political media, networked logics of digital spaces, and reduced gatekeeping authority of traditional news organizations all change the context within which cognitive theories of media e ects occur. Take agenda-setting, for instance. Although we still see evidence of legacy news outlets shaping the agendas of the public (Tran, 2013), agenda-setting e ects are far smaller among those who rely on alternative and online information sources (Shehata & Stromback, 2013). Feezell (2018) found that people had their agendas shaped by “incidentally encountered” stories on social media, with the strongest Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 agenda-setting e ects found among those lowest in political interest. Meanwhile, the story-selection decisions of journalists and editors are themselves in uenced by online and alternative content as well, as news organizations include stories and events that have circulated through social media and niche outlets (Jacobson, 2013). Gilardi et al. (2022) found that the news’ agenda was set by the social media agenda of politicians and political organizations, and their agendas were, in turn, shaped by the news and by each other. This “intermedia agenda-setting” research (McCombs, 2004), exploring who is setting whose agenda, illustrates that although traditional news is still a dominant setter of the public’s agenda (Djerf-Pierre & Shehata, 2017), Twitter—especially media-related Twitter—helps to set the agenda of mainstream news (Harder et al., 2017; Valenzuela et al., 2017), especially in the context of breaking news stories (Su & Borah, 2019). Additionally, partisan news outlets are driving the agendas of some traditional legacy news outlets (Vargo & Guo, 2017); by covering stories that serve their ideological perspective, they indirectly pressure mainstream outlets to cover those stories as well. These complex reciprocal agenda-setting dynamics have signi cant consequences, especially if the online content shaping the agenda of a news organization is false, something that is particularly likely for partisan news organizations (Guo & Vargo, 2020; Vargo et al., 2018). Networked and intermedia dynamics permit bottom-up agenda-setting processes that may empower issue publics and activists (Luo, 2014; Su & Hu, 2020). With the public’s agenda constructed through these complex relationships, the nature of Downs (1972) “issue attention cycle” has changed. The “alarmed- discovery” phase that used to catapult the public into action is no longer driven by the media’s agenda alone, but also through interpersonal discussion and mobilization through social media (Zhang et al., 2017). Issue publics mobilized through “hashtag activism” help shape media and policy agendas, framing issues from racial justice (Bonilla & Rosa, 2015; Yang, 2016) to gendered terrorism attacks (Carter Olson, 2016) to sexual harassment (Xiong et al., 2019). p. 576 In the largely homogeneous, mass-audience-oriented media system of the 1970s, articulating a distinction between signaling to people what issues to pay attention to versus persuading people to think a certain way made sense. But by 2005, as that distinction was becoming obsolete, McCombs circled back again. “The media not only can be successful in telling us what to think about,” he wrote, “they also can be successful in telling us how to think about it” (McCombs, 2004, p. 546). So-called attribute agenda-setting or second- level agenda-setting is similar to framing in that it involves elements of a news story being made more salient in ways that in uence how audiences think about the issue (Entman, 1993). With fragmented partisan media outlets reporting on stories in ways that complement their networks’ ideological leanings, attribute agenda-setting (Muddiman et al., 2014) and framing have become an essential part of contemporary media e ects research. Recent studies illustrate how partisan news frames shape viewer perceptions of issues ranging from immigration (Gil de Zuniga et al., 2012) to climate change (Feldman et al. 2012), to COVID vaccines (Jamieson & Albarracin, 2020; but see also Trujillo & Motta, 2021 for data from 144 countries indicating that mere internet access itself is related to vaccine hesitancy) to election integrity and voter ID laws (Wilson & Brewer, 2013).  8.3. Political Polarization The new media landscape is both interpersonal and identity-reinforcing by design. As discussed above, media and political elites and non-elites alike have incentives to create emotion-laden content (whether it be information, misinformation, or disinformation) that taps into the intended audience’s increasingly aligned (Mason 2018) social and political identities. And audiences have a greater ability than ever before to selectively expose themselves to content that meets their informational, entertainment, identity- reinforcing, and/or attitude-reinforcing needs. This perfect storm of content production and content Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 selection is what Slater (2007, 2015) calls a reinforcing media spiral—social identity drives media behaviors in ways that reinforce those identities. The “spiral” aspect of these relationships stems from the recursive dynamics between exposure and polarization that numerous scholars have highlighted (Prior, 2013; Stroud, 2010). Such studies point both to media contributing to polarization, and to polarization-fueling exposure to like-minded media. Given the central role that identity plays in political polarization (Iyengar et al., 2012), recent research has explored the impact of political communication in this new media environment on both attitude and a ective polarization (see Tucker et al., 2018, for a comprehensive review of the role that misinformation plays in this process). With regard to attitude polarization, Song and Boomgaarten (2017) use an agent-based modeling approach to demonstrate the role of interpersonal (dis)agreement in the reinforcing media spiral process, concluding that, “reciprocal dynamics between attitude polarization and attitudinally congruent partisan media exposure over time critically hinge on the attitudinal composition of one’s discussion network and contextual variations in which such media exposure occurs” (p. 18). Feldman et al. (2014) use a two-wave panel study to test the reinforcing spirals model in the context of attitudes about global warming, nding evidence for an “ongoing, reinforcing cycle in which media use in uences beliefs, and these beliefs then a ect subsequent media use, which, in turn, reinforces beliefs” (p. 603). And Lee et al. (2014) nd that the p. 577 positive relationship between social network homogeneity and attitude polarization is strongest among people who engage in political discussions more often than those who do not. These polarization e ects are also moderated by the variation in a ordances of di erent social media platforms (e.g., Facebook, Twitter, and WhatsApp; Yarachi et al., 2020). In her book, Niche News, Stroud (2011) demonstrates how partisan selective exposure, exacerbated by the proliferation of partisan media outlets, can simultaneously increase (not necessarily equitable) political participation and contribute to attitude polarization and fragmented interests. Pointing to the potential deleterious e ects of this process for democracy, Stroud (2011, p. 10) argues that: “the onslaught of diverse partisan media outlets may undercut the development of common goals…Without a shared issue agenda, allocation of limited resources, such as time and money, become more di cult.” Arceneaux and Johnson (2013) argue that some of the polarizing e ects demonstrated in this literature are a methodological artifact—experiments that force exposure to pro- or anti-attitudinal content. In a series of innovative experiments, Arceneaux and Johnson (2013) demonstrate that the polarizing e ects of exposure to partisan media sources are limited by audience characteristics and the available content choices (i.e., whether people can opt out of news/political content altogether). Prior’s (2013) work indicates that attitude polarization from selective media exposure is limited to those individuals who are highly politically interested and engaged, a subpopulation that is arguably more likely to wield political in uence. Recent research also points to the role of digital algorithms and curated information ows in exacerbating political divisions (Cho et al. 2020). In contrast, Munger and Phillips’ (2022) analysis of right-wing political content on YouTube challenges the notion that algorithms themselves cause polarization and/or radicalization. They propose a supply-and-demand framework for analyzing the impact of political content on YouTube. Rather than the supply of right-wing YouTube content turning moderates into radicals (in a top-down, “hypodermic” way), the authors argue for the central role of content demand in this process: “the true threat posed by some right-wing content on YouTube is the capacity for creators to draw  communities of committed viewers that mutually create and reinforce radical political canons, including some that promote hatred. There is a cap on how much news media a person can consume in a given day; YouTube has dramatically increased the number of distinct political communities which are able to hit that content cap” (p. 3). Although more research is needed to pin down the precise nature of the causal impact of algorithms, it is clear that dominance of algorithms and curated ows of information in the new media environment necessitate methodological advances that can distinguish the e ects of selective exposure (demand), on the one hand, and selected (e.g., by algorithms) exposure (supply), on the other (Thorson et al., 2019; Thorson & Wells, 2016). Downloaded from https://academic.oup.com/edited-volume/51639/chapter/418866339 by Universiteit Leiden - LUMC user on 22 October 2024 While attitude polarization concerns the shifting of policy positions toward partisan extremes, a ective polarization refers to the rise in negative attitudes toward the opposing political party. In the latter context, emerging evidence highlights the role of media exposure, with subtle and contingent e ects of media as a function of audience and content characteristics. For example, Druckman et al. (2019) nd that the relationship between exposure to partisan incivility and positive a ect toward, and trust in, the party of the source of the incivility is moderated by whether the content is pro- or counter-attitudinal; such exposure is depolarizing for in-party members, but polarizing for out-party members Work by Hasell and Weeks (2016) illustrates the implications of a ective polarization beyond mere attitudes, nding that exposure to pro- p. 578 attitudinal content increases anger toward the opposing party, which, in turn, increases the likelihood of sharing the content. Lelkes, Sood