Polarisation and misinformation
Media choice, echo chambers, filter bubbles, and other characteristics of the communication environment are often blamed for political polarisation and a post-truth era (Sunstein 2017; Weeks 2018). It is reasonable; expanding media choice and rising polarisation incentivise media selectivity and fact avoidance while increasing vulnerability to misinformation. The potential consequences are troubling amidst rising populism, in which elites utilise mass media to employ polarising communication strategies divorced from fact (Waisbord 2018). Yet existing evidence emphasises the centrality of predispositions in motivating information seeking and processing, cognitive biases, and the relative unimportance of facts in belief formation (Kunda 1990; Lodge and Taber 2013; Swire et al. 2017a). These perspectives suggest political misperceptions are a product of cognitive and affective processes that operate independently from or in tandem with exposure to misinformation (Thorson et al. 2018).
A review of the literature suggests that rather than remaining fixated on how the information environment facilitates misinformation’s dissemination and exposure, we should more carefully consider whether and how cognitive biases, motivated reasoning, and affective polarisation foster permissiveness towards misinformation, as well as its processing, acceptance, and endorsement. The central argument in this chapter is that despite the media ecology's potential for heightening misinformation exposure (Allcott and Gentzkow 2017), cognitive biases, affective polarisation, and sorting are as much to blame for susceptibility to misinformation and the development of misperceptions as structural changes to the media.
There is ample evidence for enduring, widely held misperceptions in the US (Kuklinski et al. 2000; Lazer et al. 2018; Jerit and Zhao 2020). Misperceptions persist across several domains, including policy (Jerit and Barabas 2012), the evaluation of politicians (Nyhan and Reifler 2010; Miller et al. 2016), and even parties and partisans (Ahler and Sood 2018). Misperceptions plague citizens of both parties when supporting of their political predispositions (Jerit and Barabas 2012). Because misperceptions are often characterised as being on the rise and correlated with changes to media and rising polarisation (Sunstein 2017; Weeks 2018), both are blamed for the spread and effects of misinformation. Recent Russian interference in the 2016 US presidential election and rising populism exacerbate these concerns, making it easy to understand references to an ‘epidemic of misinformation’ and a surge in related research.
Yet neither political misinformation nor misperceptions are new (Waisbord 2018). Misinformation has been around as long as politics, and researchers have been explaining why Americans cannot articulate coherent policy preferences since the 1950s (Converse 1964; Zaller and Feldman 1992; Deli Carpini and Keeter 1996). Nevertheless, both are often characterised as on the rise, enabled by digital media, and capable of affecting citizen beliefs and judgments (Weeks
Most explanations for misinformation spread and misperceptions are based on changing media structures or cognitive and affective processes. Media explanations focus on two structural aspects of the digital media environment: those allowing for ideological segregation through media selectivity and algorithmic filtering and those facilitating massive and rapid exposure to misinformation.
High choice, selectivity, and filtering
Media-based accounts depict exposure to misinformation as both a cause and a consequence of polarisation, where polarisation facilitates selective exposure in a high-choice context, and the misinformation is persuasive because it comes from like-minded sources. Assuming that much misinformation comes from biased sources of partisan information, high choice encourages exposure by allowing for partisan media selectivity and through algorithmic filtering and homogeneous social networks (Sunstein 2017). Under this view, echo chambers and filter bubbles are harmful contexts for misinformation because attitude-consistent information is viewed as credible and is more likely to be accepted and shared, especially within networks (Bakshy et al. 2015).
But evidence for ideological segregation is mixed (e.g. Weeks et al. 2017), as is evidence about its effects (Flaxman et al. 2016; Peterson et al. 2018). Some studies link digital media to ideological segregation (Sunstein 2017); others find that high choice and networks facilitate exposure diversity (e.g. Gil de Zuniga et al. 2017). Cross-ideology' exposure is frequent (Bakshy et al. 2015), attributable to the ‘social’ nature of cross-ideology network ties and endorsements (Messing and Westwood 2014). Additionally, media audiences are highly concentrated on neutral mainstream sites (Hindman 2018; Flaxman et al. 2016). When echo chambers are found, there is little evidence of attitudinal or behavioural effects (Peterson et al. 2018). Despite evidence of echo chambers, they are not impenetrable; their potential to intensify exposure to misinformation and misperceptions may be overstated, especially amidst polarisation.
Rapid information sharing and network structures
A defining feature of the digital media environment is its ability to concentrate attention even while fragmented (Hindman 2018). Thus, some concerns about misinformation are related to digital media’s vast and instantaneous reach. Low reproduction costs, network structures, and peer-to-peer sharing capabilities facilitate rapid information dissemination and information cascades, allowing the opportunity for intense exposure to misinformation (Vosoughi et al. 2018). The digital media environment lacks the gatekeeping infrastructure of traditional media, making it amenable to misinformation spread (Rojecki and Meraz 2016). Elites have an advantage for spreading rumours online and spurring informational cascades (Bakshy et al. 2011) and are known to enlist digital media to spread persuasive and divisive misinformation to encourage polarisation and mistrust (Engesser et al. 2017; Ernst et al. 2017; de Vreese et al. 2018).
Yet concerns about the persuasive power of mass misinformation exposure are reflective of now-debunked theories of massive media impact from the earliest phases of media effects research, which was motivated by fears about elites’ use of mass media to disseminate powerful and persuasive propaganda (Iyengar 2017). Given what we have since learned about active and selective audiences (Prior 2007; Bennett and Iyengar 2008; Arceneaux and Johnson 2013), the power of predispositions (Kunda 1990; Zaller 1992), and competing demands on the attention of modern audiences (Hindman 2018), we should not necessarily expect persuasion from misinformation except under narrow conditions. Though polarised audiences may be exposed to more misinformation through media choice and personalisation, its potential to significantly change attitudes and behavior is unknown.
However, digital and social media may compound the effects of misinformation in ways other than through persuasion. One is by increasing familiarity. Repeated information exposure increases perceived accuracy even when false (Swire et al. 2017b; Pennycook et al. 2018). When misinformation is disseminated widely and continually circulates within networks, it increases familiarity (Weaver et al. 2007), cognitive accessibility (e.g. DiFonzo et al. 2016), ease of processing (e.g. Schwarz et al. 2016), and perceptions of consensus (e.g. Leviston et al. 2013), all of which influence judgments about accuracy. Thus, even allowing for the limited circumstances under which misinformation should persuade (Zaller 1992; Bennett and Iyengar 2008), the information environments amenability to disseminating and recirculating misinformation prompts concerns that repeated exposure will increase its perceived accuracy.
The effects of even highly intense doses of misinformation should, however, be dependent on the message, political context, and individual-level characteristics like political awareness and predispositions (Zaller 1992; Bennett and Iyengar 2008). These factors and high polarisation may help explain why, despite the proliferation of misinformation on social media, there is little direct evidence that it promotes misperceptions (e.g. Allcott and Gentzkow 2017; Garrett