More Harm Than Good? Online Media Use and Political Disaffection Among College Students in the 2008 Election

Authors


  • Accepted by previous editor Maria Bakardjieva

Abstract

This study examines the ways in which online media influenced political disaffection among young adults during the 2008 presidential election campaign. The effects of social media attention, online expression, and traditional Internet sources on political cynicism, apathy, and skepticism were evaluated using data from an online survey of college students. Results show that attention to social media for campaign information is positively related to cynicism and apathy and negatively related to skepticism. Online expression has a positive effect on skepticism. Implications are discussed for the role of social media in bringing a historically disengaged demographic into the political process.

Citing indicators of declines in political participation, voter turnout, attention to public affairs media, and trust in fellow citizens and government, many scholars have bemoaned the rise of a disengaged and alienated citizenry (Bennett & Rademacher, 1997; Crotty & Jacobson, 1980; Delli Carpini, 2000). Concerns over an epidemic of disengagement have focused mainly on young adults, who with each passing generation appear to become less interested in public affairs, less knowledgeable, and less likely to vote than previous cohorts of young adults (Bennet & Rademacher, 1997; Delli Carpini, 2000). Indeed, young adult voter turnout declined between the 1972 and 2000 elections, except for a surge in 1992 (Lopez, Kirby, Sagoff, & Herbst, 2005). Such continued nonparticipation among certain portions of the population prevents those groups from having a voice in their democracy (Crotty & Jacobson, 1980), and sustained disaffection among the public may undermine the efficacy of the democratic system (Erber & Lau, 1990).

The recent introduction of so-called social media tools such as Facebook, YouTube, and Twitter has raised new questions about the democratic participation of young adults. Some observers have credited social media with spurring the involvement of young adults in the 2008 election by facilitating the exchange and creation of political content online (Hesseldahl, MacMillan, & Kharif, 2008; Smith, 2009). For example, young adults used social media to share campaign news with others, create political content (e.g., photos, videos, podcasts), express support for a candidate, and discuss the election (Smith, 2009). It is also often argued that the Obama campaign successfully used social media to energize and mobilize supporters, particularly young adult voters (Sullivan, 2008). In the 2010 midterm election, roughly 21 percent of online adults used social media to obtain political information or get involved in the campaign, with young adults aged 18–29 leading the political use of social media (Smith, 2011).

Not surprisingly, a growing body of research attempts to evaluate the potentially democratizing capacity of social media. Previous studies have examined the effects of social media use on political participation and decision-making processes leading to participation such as external political efficacy, interest, and knowledge (Baumgartner & Morris, 2010; Groshek & Dimitrova, 2010; Kushin & Yamamoto, 2010), while limited research has assessed the impact of social media on political disaffection, an attitudinal orientation toward politics that includes cynicism, apathy, and skepticism. Research on social media and political disaffection to date has focused mainly on cynicism and examination of social media use is limited to cognitive dimensions or frequency (Hanson, Haridakis, Cunningham, Sharma, & Ponder, 2010; Towner & Dulio, 2011), measures that may not capture the varying dimensions of social media use such as content creation and opinion expression.

To address these limitations in the literature on social media and political disaffection, this study assesses the impact of cognitive and behavioral aspects of social media use, along with the use of traditional online sources, on young adults' political cynicism, apathy, and skepticism. To this end, this study analyzes data from an online survey of college students during the 2008 presidential election. Findings from this study will further clarify the role of social media in politics, helping demystify whether social media help bring a historically disengaged demographic into the political process or contribute to its disengagement.

Political Disaffection: Cynicism, Apathy, and Skepticism

A number of important constructs have been identified in the study of political disaffection. In general terms, political disaffection refers to negative dispositions towards politics (Austin & Pinkleton, 1995). Despite its connotations, political disaffection is not limited to harmful, damaging effects on political participation. As explained below, disaffection has both negative and positive qualities that can hamper or motivate participation in the political process (e.g., Austin & Pinkleton, 1995, 1999; Pinkleton & Austin, 2001; Pinkleton, Austin, Zhou, Willoughby, & Reiser, 2012). The present study, derived from the existing literature, examines three major dimensions of political disaffection: cynicism, apathy, and skepticism.

Political cynicism is defined as a mistrustful disposition towards, and an absence of confidence in, the political system (Austin & Pinkleton, 1995, 1999). The cynical citizen tends to view politics as personally irrelevant, and abstains generally from the process of politics (Austin & Pinkleton, 1995, 1999; Crotty & Jacobson, 1980). Some scholars argue that cynicism breeds disinterest in public affairs and political disengagement (Cappella & Jamieson, 1997; Craig, Niemi, & Silver, 1990; Crotty & Jacobson, 1980; Erber & Lau, 1990). Further, cynicism is negatively associated with external political efficacy, beliefs about the responsiveness of government to the will of the public (Hanson et al., 2010; Pinkleton et al., 2012).

Some fear that cynical citizens are less likely to purposively use news media to learn about public affairs (Crotty & Jacobson, 1980). Indeed, past studies have shown that cynicism is negatively related to active media use and situational political involvement, a state of motivated interest in politics that stimulates information seeking (Austin & Pinkleton, 1999). Findings suggest that because cynical citizens do not consume news media, they become less informed and increasingly apathetic in a downward spiral of disaffection (Austin & Pinkleton, 1999).

Apathy is another key component of political disaffection. It can be defined as indifference towards, lack of interest in, or lack of attention to politics (Bennett, 1986). Scholars have used it both as an “indicator for potential for participation or an estimate of psychological engagement in the governmental process” (Bennett, 1986, p. 39). Following the spiral of disaffection thesis, apathy and cynicism may feed each other, with the cynical citizen becoming more apathetic and thus more cynical leading towards disengagement and disinterest (Austin & Pinkleton, 1995). Not surprisingly then, cynicism and apathy are positively associated with each other (Pinkleton & Austin, 2004), and political apathy is less likely for citizens with increased levels of external political efficacy and situational political involvement (Pinkleton & Austin, 2004). This may explain why apathy is negatively related to skepticism, a construct discussed below, as the skeptic is inclined toward further information gathering.

Not all so-called negative dispositions towards the media and politics have deleterious effects on citizens. Research has shown that some negative sentiments towards politics increase intent to participate in politics (Austin & Pinkleton, 1995). Skepticism, defined as a disbelief in the political process but not a rejection of it (Cappella & Jamieson, 1997), is thought to be good for democracy. Skepticism is characterized by inquiry towards fact checking information received and is thus considered a positive characteristic of political disaffection (Austin & Pinkleton, 1999; Pinkleton & Austin, 2001). In contrast, cynicism reflects disengagement and distrust in politics. Indeed, skepticism has been shown to be distinct from cynicism (Pinkleton et al., 2012). Skepticism motivates information seeking because the individual sees media coverage of an issue as incomplete (Austin & Pinkleton, 1999). Persons who are skeptical and whose information needs are not fulfilled by media coverage tend to seek additional sources of information. Accordingly, both skepticism and perceiving media coverage as incomplete are associated with situational political involvement (Austin & Pinkleton, 1999).

The Many Faces of the Internet

Use of the Internet for political purposes has become an important part of the political environment over recent election cycles, particularly among young adults (Smith, 2009). Traditional Internet information sources such as online editions of news organizations, candidate websites, and portal websites have become a regular repertoire of political information sources for many citizens. During the 2008 presidential election and the 2010 midterm campaign, citizens increasingly attended to social media for political information such as getting candidate and campaign information, posting and exchanging political messages, and getting involved in a campaign (Smith, 2011). As such, scholarly interest in the political utility of the Internet and social media has grown tremendously in recent years.

In studying the role of the Internet in politics and public affairs, scholars have begun parsing different aspects of online media, finding that different online media relate to different user outcomes at cognitive and behavioral levels (Shah, McLeod, & Lee, 2009). Furthermore, scholars have broken down Internet use into information-seeking and interpersonal interaction (Katz, Rice, & Aspden, 2001; Shah, Cho, Eveland & Kwak, 2005).

In recent years, a new era of online communication technology termed social media has emerged. Scholars have differentiated this new online form from traditional Internet sources (Baumgartner & Morris, 2010; Correa, Hinsley, & Gil de Zúñiga, 2010; Towner & Dulio, 2011). Such sites are said to contain elements of Web 2.0, emphasizing social interaction, rich media content exchange, and programming architecture that enable support of third-party applications (Cormode & Krishnamurthy, 2008). Such sites foster communal interaction around user-created content, in which the user becomes both content creator and consumer (Bruns, 2006).

The field lacks definitional consensus for social media. However, communication scholars have used the term social media and Web 2.0 interchangeably and have generally accepted social network sites, video sharing sites, blogging, instant messaging, and Twitter to fit within the rubric (Correa et al., 2010; Gil de Zúñiga, Puig, & Rojas, 2009; Groshek & Dimitrova, 2010; Hanson et al., 2010). In a political context, Kushin and Yamamoto (2010) measured these interactive media platforms at the information-seeking (i.e., attention to social media for campaign information) and expression dimensions (i.e., online expression about the campaign).

The Internet and Political Disaffection

A growing body of evidence suggests a positive relationship between political Internet use and political cognitions and behaviors, including internal and external political efficacy, knowledge, political participation, and voting (Gil de Zúñiga, Puig, & Rojas, 2009; Johnson & Kaye, 2003; Wang, 2007). There is, however, limited research on the relationship between Internet use and political disaffection.

The Internet has been championed as a key location for active political information seeking because it allows users to obtain more updated and detailed information (Delli Carpini, 2000; Zhang & Pinkleton, 2009). Active information searchers have been shown to have a lower level of disaffection, likely due to increased interest and more satisfying and fulfilling information search results (Pinkleton & Austin, 2001). Also, a greater diversity of information sources allows users to encounter and acquire various ideas and perspectives that might not receive enough attention in offline platforms. Thus, using traditional Internet sources may help direct users to the political process and lower cynicism and apathy. Indeed, prior studies have shown that using traditional Internet sources for public affairs content is related to external political efficacy and situational political involvement, negative predictors of cynicism and apathy (Kushin & Yamamoto, 2010; Zhang & Pinkleton, 2009). Also, although traditional online sources and social media were not differentiated in their measure of Internet use, Zhang and Pinkleton (2009) found a negative relationship between Internet use and cynicism.

Skepticism, on the other hand, leads to additional information searching and is thus positively related to active media use resulting from increased involvement with politics (Austin & Pinkleton, 1999). Given the ability of the Internet to foster information gathering, Internet users will be more likely to engage in active information searches to further confirm or disconfirm news content received. Zhang and Pinkleton (2009) found a positive relationship between their measure of Internet use and skepticism.

Considering the above research, the following hypotheses are proposed:

H1: Attention to traditional Internet sources for campaign information will be (a) negatively associated with cynicism; (b) negatively associated with apathy; and (c) positively associated with skepticism.

Political Disaffection, Social Media, and Online Expression

A growing number of studies have investigated the effects of social media use on political outcomes. Extant literature yields mixed results with regard to the political utility of social media. Some studies have reported that political use of social media enhances political participation (Gil de Zúñiga et al., 2009; Park, Kee, & Valenzuela, 2009; Utz, 2009). These studies favor the democratic potential of social media given their capacity to lower barriers to information, encourage online interaction, expression, and discussion, and presumably enable all citizens to participate in the democratic process (Castells, 2007). Thus, social media are seen as places of interactive equality where all users can and may participate as unrestricted equals (Gil de Zúñiga et al., 2009).

While such interaction is theoretically possible, recent evidence questions the actual democratic nature of social media. Rather than facilitate the participation of all, social media may foster social fragmentation, whereby users retreat into sites to gather content that supports their viewpoints and engage in partisan expression with like-minded individuals (Baumgartner & Morris, 2010; Towner & Dulio, 2011; Woolley, Limperos, & Oliver, 2010). For example, studying an array of social media formats, Groshek and Dimitrova (2010) reported that political social media use was not related to political knowledge, campaign interest, and voting likelihood. Also, Baumgartner and Morris (2010) found that social network site use was unrelated to offline political participation. Towner and Dulio (2011) found that exposure to YouTube's political channel had no effect on candidate evaluations and external political efficacy.

Why is it that social media may be unrelated to positive political outcomes, and potentially related to political disaffection? One possible explanation offered by Baumgartner and Morris (2010) is that any news on social network sites is “soft news,” or entertainment programming with ancillary political content. This may apply to social media in general. With soft news, political learning is an unintended outcome of consumption, and this side effect may bring otherwise disengaged citizens into the political process (Baum, 2003). However, because social media users rely on social networks for information, disengaged young adults may not be in an online cohort of individuals sharing political content, which renders exposure to political content limited (Baumgartner & Morris, 2010).

While incidental exposure to political content may break through, nonpurposive information gathering may have its own drawbacks. While purposive media use predicts low levels of disaffection (Pinkleton & Austin, 2001), those who are cynical are less likely to seek political information online (Zhang & Pinkleton, 2009). Thus, exposure to incidental political content among an otherwise disengaged cohort who is not seeking political content may relate to political disaffection. In effect, those who are less politically engaged may be the people who report relying on social media for political information because it takes very little effort to get political information, and the information they get tends to be incidental.

In contrast, it is less likely that online expression would be associated with political disaffection. Online expression behavior is a complex phenomenon, predicted by feelings of ability and motivations for self-satisfaction and extrinsic awards (Correa, 2010). Individuals choose to create content to enhance self-awareness and express their identity (Leung, 2009). Those engaging in online communities tend to gravitate toward their ideological camps (Baumgartner & Morris, 2010; Smith, 2009; Woolley et al., 2010), which further solidifies the political self within its place among the likeminded. Online expression may also enhance political dispositions, as it has been shown to lead to political trust (Wang, 2007) – evidence suggesting a possible negative relationship with cynicism and apathy.

Thus, it is less likely that we should see persons engaging in political expression apt to political disaffection, unless the individual finds the experience unrewarding or discouraging. Online interaction has been repeatedly shown to contain elements of flaming, vitriolic language that may discourage participation among a wide audience and harm the potential for deliberative democracy (Davis, 1999). As gratifications received from online content generation is key to determining its impact on psychological empowerment (Leung, 2009), engaging in polarized online environments, or being the recipient of criticism could have negative effects on self-image as well as impeding actualization of intrinsic and extrinsic awards. In the political sphere, this could lead to disaffection.

A few studies have assessed the relationship between social media and political disaffection. Exploring social network sites, video sharing sites, and blogs, Hanson et al. (2010) reported that frequency of social network site use was associated with a lower level of cynicism, while the effects of frequency of video sharing site and blog use was not significant. Examining exposure to YouTube's politics channel, Towner and Dulio (2011) reported increased levels of cynicism as compared with traditional online news. While providing useful insights into the effects of social media on cynicism, these studies are limited primarily to cognitive aspects of social media use or simple frequency measures. As noted above, social media are characterized by the principles of Web 2.0 that embrace the participation of users, and thus it is important to examine online expression, behavioral aspects of social media such as posting political comments and sharing news stories. Past studies also have focused mainly on cynicism as an indicator of political disaffection. Other dimensions of disaffection such as apathy and skepticism have not been fully explored.

Considering the existing literature, the following research questions are proposed:

RQ1: In what ways will attention to social media for campaign information be associated with (a) cynicism; (b) apathy; and (c) skepticism?

RQ2: In what ways will online expression about the campaign be associated with (a) cynicism; (b) apathy; and (c) skepticism?

Method

An online survey of college students was conducted during the 2 weeks before the November 4, 2008 election at a large public university in the northwestern United States. College students are a reasonable population for the present study, because young adults including college students are heavy users of social media. Although online surveys are limited in coverage for a survey of the general population (Dillman, Smyth, & Christian, 2009), college students are known to be among the most Internet connected groups of the population.

A probability-based sample of the student body was obtained from the university registrar. An initial invitation e-mail was delivered to the university e-mail address of each individual in the sample. An e-mail reminder was sent during the week before the election. The survey completion rate was 10.85%, which is admittedly unfavorable. Two respondents were below 18 years old and, hence, ineligible to participate in the survey. Those who were above the age of 29 were not included in the present analysis because the focus of this study was on young adults. Twenty-five respondents were removed for this reason, with the resulting sample size of 407.

The measures of attention to traditional Internet sources and social media, and online expression were conceptually derived and consistent with prior studies (Kushin & Yamamoto, 2010). Attention to social media and traditional Internet sources represent a cognitive aspect of online political activity, while online expression represents its behavioral aspect (Katz et al., 2001; Shah et al., 2005; Wang, 2007). Social media and traditional Internet sources were conceptually distinct along the content creator line as noted above. Attention measures were used to assess the cognitive dimension, because they are more adequate than exposure measures in capturing a recipient's cognitive engagement with an information source (Chaffee & Schleuder, 1986).

Controls

To evaluate the unique effects of traditional Internet sources, social media, and online expression, the present study included age, sex, and political ideology as controls. Also, attention to traditional news sources was controlled. Using a 7-point Likert-type scale with no attention and a lot of attention as anchors, respondents were asked how much attention they paid to newspapers, television, radio, and magazine for campaign information.

Attention to social media

Attention to social media for campaign information was measured by five items. Using a 7-point Likert-type scale with no attention and a lot of attention as anchors, respondents were asked “For information about the election, how much attention have you been paying to each of the following?”: (1) “Personal blogs,” (2) “Video-sharing websites (Youtube),” (3) “Microblogs,” (4) “Social networking websites (e.g., Facebook or Myspace),” and (5) “Online forums and discussion boards.” These items were summed to form an additive index (M = 10.74, SD = 6.09, α = .78).

Online expression

Online expression about the campaign was measured by four items.1 Using a 7-point Likert-type scale with none and a lot as anchors, respondents were asked “In regard to the election, how much have you engaged in each of the following?”: (1) “Writing blog posts on political issues,” (2) “Creating and posting online audio, video, animation, photos or computer artwork to express political views,” (3) “Sharing political news, video clips, or others' blog posts online,” and (4) “Exchanging opinions about politics via e-mail, social networking websites, microblogging (such as Twitter) or instant messenger.” These items were combined to create an additive index (M = 7.79, SD = 5.03, α = .75).

Attention to traditional Internet sources

Attention to traditional Internet sources for campaign information was measured by five items. Using a 7-point Likert-type scale with no attention and a lot of attention as anchors, respondents were asked “For information about the election, how much attention have you been paying to each of the following?”: (1) “Government websites (e.g., local, state, or national),” (2) “Candidate's websites,” (3) “Network TV news websites (e.g., CNN.com, ABCnews.com, or MSNBC.com),” (4) Print media news websites (e.g., New York Times or US News and World Report Websites),' and (5) “News portal services (e.g., Google News or Yahoo News).” These items were summed into an additive index (M = 19.70, SD = 6.53, α = .66).

Cynicism

Based on previous research (Austin & Pinkleton, 1995; Craig et al., 1990; Pinkleton & Austin, 2001, 2004; Pinkleton et al., 2012), cynicism was measured by five items using a 7-point Likert-type scale with strongly disagree and strongly agree as anchors: (1) “It seems like politicians only care about special interests,” (2) “Politicians lose touch quickly with the public after they get elected,” (3) “Politicians put their own interests ahead of the public interest,” (4) “It seems like our government is run by a few big interests who are just looking out for themselves,” and (5) “Candidates for office are interested only in people's vote, not in their opinions.” These items were summed into an additive index (M = 23.19, SD = 5.42, α = .80).2

Apathy

Based on questions used in prior studies (Austin & Pinkleton, 1995; Pinkleton & Austin, 2004; Pinkleton et al., 2012), apathy was measured by four items using a 7-point Likert-type scale with strongly disagree and strongly agree as anchors: (1) “Voting takes too much time,” (2) “Participating in elections is more trouble than it's worth,” (3) “Staying informed about the government is too much trouble,” and (4) “Keeping up on political issues takes too much time.” These items were combined to form an additive index (M = 8.18, SD = 4.61, α = .80). It is important to point out that although these items have been shown to measure the same underlying construct (Pinkleton & Austin, 2004), there might be a discrepancy between the conceptual and operational definitions of apathy. That is, it is possible that the items used to measure apathy capture a different theoretical concept such as political laziness rather than a lack of interest in or attention to politics. Therefore, further examination of this construct is required in future studies.

Skepticism

Similar to cynicism and apathy, the items designed to measure skepticism were derived from the political disaffection literature (Austin & Pinkleton, 1999; Pinkleton et al., 2012; Zhang & Pinkleton, 2009). Specifically, based on Pinkleton et al. (2012), skepticism was measured by five items using a 7-point Likert-type scale with strongly disagree and strongly agree as anchors: (1) “I think about the things elected officials say before I accept them as believable,” (2) “It's important to critically evaluate what news stories say,” (3) “I think about news stories before I accept them as believable,” (4) “I always think twice about statements made in news stories,” and (5) “It's important to critically evaluate statements made by government officials.” These items were combined into an additive index (M = 30.77, SD = 4.21, α = .80).3 It is importunate to note that although these items were directly obtained from the political disaffection literature cited above, their wording may have produced measures of skepticism that were biased toward higher scores. That is, it is possible that respondents did not provide low ratings to these items due to the way the statements were phrased. As such, this skepticism index may have failed to adequately measure the intended construct. Thus, future studies should consider constructing new measures of skepticism with high content and construct validity, which would provide a more accurate examination of skepticism.

A preliminary analysis was performed for online media and political disaffection variables. Cynicism positively correlated with skepticism (r = .17, p < .01) and apathy (r = .25, p < .001). Skepticism negatively correlated with apathy (r = −.25, p < .001). Attention to social media positively correlated with online expression (r = .51, p < .001) and attention to traditional Internet sources (r = .28, p < .001). Online expression and attention to traditional Internet sources positively correlated with each other (r = .37, p < .001). Overall, the correlations support construct and discriminant validity of the measures of online media use and political disaffection. To test the hypotheses and answer the research questions, three ordinary least squares (OLS) regression models were constructed. Demographic variables and attention to traditional news sources were entered first, followed by the three online media variables.

Results

Table 1 presents regression models for cynicism, apathy, and skepticism. The first column indicates that age (β = .13) and sex (β = −.13) were significant predictors of political cynicism. Political ideology and traditional news sources were not related to cynicism. These variables combined to explain 4.8% of the variation.

Table 1. OLS Regression Predicting Political Cynicism, Apathy, and Skepticism
 CynicismApathySkepticism
 βββ
  1. Note. ***p < .001, **p < .01, *p < .05, †p < .10.

Age.13*−.04.13*
Sex (female)−.13*−.02−.09
Political ideology.05.04-.04
Newspaper attention.03-.13*.08
TV news attention.00−.03−.11
Radio attention.02.06−.01
Magazine attention.02.06−.16**
R2 (%)4.8*5.2**7.1***
Attention to social media.11.13*−.12*
Online expression.07−.09.14*
Attention to traditional Internet sources−.27***−.23***.10
Block R2 (%)5.5***4.9***2.6*
Total R2 (%)10.3***10.1***9.7***

H1a predicted that attention to traditional Internet sources for campaign information would be negatively associated with cynicism. The data support this hypothesis. Attention to traditional Internet sources for campaign information had a relatively strong negative effect on cynicism (β = −.27), while controlling for the influence of three demographic controls, attention to traditional news sources, attention to social media, and online expression. Respondents who paid greater attention to traditional Internet sources showed a lower level of cynicism.

RQ2a and RQ3a asked how attention to social media for campaign information and online expression would be associated with cynicism. Although statistically marginal, social media attention for campaign information had a positive effect on cynicism (β = .11, p = .073), which indicates that those who attended more to campaign information on social media were more politically cynical. Online expression was not significantly associated with cynicism. The measures of these three online media use variables explained an additional 5.5% of the variation.

The second column displays an OLS regression model predicting political apathy. Of demographics and traditional news sources, only newspaper attention significantly predicted apathy (β = −.13). Those who paid greater attention to newspapers for campaign information showed a lower level of apathy. These variables combined to explain 5.2% of the variance.

H1b predicted that attention to traditional Internet sources for campaign information would be negatively associated with political apathy. This hypothesis was supported. Attention to traditional Internet sources had a negative effect on apathy (β = −.23), with those who attended more to traditional Internet sources for campaign information showing a lower level of apathy.

RQ2b and RQ3b asked how attention to social media for campaign information and online expression would be associated with political apathy. Social media attention had positive effects on apathy (β = .13), with those who paid greater attention to campaign information on social media more politically apathetic. Online expression was not significantly related to apathy. These online media use variables as a whole explain an additional 4.9% of the variance.

The third column of table 1 presents an OLS regression model predicting political skepticism. Age (β = .13) was positively related to skepticism, with older respondents being more politically skeptical. Magazine attention also significantly predicted skepticism (β = −.16). Respondents who attended more to magazines for campaign information exhibited a lower level of skepticism. Demographics and traditional news sources explained 7.1% of the variance.

H1c predicted that attention to traditional Internet sources for campaign information would be positively associated with political skepticism. This hypothesis was not supported, as traditional Internet sources were not significantly related to skepticism. Attention to social media for campaign information had an independent negative effect on skepticism (β = −.12). Respondents who paid more attention to the campaign on social media were less politically skeptical. The effects of online expression were positive (β = .14), with respondents who expressed their political views using social media exhibiting a higher level of skepticism. These findings answer RQ2c and RQ3c. Overall, the three online media use variables accounted for an additional 2.6% of the variance.

Discussion

This study examined the relationship between social media attention, online expression, and traditional Internet attention, and political disaffection among a sample of college students during the 2008 presidential election campaign. In particular, we were interested in investigating whether social media and online expression would be related to political disaffection among young adults, a historically disengaged demographic group.

Traditional Internet Sources

The literature indicates that political Internet use has positive effects on political outcomes including knowledge, internal and external efficacy, and participation (e.g., Gil de Zúñiga, Puig, & Rojas, 2009; Johnson & Kaye, 2003; Wang, 2007). In support of the democratic potential of the Internet, attention to traditional Internet sources for campaign information was found to be negatively associated with cynicism and apathy. These results are consistent with past research showing negative effects of Internet use on cynicism (Zhang & Pinkleton, 2009). These results are also in concert with prior studies that have shown positive relationships between traditional Internet attention and external political efficacy and situational involvement (Kushin & Yamamoto, 2010), negative predictors of cynicism and apathy (Austin & Pinkleton, 1995; Pinkleton & Austin, 2004; Pinkleton et al., 2012).

Interestingly, traditional Internet attention was not significantly related to skepticism. This is inconsistent with prior studies that have found positive effects of political Internet use on skepticism and active information seeking (Kushin & Yamamoto, 2010; Zhang & Pinkleton, 2009). Traditional online information sources were posited to foster skepticism by providing more detailed information to those wanting to verify the first level of political content received. It might be that political information provided by online editions of traditional news media and government and candidate websites are deemed complete, and thus, readers do not perceive a need to verify facts and/or seek out additional sources for further details. The mixed cumulative evidence requires further examining the effect of traditional Internet sources on skepticism.

Social Media

The present findings provide some support for studies questioning the democratic potential of social media (Baumgartner & Morris, 2010; Groshek & Dimitrova, 2010; Towner & Dulio, 2011). Attention to social media for campaign information had a negative effect on skepticism, while its effect on apathy was positive. Although statistically marginal, attention to social media was positively related to cynicism. These results imply that political information on social media has negative political implications. To understand the observed effects of attention to social media on negative aspects of political disaffection, it is important to study the characteristics of social media platforms and the patterns of political content on social media.

Unlike traditional journalism committed to fair and balanced coverage, fact-checking, and quality of writing, political content on social media can rely on personal views and opinions. Previous studies have revealed that social media such as campaign blogs often include negative attack messages such as attacking an opponent's record and personal qualities (Trammell, 2006). Also, the messages exchanged on social media often contain uncivil elements that discourage participation (Davis, 1999). Reading such content might breed cynical and apathetic attitudes toward politics.

Regarding skepticism, social media might decrease the motivation for fact-checking. Social media deliver content to users in a stream of information from “friends” and those “followed,” via RSS feeds, and other push features. Hence, social media use requires very little active information searching. Also, the information shared through online social networks such as friends and family members may have stronger persuasive influence on users (Rogers, 1995). As individuals trust the information circulated through their personal networks, this may lessen the need for information verification. The social media user thus has the luxury of laziness, expending very little effort to learn about politics in correspondence to their low levels of interest, as suggested by heightened levels of apathy found in this study.

Social media also may foster social fragmentation, prompting users to build walled gardens to retreat into websites where selectivity is the governing ethos of content exposure (Baumgartner & Morris, 2010; Woolley et al., 2010; Towner & Dulio, 2011).4 Such selective exposure may reduce the likelihood of incidental exposure to political content that counters one's point of view (Baumgartner & Morris, 2010). This may help further explain the observed relationships between social media attention and apathy and cynicism.

Online Expression

The present results indicate a more positive outlook for online expression. Online expression was not associated with cynicism and apathy, while its effect on skepticism was positive. These results supported prior research, which showed a positive association between online expression and situational political involvement (Kushin & Yamamoto, 2010). A possible explanation for these findings is that as users express themselves and talk about politics with others, they are more likely to make sure that what they hear, know, and post online is accurate. Indeed, interpersonal sources have been cited as potential channels for greater information gathering by skeptics (Austin & Pinkleton, 1999).

Online expression, then, may be a pathway to political participation distinct from the more passive function of information acquisition through social media. These results suggest that online expression, a central function of social media, can help bring young adults into the political sphere. This finding bodes well for the democratic potential of the Internet and online social communities, supporting past research in this area (Hardy & Scheufele, 2005; Kushin & Yamamoto, 2010; Wang, 2007).

Limitations

Several limitations are of note. First, the present findings are based on a sample of college students at one institution and are limited in generalizability. Previous research on the political utility of online media among young adults has similarly suffered from this limitation (e.g., Baumgartner & Morris, 2009; Correa, 2010; Kushin & Yamamoto, 2010; Towner & Dulio, 2011). College students are on the younger end of the young adult population, and more females attend college than males (Aud, Fox, & KwealRamani, 2010). Furthermore, college students differ politically from others in the young adult population. For example, college students are more civically and politically involved, more likely to vote, and more likely to engage in political discussion online than young adults without any college education (Portney & O'Leary, 2007). Also, among 18- to 24-year-old young adults, those with more advanced educational attainment (e.g., a bachelor's degree or more) are more likely to vote than those with a lower level of educational attainment (File & Crissey, 2010). Therefore, although college students are heavy users of social media (Smith, Rainie, & Zickuhr, 2011), they might not use the Internet for political purposes in the same manner as non-college students. As such, caution is needed in interpreting the present findings. Future studies should span beyond the college campus to include a wider range of young adults. Second, the online survey suffered from a low response rate, which might have produced inaccurate estimates. Third, the results should be weighed in consideration of the relatively limited use of social media for politics during the 2008 election. As political use of social media grows, and further development of social media is expected, continued research on the political utility of social media is required. Fourth, as discussed in the method section, future research should verify the present findings with improved measures of skepticism and apathy. Finally, there are other aspects of social media for political use not examined in the present study such as user satisfaction with and perceived credibility of the nature of political information on social media and online expression (Towner & Dulio, 2011).

Conclusion

These limitations notwithstanding, this study has presented important insight into the role of traditional Internet sources, social media, and online expression in political disaffection. The relationship between social media and political disaffection is complicated by the ways in which they are used, particularly along the boundary of information consumption and content creation.

For some, social media may breed political disaffection, as users tend to be limited to incidental exposure to political information. This study suggests that consuming political information on social media increases cynicism and apathy, negative aspects of political disaffection. Despite popular assumptions that social media are inherently good for politics, this study adds to a growing body of evidence to the contrary. However, online expression was found to be positively associated with skepticism. Social media not only allow incidental consumption of political information but also enable exploration and development of political identity through interaction and expression with others about politics. Online expression is a complex phenomenon involving motivations, reward seeking, and personal identity (Correa, 2010; Leung, 2009). As such, greater understanding is needed regarding what encourages the individual to bridge the gap from passive reception of political information on social media to active expression of political viewpoints. The challenge for those hoping social media can bring young adults into the political sphere, and help reverse the political disaffection plaguing young adults, is to activate individuals on social media toward greater engagement through online expression.

Notes

  1. 1

    The items used in this study were shown to load on the same underlying component in previous research (Kushin & Yamamoto, 2010). The items consist of content creation and information sharing and exchange through popular avenues of interactivity online, capturing a wide array of expressive behaviors that can be performed in social media spaces such as Twitter, blogging, and social network sites. The fourth item measuring the extent to which respondents exchange opinions about politics includes email and instant messenger, as many social media sites such as Facebook and YouTube integrate dynamic e-mail and instant messenger applications into their communication functions. These applications allow users to communicate with each other without leaving the site they are presently using or activating separate applications and enhance the capacity of users to exchange political views with others. Indeed, Correa et al. (2010) acknowledged the benefits of instant messenger in facilitating user communication and conceived it as part of social media.

  2. 2

    Of note, several of the items used in the literature to measure cynicism are similar to several used previously to measure external political efficacy. For example, two measures of four used to measure external political efficacy in the 1988 National Election Study are (1) “People like me don't have any say about what the government does” and (2) “I don't think public officials care much what people like me think” (Niemi, Craig, and Mattei, 1991, p. 1408). For both questions, a low score on an interval measure indicates greater external political efficacy.

  3. 3

    With the high mean value of the skepticism index, it is possible that a single or a small number of observations influenced the estimate of the regression line. To examine the extent to which individual cases disproportionately influenced the outcome of the analysis, studentized deleted residuals were requested. Four cases exceeded the cut-off value of +/− 3.00 (Cohen, Cohen, West, & Aiken, 2003). The regression model for skepticism was re-estimated with the four cases removed. Results were substantively equivalent to the analysis reported in this article.

  4. 4

    The term walled garden is a popular term in online sites. It refers to closed systems. Web 2.0 systems are said to be open and collaborative. In their popular book about the rise and success of Web 2.0 over traditional “Web 1.0” sites, Tapscott and Williams (2006) commented that: “The losers built walled gardens. The winners built public squares” (p. 39). In an ironic twist, social media may be enabling users to create their own walled gardens through selective exposure.

Biographies

  • Masahiro Yamamoto (e-mail: myamamoto@uwlax.edu), Ph.D., is an Assistant Professor in the Communication Studies Department at the University of Wisconsin-La Crosse. His research interests include mass media and community social control and social media and politics.

    Address: 1725 State Street, 202 Wing, La Crosse, WI 54601

  • Matthew J. Kushin (e-mail: mkushin@shepherd.edu), Ph.D., is an Assistant Professor in the Department of Communication at Shepherd University. His research interests include political campaigns, civic and political engagement, online media, and social media.

    Address: Knutti Hall L10K, PO Box 5000, Shepherdstown, WV 25443