Social Media and the Political Engagement of Young Adults: Between Mobilization and Distraction

Purpose: Scholars have expressed great hopes that social media use can foster the democratic engagement of young adults. However, this research has largely ignored non-political, entertainment-oriented uses of social media. In this essay, I theorize that social media use can significantly dampen political engagement because, by and large, young adults use social media primarily for non-political purposes, which distracts rather than mobilizes. Design/methodology/approach: I illustrate this argument using aggregate level data from the U.S., Germany, Switzerland, and Japan by comparing relative voter turnout and social media use data of young adults. Findings: Data suggest a so called Social Media Political Participation paradox in those countries: The gap in voter turnout between young adults and older generations has not significantly decreased, despite a skyrocketing rise of social media use on the side of young adults, and the overwhelming research evidence that social media use fosters offline political participation. Implications: When trying to understand the implications of social media for democracy across the globe, entertainment-oriented content needs to be brought back in. Originality/value: This essay challenges the dominant research paradigm on social media use and political participation. It urges future research to theoretically develop, describe, and empirically test a comprehensive model of how social media use has the potential to mobilize and to distract.


Introduction
Around the globe, social media have become a centerpiece in young adults' lives. Particularly with their smartphones, young adults can literally be on social media 24/7, permanently connected to the world and their peers (Vorderer and Kohring 2013). In fact, when comparing the current young generation to their older counterparts, there is a fundamental difference in media use behaviors: While young adults, aged 16-25, rely on digital platforms or messenger services, such as Facebook, TikTok, YouTube, Twitter, Instagram, WhatsApp, or WeChat, to get the news, the older generation is much more likely to be exposed to traditional news sources such as television or newspapers. At the same time, there are dozens of studies around the globe demonstrating that, traditionally, young adults are less interested in traditional politics compared to older generations (Delli Carpini 2017), less likely to vote, and generally less politically sophisticated (Binder et al. 2021). In short, political parties had been, before the emergence of digital media, struggling to reach out to the younger generation. Especially when it comes to traditional institutions, you adults are often described as detached and apathetic (e.g., Binder et al. 2021;Loader et al. 2014).
Yet with social media, scholars have expressed great hopes regarding young adults' democratic engagement (see Binder et al. 2021;Oser and Boulianne 2020): It has been argued that particularly social media can build new relationships between political actors and young adults, enable social interaction about political topics, connect people, enhance political opinion expression, equalize engagement and generally foster participation as well as boost voter turnout or contribute to social cohesion (e.g., Boulianne 2011Boulianne , 2015Boulianne , 2020Goh et al. 2019;Loader et al. 2014). So, with digital media, there are grounds to believe that the generational engagement gap may be reduced, and that young citizens could be reengaged into the political world. In fact, scholars working on digital media and political engagement have been fascinated by this idea, largely pointing to democratically welcomed outcomes of social media use, such as learning or participation. For instance, researchers observed a positive relationship between the frequency of social media use and protest participation among the youth (Valenzuala et al. 2014), and more generally, it has been found that political social media use is positively related to various forms of political participation (e.g., Ekström et al. 2014;. With recent meta-analyses on the topic, the evidence for the democratically positive outcomes of social media use is simply overwhelming, particularly in cross-sectional survey research (Boulianne 2011;Chae et al. 2019; and also with respect to young adults (Boulianne and Theocharis 2020).
However, scholarship on the democratic outcomes of social media frequently seem to overlook the fact that social media are primarily used for entertainment and relational purposes, especially when it comes to young adults (Dimitrova and Matthes 2018; but see Theocharis and Quintelier 2016). That is, the social media use of young people is clearly dominated by non-political content (Binder et al. 2021). Yet the vast majority of studies do not take these forms of exposure into account, eventually ignoring a large share of the diversity in content on social media. As a consequence, scholars have turned a blind eye on potentially distractive effects of social media use on political engagement, leading to a skewed overall picture of this research field. In this conceptual paper, I take a different approach by theorizing that social media use can significantly dampen political engagement. The main reason is that social media are primarily used for entertainment and social networking purposes, which has the potential to distract rather than mobilize (Heiss and Matthes 2021).
In what follows, I will briefly summarize the literature on digital media and young adults' political engagement. Then, using illustrative, aggregate level data from the U.S., Germany, Switzerland, and Japan, I will then describe the so called Social Media Political Participation Paradox, that is, the gap in voter turnout between young adults and older generations has not significantly decreased, despite a skyrocketing rise of social media use, on the side of young adults, and the huge amount of studies suggesting a significant relation between social media use and political participation. Then, I will develop a theoretical argument for why social media use is also likely to disengage, challenging the prevailing academic reading of the literature as well as the existing empirical evidence.

Digital Media and Young Adults' Political Engagement
Political engagement, often also referred to as political participation, is understood as "actions or activities by ordinary citizens that in some way are directed toward influencing political outcomes in society" (Ekman and Amnå 2012, p. 287). This entails a diverse repertoire ranging from traditional (e.g., voting) and nontraditional (e.g., political online discussion; see Hopmann et al. 2015) forms of engagement to political consumerism . Political engagement can be conceptualized along the lines of individual and collective actions (Adler and Goggin 2005). In addition, formal political participation, such as voting or party membership, and activism, which allows for influence in the political decision making process though protests represent distinct dimensions of political engagement (Ekman and Amnå 2012). Social media has given rise to entirely new forms of action and interaction that can only happen in a digitally networked space, such as online petitions or commenting on politicians' posts (Sloam 2014). Social networks allow for noninstitutionalized and horizontal modes of engagement, which are often favored by young adults (Sloam 2014). Since such forms of political engagement only exist and can be exercised within social media, they also have to be treated as a separate subdimension of political engagement. Overall, the various forms of engagement can be described as either institutionalized (e.g., voting) or noninstitutionalized (e.g., protest behavior). This distinction is relevant because "young adults are disproportionately more likely to participate through noninstitutionalized means." (Weiss 2020, p. 3), particularly in the online world. This is by no means a new insight. Since decades, scholars have bemoaned a participation gap between younger and older individuals (Quintelier 2007), as "in almost every election young people are the least likely to vote, and these participation rates are continuously declining" (Quintelier 2007, p. 165). For instance, in Austria, young people are allowed to vote at the age of 16, but nevertheless, their turnout rates are comparably low (Binder et al. 2021). Findings from other countries also suggest that young people have comparatively negative attitudes toward politics and low trust in the political system (Quintelier 2007;see Binder et al. 2021).
But there is also hope. In this research area, "[y]outh's digital media use is often seen as a partial remedy to the decline of youth participation in political and civic life" (Boulianne and Theocharis 2020, p. 112). The argument is that digital media is an important source of information for those not primarily interested in institutional politics. More specifically, it is generally assumed that social media fosters young adults' political engagement because the networking function of social media helps young citizens to develop skills and psychological dispositions that promote offline participation (Kahne and Boyer 2018). Also, a seminal, crosssectional study by Gil de Zuniga et al. (2012) suggests that social media indirectly promote participation by fostering opinion expression, which is a key driver of political participation. As another indirect effect of social media on political participation, it has been argued that some forms of social media use, such as news sharing, can strengthen social relationships and increase social cohesion, which in turn, may be the driver for collective action and group engagement (e.g., Goh et al. 2019;Hwang and Kim 2015). Similarly, social media shapes civic attitudes, and such "civic identity is a central factor that fosters civic engagement because it helps individuals to see society as a construction of human actors with political and moral goals" (Chen et al. 2015, p. 445).
In addition, it has been theorized that social media fosters incidental exposure to political information, which leads to learning effects, and ultimately, an increase in traditional forms of political participation (Matthes et al. 2020). Furthermore, as explained by Boukes (2019), social media increase the opportunities for citizens to learn because "never before has so much information, mostly without extra financial costs, been so close at hand for citizens" (p. 39). In line with this argument, the timeline structure of social networking platforms can lead to a "by-product" learning mechanism through which knowledge gaps are reduced (Chadwick 2012), ultimately leading to political participation. And, more germane to young adults, the Impressionable Years Hypothesis (Sears and Levy 2003) suggests that young people are more susceptible to social media effects compared to adults, as political efficacy and engagement are not fully consolidated during adolescence (e.g., Eckstein et al. 2012). That is, when exposed to social media, the notion of efficacy can be strengthened in young adults, leading to participation.
So overall, the dominant argument is that social media use can foster softthat is, non-institutionalized-forms of engagement in a first step, such as online political expression or low-effort forms of political participation. This, in turn, will then shape classical, institutionalized types of political engagement in a second step. As Bode et al. (2014) have put it, "When adolescents use social networking sitessomething many of them do in their daily livesin a new, politically oriented way, it may actually increase their likelihood of participating in politics in other ways." (pp. 424-425). In addition, social media can also directly lead to a rise in offline participation among young adults (Boulianne and Theocharis 2020). This theory, in line with the available meta-analyses, leads to the clear prediction that social media use should increase not only non-institutionalized forms of participation, but also-indirectly or directly-traditional forms such as voting. The available "findings offer a strong, conclusive statement that online and offline forms of engagement are highly correlated; youth engage in both environments" (Boulianne and Theocharis 2020, p. 123). But how is this observable at the aggregate level when it comes to the institutionalized forms of participation, such as voting in national parliamentary elections?

The Generational Political Participation Gap
To reiterate, the previously available literature unequivocally suggests that social media use, particularly political forms of use, can foster online and offline political engagement (e.g., Boulianne and Theocharis 2020). With these insights at hand, the traditional generational gap between the young and older generations in terms of institutionalized participation needs to be revisited. One could argue that, over the past decades, social media use of young adults, unlike older generations, increased from basically zero to very high levels. That is, assuming there is a significant small to medium effect size for the relationship between political social media use on online participation, and a medium to large effect size for the relationship between online and offline participation (Boulianne and Theocharis 2020), we would expect that, in the past decades, the generational gap should have decreased.

The Social Media Political Participation Paradox
To test this idea, I draw on official voter data (i.e., not self-report survey data) from the National Parliamentary Elections in Germany. As can be seen in Figure 1, as can be expected, social media use of young individuals has increased from zero in the early 2000s to more than 80% at the time of the last election in 2017. So, in terms of the meta-analyses reported above and although such aggregate level analyses do not allow causal conclusions, one of the theorized drivers of engagement has witnessed a skyrocketing increase. But paradoxically, as visible in Figure 1, there were no significant increases in voter turnout for individuals aged 18-24 over the years. By contrast, in 1983, young individuals showed a turnout of 84.30 (18-20 year old) and 81.50 (21-24 year old)% in the national election. In 2017, these numbers were significantly and substantially lower (69.90% and 67.00% respectively). That is, if anything changed, the numbers are more suggestive of an Source for social media data: Koch and Frees (2007), source for election data: Bundeswahlleiter (2017).
Political Engagement of Young Adults increase in the generational gap, rather than a decrease. Although these are only aggregate level data, they do not suggest that social media use-emerging in the last two decades-has completely changed the picture.
Data from the United States suggest a similar, albeit not identical picture. As can be seen in Figure 2, the generational gap in voter turnout is clearly visible over time, although young individuals slightly increased in turnout (e.g., from 44.30% in 2008 to 48% in 2020), whereas turnout for older generations remained largely at the same level. However, a fundamental shift in turnout when it comes to young adults is clearly not visible in the data, the generational gap is still evident. In Switzerland ( So overall, these aggregate data deliver a clear message for these four countries, none of which has witnessed huge protest movements outside traditional party politics. Young adults are significantly less likely to vote in national elections compared to older generations, and it seems this gap has not decreased over the past decades, as one would expect from the enthusiastic voices in research on social media use and political participation. In fact, the findings appear somewhat paradox: The social media use has increased from zero to almost 100% in the last two decades, which should, considering the effect sizes observed in previous research, also become visible at the aggregate level. So, on the one hand, we learn from previous research that social media significantly fosters political participation, online and offline (e.g., Boulianne and Theocharis 2020), on the other hand, we don't see a higher turnout compared to older generations, who are much less likely to use social media compared to their younger counterparts. And even if they did, we would expect an increase in turnout for this group as well, which has not been the case. This phenomenon can be called the Social Media Political Participation Paradox. In short, at the aggregate level, it appears as if social media has not changed a thing, although social science research suggests quite the opposite.

Explaining the Paradox
Of course, on the methodological the most obvious explanation is ecological fallacy. We can't derive any assumptions on individual level effects when looking at aggregate level data. Ultimately, no causal claims can be made, nor can we say anything about intraindividual change or processes. There may also be simultaneous, competing processes, and third variables cannot be controlled at all. Yet what we learn is that there is a generational gap in participation at the aggregate level which is, by and large, rather substantial. So at best, the data reported above can be understood as anecdotal evidence. Yet on theoretical grounds, the findings could inspire us to elaborate on why social media use may not automatically lead to strong shifts in institutionalized forms of political participation.
Several reasons can be found for this in the literature. On the theoretical side, the recent Social Media Political Participation model (Knoll et al. 2020) holds that social media can foster participation only when a chain of subsequent conditions are met. According to this model, young adults have to be (intentionally or incidentally) exposed to political content on social media, they need to appraise political content on social media as relevant (as compared to other content that is simultaneously present), there needs to be a discrepancy between a current state and a future, desired state, they need to believe that a participation goal is attainable, and this goal must then be dominant in a real behavioral situation, in which other behavioral goals may be present as well (see Knoll et al. 2020). If one of the conditions is not met, social media use may not increase offline forms of political participation, according to the model. This model can explain why participation efforts of young adults are often short-lived, they can rise to substantial amounts during times of protest, but remain low in times of national elections. It would suffice if one of the appraisals is negative, as for instance, when other goals appear to be more important in an behavioral situation. Obviously, typical self-report survey studies cannot fully grasp the process described in the model because the behavioral situation is hardly taken into account in the typical research designs and the processes can hardly be measured in retrospect.
Second, it has been argued that social media is more likely to impact noninstitutionalized forms of participation, rather than institutionalized ones (Sloam 2014). That is, social media may engage young people politically, but that doesn't necessarily make them more likely to participate in elections. In other words, social media can have the potential to engage, but such engagement may be short-lived, conditional, and bound to specific topics such as environmental engagement, animal rights, or social protest.
Third, and more importantly, when looking at the democratically relevant effects of social media use, scholars rarely consider (or control) non-political, entertainment-oriented content (Boulianne and Theocharis 2020). Non-political, entertainment-oriented content can be information on lifestyles, products, leisure, sports, or social relationships on social media (see Hanitzsch and Vos 2018). Yet when we estimate the effects of political social media use without at least controlling non-political forms, we may end up with erroneous conclusions about the mobilizing potential of social media. Political contents and entertainmentoriented contents are simultaneously present on social media. A typical newsfeed completely mixes both.
As expressed by Boulianne and Theocharis (2020), "purely social-, entertainment-, and leisure-oriented activities carried out on digital media do not necessarily mobilize individuals for civic or political action." (p. 114) Entertainment-oriented use of social media may serve the purpose to create an "emotional relief generated by temporarily recreating or recessing from daily routines" (Buzeta et al. 2020, p. 81). When compared to political social media use, entertainment-oriented use accounts for a large proportion of everyday Internet use, especially among adolescents (Büchi et al. 2016). For instance, in a recent over time experiment using behavioral browsing data, Wojcieszak et al. (2021) found that news websites comprised only 3.54% of the browsing behavior. The authors conclude that "news content is nearly unnoticeable in the context of overall information and communication ecology of most individuals" (p. 8).
Despite the relevance of non-political content on social media, scholars have hardly looked at the relationship of entertainment-oriented use of social media and political participation . As one rare exception, Kahne and Bowyer (2018), observed that non-political content can serve as a gateway to various forms of civic and political participation. Yet other studies found no such gateway effects. In a cross-sectional study conducted in the Netherlands, Bakker and de Vreese (2011) observed non-political social media use had negative consequences for political participation. Additionally, data collected by Chan et al. (2012) suggest that the effect of Weibo use on the willingness to express one's political views was weakened when entertainment motives prevailed. An experiment by Theocharis and Lowe (2016) even suggests negative consequences of Facebook use for participation because it can distract from politics. So taken together, only a few studies looked into the effects of non-political social media use on participation, some suggest that entertainment activities on social media may serve as a gateway to participation (Kahne and Bowyer 2018), others speak of distraction from politics (Bakker and de Vreese 2011;Boukes 2019;Chan et al. 2012;Theocharis and Lowe 2016).
Besides adding entertainment-oriented exposure as a predictor, it is also important to understand how entertainment-oriented and political uses of social media interact. If we accept both types of uses as separate dimensions, we arrive at four different types of audience members: "the Avoiders" (political use: low and entertainment-oriented use: high), "the Inactive" (political use: low and entertainment-oriented use: low), "the Distracted" (political use: high and entertainment-oriented use: high), and "the Focused" (political use: high and entertainment-oriented use: low). This typology of ideal groups is depicted in Figure 3 (see Matthes et al. 2021).
For the purposes of the present paper, the theoretically most relevant group are "the Distracted". I use this term because I theorize that high loads of entertainment content may potentially distract the processing of political content. There are several theoretical reasons for that. As suggested by the priming literature (Higgins 1996), the accessibility of concepts can drive cognitions and behaviors. Thus, when young adults evaluate the importance of concepts, they do not use all of the information they have available in memory. Instead, they often rely highly accessible information (Iyengar and Kinder 1987). Here, accessibility refers to the ease or speed with which available information can be retrieved from memory. Information that is recently and frequency activated, is more likely to be temporary accessible (Arendt and Matthes 2014). At the same time, of course, individuals' cognitions are also driven by their basic political and social orientations which are be chronically accessible. But the influence of chronically accessible concepts can be weakened when individuals are frequently and recently exposed to other considerations. When, for instance, people are exposed to entertainment-oriented content, these non-political considerations become cognitively accessible. That is, when young adults are permanently confronted with non-political content, as for instance, about friends, family, movies, or sports, this content gets situated at the top of the head, making it more cognitively accessible when making judgments about political matters, including engagement.
That means, non-political content on social media competes with (and may even impede) the accessibility of political content. The Social Media Political Participation model (Knoll et al. 2020) holds that political content on social media can only affect offline political participation when the content is appraised as relevant. However, when young adults are primarily exposed to non-political content, this may shift the perceived relevance of political considerations. In other words, when there is an abundance of entertaining, non-political news, the current political issues may appear less severe, and therefore, also less personally relevant. As a consequence, young adults may engage less with political content, which is a prerequisite to political participation. By the same token, the model holds that the goal to vote needs to be dominant in a real behavioral situation. When other goals are present as well, such as visiting friends or going shopping, then young adults will only cast their vote when this goal is dominant, that is, more important than other goals. So again, the argument is that non-political content on social media can shift the appraisal of goals.
What is more, we know from entertainment research that particularly hedonic entertainment content has a high absorption potential, and based on that, it can interfere with the cognitive elaboration of political content, "because it absorbs attentional resources to a degree that interferes with further elaboration" (Bartsch and Schneider 2014, p. 376). Also, on a perceptual level, non-political content is often perceived as more eye-catching as compared to political content. The reason is that the former typically relates to close ties such as friends and family and it is often associated with immediate positive gratifications.
Moreover, not all individuals may be exposed to political and non-political content on social media simultaneously. "The Avoiders", albeit high in general social media use, may try to bypass exposure to political content, as for instance when an "individual exposes him-or herself to a limited amount of news because other content has more appeal to him or her, many algorithms will make future content decisions in favor of other content, e.g., entertainment, and give news stories a lower priority" (Skovsgaard and Andersen 2020, p. 466). Young adults interested in non-political contents may curate their newsfeed on SNS platforms in ways which exposes them to high degrees of entertainment-oriented content with low likelihood of exposure to political contents. This may lead to low-effort, feelgood types of engagement at best, making high-effort, offline political participation less likely.
But still, but even for news avoiders, scholars have argued that social media has the potential to foster participation because people can be accidentally exposed to political information in their newsfeeds. So even though young adults may not want to see political information, they may see it based on their social environments and networks (Matthes et al. 2020). This exposure to political information may then lead to learning, and ultimately, participation. While scholars agree on the importance of incidental exposure for participatory outcomes, they have, at the same time, entirely ignored the opposite logic: Social media also exposed individuals accidentally to non-political information. Such incidental exposure to non-political content "can have important effects on political outcomes such as learning or participation. The more people are confronted with non-political content (without actively looking for it), the more they are potentially distracted from their primary political processing goal" (Matthes et al. 2020(Matthes et al. , pp. 1137(Matthes et al. -1038. So as much as incidental exposure to political content can engage, so can incidental exposure to non-political content distract and disengage. Unfortunately, while there are dozens of studies on the former phenomenon, we lack studies on the latter.

Testing the Paradox
The paradox described here is situated at the aggregate level. However, additional evidence is needed to explain why this paradox occurs. For this, it is therefore important to look at the outcomes of entertainment-oriented content at the individual level. That is, we should not only model political media use as a predictor of participatory outcomes, but simultaneously access (or at least control) exposure to non-political content. These two dimensions can then not only be used as focal predictors, they may also interact in a regression model (see Figure 3). High levels of entertainment-oriented non-political content on social media may have a "vampire" effect, when young adults are simultaneously exposed to political content on social media as well. Some young adults, those with high levels of political interest and sophistication, may primarily be exposed to political content on social media, which in fact, can be theorized to increase levels of online and offline political participation (Knoll et al. 2020). These conjectures suggest that social media use, as a generic category, is of limited use in global research on digital media.
A similar argument can be made for the measurement of political participation. Oftentimes, scholars create participation indices by averaging several, distinct acts, such as wearing buttons of a party, sharing personal political experiences on social media, signing a petition, or voting in an election (see for a discussion, Theocharis and van Deth 2018). These measures blur the differences between institutionalized and non-institutionalized forms of participation. Differentiating between these forms beyond the classic online versus offline notion, however, is crucial to understand the political consequences of social media use.
In addition to that, it is important to note that the typical survey approach used in political communication scholarship has its limitations as well. The reason is that self-report measures of social media use are largely biased (Scharkow 2019). Also, typical experiments use forced-exposure settings, which cannot be compared to news reception situations in the real world (Wojcieszak et al. 2021) Thus, to empirically examine the paradox, future studies need to rely on more naturalistic settings, such as web-tracking data or mobile experience sampling. With such measures, we can more precisely estimate the amount of exposure to political and non-political content.

Conclusion
With the emergence of social media, great hopes have been expressed across the globe that young adults may reengage into traditional politics, directly due to the various activities happening on social media, and indirectly by shaping low-effort forms of online-engagement in the first place, which is then assumed to shape offline participation in a second step. In fact, the available evidence clearly suggests that social media use, particularly political one, has an impact on offline forms of participation (Boulianne and Theocharis 2020). However, aggregate level data from Germany, the U.S., Switzerland, and Japan indicate that generational gaps between younger and older adults have not or only marginally decreased. Although there was an unprecedented up rise of social media use over the past decades (including an increase in political uses), voter turnout did not change significantly, and was substantially lower as compared to the older generation. At the same time, the entire body of research on social media and political participation would suggest an increase in participatory activities of young adults. In explaining this paradox, this paper pointed to the potentially distracting functions of social media, mainly due to non-political, entertainment-oriented content. Such content can make non-political information more accessible, ultimately impeding the processing and salience of political considerations, dampening the activation and execution of participatory goals. In order to test these conjectures, future research needs to carefully distinguish several types of content on social media, on several platforms and channels, and access motivations, gratifications of usage as well as contents. This may lead to a more nuanced picture about the social media based political engagement of young adults, particularly when it comes to the democratically most relevant outcome: voting in an election.