Poking People to Participate: Facebook and Political Participation in the 2008 Election

Transcription

1 Facebook and Political Participation 1 Running head: FACEBOOK AND POLITICAL PARTICIPATION Poking People to Participate: Facebook and Political Participation in the 2008 Election Jessica Vitak, Paul Zube, Andrew Smock, Caleb Carr, Nicole Ellison, and Cliff Lampe Michigan State University To cite: Vitak, J., Zube, P., Smock, A., Carr, C., Ellison, N., & Lampe, C. (2009, May). Poking people to participate: Facebook and political participation in the 2008 election. Paper presented at the annual conference of the International Communication Association, Chicago.

2 Facebook and Political Participation 2 Abstract In the 2008 presidential election, social network sites such as Facebook allowed users to share their political beliefs, support specific candidates, and interact with others on political issues. But do political activities on Facebook affect political participation among young voters, a group traditionally perceived as apathetic in regard to civic engagement? Results from a survey of undergraduate students (N=683) at a large public university in the Midwestern United States, conducted in the month prior to the election, found that political activity on Facebook is a significant predictor of political participation. Students perceptions regarding the appropriateness of political activity on Facebook, as well as the specific kinds of political activities they engaged in and witnessed within the site, were also explored.

3 Facebook and Political Participation 3 Poking People to Participate: Facebook and Political Participation in the 2008 Election On January 5, 2008, candidates for the U.S. presidency met in New Hampshire for one of many debates during the primary season. However, this debate offered a new opportunity for candidates to engage young people: it was sponsored by Facebook, a popular social network site with 75% adoption by year-olds (Lenhart, 2009). Facebook members watching the debates could participate in online polls and post comments in real time. Immediately following both debates, 81% of users participating in a Facebook poll said the debates helped inform their voting decisions (Goldman, 2008). The 2008 U.S. presidential election continued a trend in political campaigning as candidates adapted their message to new communication tools. The popularity of social media sites like YouTube, Facebook and MySpace especially among younger voters provided a highly visible environment for candidates to promote themselves, articulate their platforms in detail, and interact with voters in fundamentally different ways. Likewise, these sites allowed users to interact with each other about political issues and to share and discuss their opinions. Data from Pew Internet reveal that 65% of SNS users ages engaged in at least one of five political activities on a SNS during the 2008 campaign, including joining a political group on the site and obtaining information about a candidate (Smith, 2009). However, little is known about the impact that sites such as Facebook have on the political behavior of young people. In particular, Facebook is an important social media site to study because of its high rate of use by younger voters in the demographic. During the 2008 election, both Republican and Democratic presidential candidates utilized the site, maintaining pages that allowed users to post comments, share news and videos, and connect with other users. Furthermore, Facebook members had access to various site features that allowed

4 Facebook and Political Participation 4 them to share their political views. But did these efforts make a difference in the political participation of Facebook users? If so, social network sites may be a way to engage more young people in political processes. Therefore, this study seeks to elucidate the relationships between political activity on Facebook and more traditional forms of political participation occurring onand offline. Literature Review Putnam s (2000) argument that political participation is declining due to a reduction in civic engagement (both political and non-political) implies negative consequences for the health of a representative democracy. Recent research indicates that interaction through the Internet can replace some of these lost forms of civic engagement (Krueger, 2002; Tolbert & McNeal, 2003), although there has been a debate about the nature of that effect (see Han, 2008; Scheufele & Nisbet, 2002). However, the evidence indicates that Internet use has provided an additional channel for citizens to engage politically with each other and their government. Using the Internet to engage young voters is of special interest, as this group of Americans is statistically the least-represented subsection of the population in many political activities, including voting, working on a campaign, and general interest in politics (Delli Carpini, 2000). As measured by voter turnout, young voters have long been perceived as apathetic toward politics. With the exception of 1992, young-adult voting rates in presidential elections declined steadily between 1972 and 2000 (Nickerson, 2006). Since 2000, however, there have been significant increases in this measure: voter turnout among year-olds increased 11% between the 2000 and 2004 presidential elections (Lopez, Kirby, Sagoff, & Herbst, 2005). Furthermore, the Center for Information and Research on Civic Learning and Engagement (CIRCLE) estimates that 23 million people under 30 voted in the 2008 election, an

5 Facebook and Political Participation 5 increase of 3.4 million over 2004, making the 2008 election the highest youth voter turnout since 1972 (U.S. Census Bureau, 2008; Youth Turnout Rate, 2008). One way of reaching young adults is through social network sites (SNSs). Sites such as Facebook and MySpace offer a venue for young people to express themselves politically, locate political information, and interact with their peers about politics. Rheingold (1993) speculated that virtual communities could help citizens revitalize democracy, or they could be luring us into an attractively packaged substitute for democratic discourse (p. 295). The question then arises, which of Rheingold s predictions is now being realized? Is the Internet encouraging and increasing political discourse specifically among young voters or has it replaced more traditional forms of political and civic engagement with low-commitment activities, like forwarding bogus petitions? To answer this question, we will first review literature addressing political participation and the impact of the Internet interaction on political engagement. Political participation on- and offline Political participation can take many forms, including such activities as campaign donations, attempting to persuade others, and taking part in activities related to politics (Kenski & Stroud, 2006). Conway (2000) conceptualized political participation as the set of activities that citizens perform in order to influence different levels of the government, such as its structure, policies, or officials. Taken together, political participation may be considered as one s intent to influence government actions through different activities, either directly, by affecting the creation or implementation of public policy, or indirectly, by influencing the people that make those choices (Burns & Schlozman, 2001). It is through political participation that individuals can control and take part in the policy decisions that might directly affect them (Best & Krueger, 2005).

6 Facebook and Political Participation 6 Jackman (1987) has noted that political participation in the United States is one of the lowest among the industrialized democracies in measures of political engagement, and Putnam (2000) points to two longitudinal surveys that show significant decreases in the public s interest in current events, politics, and membership in civic organizations during the last 25 years of the 20 th century. Voting, a key measure of political activity, has consistently declined since the late 1960s (Conway, 2000), with young adults showing greater rates of disengagement than any other age group (Delli Carpini, 2000). Voting in presidential elections by year-olds declined from approximately half of the population in the early 1970s to less than one-third of that age group by the 1996 election (Galston, 2004). More recent research points to increases in participation among young voters (e.g., Baumgartner & Morris, 2008), with the media and especially the Internet often cited as a key factor in effecting change and increasing knowledge. Media use has been associated with greater levels of involvement in civic activities, as well as higher levels of political awareness amongst American adolescents (Pasek, Kenski, Romer, & Jamieson, 2006). Young adults ages extensively used the Internet for obtaining election information in 2008: 58% went online for political news, 48% watched a political video online, and 65% of those with SNS profiles performed at least one of five political activities on the site (Smith, 2009). While Internet access is often thought of as primarily access to information resources (Tolbert & McNeal, 2003), Internet use is also associated with small but significant increases in political participation (Kenski & Stroud, 2006; Tolbert & McNeal, 2003). Some of the increase in participation amongst young voters may be linked to increasing access to and use of the Internet for political research and engagement. The Internet both supplements traditional offline methods of participation (e.g., posting

7 Facebook and Political Participation 7 videos from campaign rallies online) and provides new outlets for participation (e.g., personal blogs tackling political issues). Elin (2003) argues that the Internet provides a virtual space that allows individuals to immerse themselves in political information, which in turn can lead to offline political activities. The website Meetup.com, for example, enables people with similar interests to find each other online and then meet up offline. In the six months leading up to the 2008 presidential election, 1,472 Meetup users utilized the site to organize offline gatherings and groups in support of John McCain, and 13,702 did the same for Barack Obama (Havenstein, 2008). Weinberg and Williams (2006) found Meetup attendance related to the presidential candidates was positively related to campaign donations, volunteering, candidate support, and advocacy. Internet access alone does not generally increase political participation. However, among those with Internet access, exposure to political material does increase participation (Kenski & Stroud, 2006). Wellman, Hasse, Witte, and Hampton (2001) found that the more politically active people are offline, the more they participate in political discussions online. Hardy and Scheufele (2005) further found that exposure to and discussion of political information online had both a main and moderating effect on political engagement. The results of their national survey (N = 787) indicated that individuals who discussed politics with others via computermediated communication (e.g., , synchronous chat, discussion forums) also reported significantly higher levels of political participation (e.g., working for a political campaign, circulating a petition). These studies provide strong support for the Internet as a new repository for political information and a new outlet to engage in political discussions, both of which may be associated with greater political participation. Social network sites and political activity

8 Facebook and Political Participation 8 While hundreds of SNSs exist, the most popular sites exhibit a number of similar characteristics. These include the creation of a user profile, a list of users with whom one is connected, and the ability to view a list of one s connections and the connections of others within one s network (boyd & Ellison, 2007). These sites are especially popular among young adults, and several studies have found usage rates of 90% or higher amongst college students (e.g., Ellison, Steinfield, & Lampe, 2007; Golder, Wilkinson, & Huberman, 2007). Research shows that connections among users on these sites typically represent preexisting or offline relationships (boyd & Ellison, 2007; Ellison et al., 2007). The affordances of SNSs suggest they might be well suited to increasing general political knowledge and, subsequently, political participation. Users can become a fan of a candidate and download candidate applications to their profile pages. They can share their political opinions through the many communication methods on the site, from status messages and wall postings to joining various politically based groups. Friends can view all of this information as it comes through aggregators such as Facebook s News Feed, which collect the actions of each member s friends and present it in a list. Friends can also comment on News Feed postings, thus engaging others in active conversation about political issues. Thus far, little research has examined the political uses of SNSs during a campaign cycle. Gulati and Williams (2007) examined candidates use of Facebook s Election Pulse feature, which provided generic profiles to all candidates running for a congressional or gubernatorial seat in the 2006 mid-term election. They found that 40% of Senate candidates and 15% of House candidates updated their profile beyond basic information provided by Facebook, with major party candidates being significantly more likely to expand their profiles than minor party candidates. These data suggest that early on, politicians recognized the value young people place

9 Facebook and Political Participation 9 on SNSs and were interested in accessing this population. Studies specifically looking at the 2008 presidential primaries found that while candidates were willing to engage young voters, young voters were not necessarily comfortable with candidate use of SNSs. Hayes (2008) found that young people reacted negatively toward the presence of politicians on SNSs. Hayes, Zube, and Isaacson (2008) found that young voters were dissatisfied with the amount of political information candidates had on their profiles, suggesting that although SNS profiles were accepted as legitimate sources of political information, candidates did not provide sufficiently detailed information on their profiles. These studies reveal a conflict of interests whereby political candidates are willing to reach out to young voters, but their target audience generally perceives messages negatively. However, these studies examine political use of SNSs between young voters and official campaigns. Considering candidates ultimate unwillingness to concede message control during a campaign (Stromer- Galley, 2000), it should not be surprising that campaigns struggle to connect with young voters in more interactive environments. What may be more empowering for young voters is the peerto-peer rather than candidate-to-citizen interaction that is central to how young people use SNSs. MySpace and Facebook have been shown to promote some types of online to offline political activity. Partnering with Declare Yourself, a nonpartisan political group, MySpace facilitated an online voter registration drive that produced a printout for potential voters to send to their state election officials. Campaigns used MySpace and Facebook not only to promote candidates, but also to recruit volunteers. For example, Peter Franchot, a candidate for Maryland state comptroller in 2006, recruited 80% of his campaign volunteers online through MySpace and Facebook (Gueorguieva, 2008).

10 Facebook and Political Participation 10 Valenzuela, Park, and Kee (2008) addressed political participation on Facebook as part of a larger study of SNS effects on social capital. The authors used Verba, Schlozman, and Brady s (1995) definition of political participation, which encompasses activities having the goal or result of impacting government action through the shaping of public policy, either directly or through the election of those who create policy (as cited in Valenzuela et al., 2008). The activities of voting, working for political campaigns, donating money to candidates, and displaying political bumper stickers were used as examples of political participation. A regression analysis found that using Facebook Groups was the only variable to have a statistically significant positive impact. Further analysis showed a strong relationship between being a member of a Facebook political group and political participation. A positive, significant interaction between intensity of Facebook use and social trust led the researchers to suggest the effects of Facebook on political engagement were very strong for trusting people. Facebook s impact on political participation Originally launched in 2004 as a website for college students, Facebook has been adopted by many young adults in the United States as the SNS platform of choice. According to Facebook (2009), more than 200 million people have active accounts with the site. Facebook offers a number of methods for users to interact with each other directly (e.g., wall postings, instant messaging, messages) and indirectly (e.g., posting notes, status updates). College students use the site heavily, and have been reported to spend more than 63 minutes on average per day on the site (Steinfield, Ellison, & Lampe, 2008). Of particular interest to this study is Facebook s News Feed. Each time users log onto the site, the News Feed displays a continually updating list of friends activities, including status messages, recently uploaded photographs, new notes, and recently joined groups. While the

11 Facebook and Political Participation 11 previous section has identified studies exploring how direct dyadic or group communication may influence political activities, research has not yet explored the potential ramifications of information from a friend that simply appears online and without the influence of direct interaction on political participation. Examining how users react to friends political information as displayed on the News Feed allows us to determine the effects of this indirect interaction on political participation. This study examines trends in Facebook use by college students in the weeks leading up to the 2008 presidential election, both to quantify their political use of Facebook and to determine what relationships exist between students political activities on the website and their political participation in general, as measured by participation in activities such as writing to one s representative or running for political office. In addition, this research contributes to the growing body of literature dealing with impacts of new media on political participation, specifically how exposure to peers attitudes and interests may affect an individual s political activity. Although some work frames these activities as offline political participation, our use of the term political participation acknowledges that some of these activities can be accomplished via the Internet or other channels, as well as through traditional media. For instance, donating money to a campaign or political organization can be accomplished by mailing in a check or completing an online form. In either case, the end result is the same. We are therefore concerned with political participation regardless of medium. One way in which Facebook might increase political participation is through lowering the barriers to communication about political events and ideas. The site s News Feed simplifies the process of sharing information quickly with other members of one s network; 1 as soon as one person updates his or her status, posts a note or link, or creates an event, that information

12 Facebook and Political Participation 12 immediately populates into the News Feed, where friends can read and respond to the shared information. Users can easily share political information and, depending on the number and orientation of their Facebook friends, may be exposed to political content that is filtered through their peer network, potentially rendering it more powerful than traditional media channels. Thus, this paper focuses on the following research questions: RQ1a: What kinds of political activities do college students engage in? RQ1b: What kinds of political activity on Facebook do college students engage in? RQ2: Do college students perceive Facebook as an appropriate venue for political activity? RQ3: Does political activity on Facebook predict political participation? RQ4: Does political interest predict political activity on Facebook? Method Participants A random sample of 4,000 students was obtained from the registrar s office of a large, Midwestern university. Students were then invited via their university address to participate in an online survey hosted on Zoomerang. Three reminders were sent after the initial invitation to participate. The survey period lasted for two weeks in October We received 683 usable responses, yielding a response rate of 17% (see Table 1 for sample demographics). Survey respondents were entered into a raffle for one of a number of prizes as an incentive to participate. Table 1

13 Facebook and Political Participation 13 Demographics of Survey Respondents (N = 683) Demographic Mean Percentage Gender Male 32% Female 68% Age 20 Year in School Freshman 23% Sophomore 21% Junior 24% Senior 33% GPA 3.29 Residence On Campus 45% Off Campus 55% Ethnicity White 86% Non-White 14% Are you a member of a fraternity or sorority? Yes 7% No 93% Are you eligible to vote in the upcoming US election?

14 Facebook and Political Participation 14 Yes 98% No 2% Are you registered to vote? Yes 96% No 4% Are you a member of Facebook? Yes 96% No 4% Note: Numbers may not equal 100 due to rounding decisions. Materials Several measures were used to test the research questions of this study. The Facebook Intensity (FBI) scale developed by Ellison et al. (2007) was used to measure Facebook usage. Table 2 displays the FBI scale, along with the individual measures that comprise the scale. Number of friends and minutes spent on Facebook were open-ended variables, while the measures of attitudes toward Facebook were Likert-type scale items. Table 2 Summary statistics for Facebook Intensity Individual Items and Scale Mean S.D. Facebook Intensity a,b (Cronbach s alpha = 0.83) Approximately how many TOTAL Facebook

15 Facebook and Political Participation 15 friends do you have? Facebook Minutes Facebook is part of my everyday activity I am proud to tell people I'm on Facebook Facebook has become part of my daily routine I feel out of touch when I haven't logged onto Facebook for a while I feel I am part of the Facebook community I would be sorry if Facebook shut down a Individual items were first standardized by converting open-ended variables using the log10 function before taking an average to create the scale due to differing item scale ranges. b Unless provided, response categories ranged from 1=strongly disagree to 5=strongly agree. Additional items were developed to determine what role Facebook plays in allowing users to express their political views as well as the extent to which they are exposed to the views of others on the site. The Political Activity on Facebook scale, shown in Table 3, includes responses to the question, In the past week, which of the following have you done in Facebook? (Click all that apply). Questions were asked in a binary fashion; consequently, for individual items, we show the percentage of the sample that responded Yes to this question. The mean reported for the scale represents the average number of items checked by an individual user. For this, and following measures adapted from nominal response categories, the scale was developed using the set of responses by individuals to indicate level of participation. For example, a person

16 Facebook and Political Participation 16 who indicated more of the behaviors shown in the series was considered to have been more politically active on Facebook. Table 3 Scale for Political Activity on Facebook Percent of sample Political Activity on Facebook scale (Cronbach s Alpha = 0.75, mean = 1.33, SD = 1.97) Posted a wall comment about politics 20.4% Posted a status update that mentions politics 18.4% Joined or left a group about politics 13.8% RSVPed for a political event 13.8% Posted a photo that has something to do with politics 10.0% Posted a photo of myself or others at a political event 9.6% Discussed political information in a Facebook message 8.9% Become a fan of a political candidate or group 8.8% Discussed political information using Facebook s instant messaging system 6.9% Posted a link about politics 6.1% Added or deleted political information from my profile 5.8% Added or deleted an application that deals with politics 3.8% Posted a Facebook Note that has something to do with politics 3.6%

17 Facebook and Political Participation 17 Took a quiz that is about politics 2.7% We are also interested in what kinds of political information respondents receive from their peers within Facebook. Table 4 shows the Exposure to Network s Political Activity on Facebook scale, which comprises responses to the question, In the last week, which of the following have you seen on your News Feed? (Check all that apply). As with above, responses were entered in a binary fashion, and consequently, percentages of the respondents affirming the statement are shown. Table 4 Scale of awareness of political messages being shown on Facebook s News Feed Percent of sample Exposure to Network s Political Activity on Facebook Scale (Cronbach s Alpha = 0.84, mean = 4.91, SD = 3.42) A status update that mentions politics 70.0% Others joining or leaving a group about politics 51.2% Someone becoming a fan of a political candidate or group 51.0% Someone posting a photo that has something to do with politics 49.3% A photo of someone at a political event 48.4% A wall comment about politics 43.2% Someone is planning to attend a political event 42.5% A link about politics 41.9%

18 Facebook and Political Participation 18 A Facebook Note that has something to do with politics 35.5% Political information added or deleted from someone s profile 26.8% Someone adding or deleting an application that deals with politics 19.8% A quiz that is about politics 11.1% To control for previous political knowledge on the part of the respondents, items were drawn from Delli Carpini and Keeter (1996) to gauge political knowledge. The Cronbach s alpha for this validated scale is lower than reported in other research. In addition, the number of correct responses in general was higher than is usually reported for this scale. There are several possible explanations. First, since the survey was conducted online, it is possible participants researched answers online before responding. Second, because our sample was comprised of students currently enrolled in a university, they may have a higher level of political knowledge than the general population. Third, conducting the survey at a charged moment in the U.S. presidential campaign could have skewed responses. Table 5 Percentages of correct responses to questions of political knowledge Percent of sample answering correctly Political Knowledge Scale (Cronbach s Alpha = 0.57, mean = 3.94, SD = 1.20) What job or political office is now held by Richard Cheney? 91.6%

19 Facebook and Political Participation 19 Whose responsibility is it to determine if a law is constitutional or not... is it the president, the Congress, or the Supreme Court? How much of a majority is required for the U.S. Senate and House to override a presidential veto? Which party has the most members in the House of Representatives in Washington before the election? Which political party, if either, would you say is more conservative at the national level? 81.9% 69.8% 62.8% 87.5% Another control scale was created to account for potential effects caused by a respondent s sense of personal ability to affect government. The Individual Personal Efficacy scale is based on work by Verba et al. (1995). Table 6 shows the scale and individual responses comprising it. The questions we used are verbatim from the established scale. Table 6 Individual Personal Efficacy scale Mean Standard Deviation Individual Personal Efficacy Scale (Cronbach s Alpha = 0.75) If you had some complaint about a local government activity and took that complaint to a member of the local government council, do you think that he or she would pay:

20 Facebook and Political Participation 20 If you had some complaint about a national government activity and took that complaint to a member of the national government, do you think that he or she would pay: How much influence do you think someone like you can have over local government decisions? How much influence do you think someone like you can have over national government decisions? Note. Respondents answering this series of questions selected from the following responses: (1) none at all; (2) very little; (3) some; and (4) a lot. To determine whether the level of interest participants had in politics affected how Facebook was used for political participation, a Political Interest scale created by Verba et al. (1995) was employed. This 5-item scale asks participants to rate their political interest from 1 (strongly disagree) to 5 (strongly agree). The individual items are displayed in Table 7. Table 7 Political Interest Scale Political Interest Scale (Cronbach s Alpha =.806) Scale Items-Yes/No Mean Standard Deviation Everyone should be involved in working with community organizations and the local government on issues that affect the community. I think it is important to get involved in improving my community

21 Facebook and Political Participation 21 Being actively involved in national, state and local issues is my responsibility. I am interested in political issues I can learn a lot from people with backgrounds and experiences that are different from mine. Note. Respondents were asked to rate each statement along a 5-point Likert-type scale ranging from 1 ( Strongly Disagree ) to 5 ( Strongly Agree ). Finally, the central dependent variable of this research political participation was measured using a 12-item scale adapted from Rosentstone and Hansen s (1993) measures of offline political activity to determine respondents political participation, both in offline and online contexts. Participants indicated their experiences with a variety of political interactions in response to the question, We are also interested if you have done any of the following. This is a list of some things people do regarding government and politics. Have you done any of these things in the past year? (Check all that apply). Table 8 Political Participation Scale Percent of sample Political Participation Scale (Cronbach s Alpha = 0.63, mean = 2.68, SD = 1.86) Watched a 2008 election debate on television 84.6%

22 Facebook and Political Participation 22 Watched a 2008 election debate (clips or in its entirety) on the Internet 66.6% Signed a paper petition 31.0% Attended a public meeting on town or school affairs 24.2% Signed an online petition 20.7% Served as an officer of some club or organization 18.2% Volunteered or worked for a political organization 13.1% Wrote my national representative or senator 9.8% Belonged to a nonprofit group that is interested in better government 8.4% Held or ran for political office (including student government) 7.9% Wrote a letter to the paper regarding a political issue 2.6% Wrote an article for a magazine or newspaper regarding political issues 1.7% Findings Our first research question asked what kinds of political activities college students engage in, both in general and on Facebook. To answer the question of general participation (RQ1a), we took the sum of affirmative responses, as shown in Table 8, and created a measure with a possible range of 0-12 (reflecting the number of questions). Responses ranged from 0-10 (i.e., no one responded yes to every statement), with a mean of 2.68 and a standard deviation of The most common individual behaviors reported were related to watching campaign debates, either online or on traditional media. In general, the frequency of reported behavior shown in Table 8 is ordered from least burdensome to most, indicating that participation is a function of effort. Additionally, given the mean reported above and the frequency of individual responses, it appears that for most of our respondents, political participation mainly involved passively

23 Facebook and Political Participation 23 watching an election event, or perhaps signing a petition. The other part of our first research question (RQ1b) asked about the types of political activities college students reported doing specifically on Facebook. The Political Activity on Facebook scale, as shown in Table 3, shows individual behaviors reported on 14 political activities that can be performed within Facebook. Responses ranged from 0-12, with a mean of 1.33 and a standard deviation of The most common activities respondents reported include posting a politically oriented wall comment (20.4%), posting a status update referencing politics (18.4%), joining or leaving a political Facebook group (13.8%), and RSVPing for a political event (13.8%). These findings indicate that most students do not use Facebook extensively for political purposes. Our second research question (RQ2) related to whether our participants saw Facebook as an appropriate venue for political activity. We looked at a number of items related to students perceptions of Facebook as an appropriate forum for political discussion and display, as shown in Table 9. Respondents show slight agreement with questions regarding Facebook s appropriateness as a platform for sharing political beliefs (M = 3.32, SD = 1.05). However, respondents objected to using Facebook as a means of persuading others. Motivating others to vote, both in general (M = 2.01, SD = 1.15) and for a particular candidate (M = 1.88, SD = 1.13), were met with relatively strong disagreement. Finally, respondents showed some positivity regarding candidates presence on Facebook (M = 3.33, SD = 1.09). The data show that students are somewhat accepting of Facebook being used as a venue for political activity when it comes to self-expression, but they view the site as an inappropriate medium for attempting to change others political beliefs or activities.

24 Facebook and Political Participation 24 Table 9 Perceptions of Facebook as a Forum for Political Discussion Question Mean Standard Deviation Facebook is an appropriate place for people to express their politics. I use Facebook to express my political views It's good that presidential candidates have pages on Facebook I use Facebook to convince my friends to vote I use Facebook to convince my friends to vote for my candidate Note. Respondents were asked to rate each statement along a 5-point Likert-type scale ranging from 1 ( Strongly Disagree ) to 5 ( Strongly Agree ). Respondents who felt most strongly that Facebook was an appropriate medium for political expression also showed a tendency to use the site to that end. A Spearman s correlation coefficient was calculated for the relationship between subjects perceptions of appropriate Facebook use and political activity on Facebook. A strong, positive correlation was found (r = 0.256, p <.001), indicating a significant linear relationship between perceptions of Facebook as an appropriate medium for political communication and the amount of political activity one engages in on Facebook. The third research question (RQ3) sought to understand whether political activity on Facebook predicts levels of other forms of political participation as measured by activities like

25 Facebook and Political Participation 25 signing petitions (online or paper), attending community meetings, and volunteering for political organizations. A model to predict political participation was constructed by combining variables into an OLS regression model to detect whether variables remained significant predictors of the reported behavior when in the presence of other possible influencers. Table 10 shows the results of this regression predicting political participation (as measured by the scale in Table 8). Independent variables included several controls for possible alternative explanations. These include demographic variables (Table 1), scales for political knowledge (Table 5), and scales for Individual Personal Efficacy (Table 6). Three measures of Facebook participation were included: Facebook Intensity (Table 2), Political Activity on Facebook (Table 3), and Exposure to Network s Political Activity on Facebook (Table 4). We tested for multicollinearity between these terms using a Variance Inflation Factor (VIF) and did not find significant collinearity between terms. We report standardized coefficients to account for different types of data included in this model. The N reported for this model is lower than reported above as the regression analysis excludes cases with missing data. Gender and Age were treated as a dummy variable, with Female and Democrat being the included variables. Table 10 OLS regression predicting political activity. Std. Error of Adjusted R 2 the Estimate N ANOVA F = , p <.001

26 Facebook and Political Participation 26 Independent Variables Standardized t-score Significance Coefficient (Constant) Gender Age Year in School GPA Political Party Political Knowledge Individual Political Efficacy Political Interest Facebook Intensity Political Activity on Facebook Exposure to Network s Political Activity on Facebook The analysis of variance (ANOVA) revealed that the overall model was significant (F (2, 458) = , p <.001, adjusted R 2 =.318). The standardized beta weights of this model are reported in Table 10. Political activity on Facebook is the most significant predictor of political participation in the model (β =.239, p <.001); exposure to political activity on Facebook is also a significant predictor of political participation (β =.141, p <.001).

27 Facebook and Political Participation 27 In this model, we find that most demographic information is not associated with political participation. The exception is political party affiliation, with the respondents who identified themselves as Democrat having an increased chance of listing more political participation behaviors. Political Knowledge and Individual Political Efficacy were both strong predictors of political participation, which is consistent with previous literature (Galston, 2004; Verba et al., 1995). Facebook Intensity is significantly related to political participation, but the coefficient indicates a negative trend, that is to say that more intense Facebook use is actually associated with a decrease in political participation. The other measures of Facebook use are strongly positively associated with political participation, with Political Activity on Facebook in particular showing the strongest positive correlation of the included variables. It may be that politically active people are more likely to express those politics on Facebook, or that the people who use Facebook for communication, as opposed to games, are more likely to express their political beliefs. Our final research question (RQ4) addressed the role of political interest on the likelihood of political participation on Facebook. An OLS regression was run using the same demographic and control variables utilized in the regression addressing our third research question. As shown in Table 11, a strong relationship between political interest and political participation on Facebook exists (β =.157, p <.001). However, political interest alone is not the most significant relationship. As might be expected, those who are heavier Facebook users are more likely to participate politically on Facebook (β =.241, p <.001). This finding, in addition to the findings reported from the prior OLS regression, suggests that while FBI appears to have a negative relationship with offline political participation, it is strongly and positively related to Facebook

28 Facebook and Political Participation 28 political participation. Table 11 OLS regression predicting political activity on Facebook. Std. Error of the Adjusted R 2 Estimate N ANOVA F = , p <.001 Standardized Independent Variables Coefficient t-score Significance (Constant) Gender Age Year in School GPA Political Party Political Knowledge Individual Political Efficacy Political Interest

29 Facebook and Political Participation 29 Facebook Intensity Political Participation Exposure to Network s Political Activity on Facebook In addition to FBI, political participation is also strongly related to Facebook political participation (β =.250, p <.001), suggesting that those who are politically engaged seek multiple outlets for their political behaviors. This may also suggest that Facebook has some legitimacy as a political tool, as those that are engaged in historically valued forms of offline participation such as volunteering and petitioning are also using Facebook to achieve their political goals. While this research did not cover the full range of political activities one can perform on Facebook or the frequency with which users engaged in such activities, we believe there must be some perceived utility in Facebook as a political tool if those who are more actively participating offline are also actively participating on Facebook. Discussion Our first research question asked about the extent to which participants were active politically in the past year and the ways in which they used Facebook for political purposes. The two most frequent forms of political participation indicated by our respondents were acts of media consumption: watching a debate on television and watching a debate on the Internet. While these two types of participation are relatively passive when compared to more active tasks such as writing a letter to a newspaper regarding a political issue, they do reflect information-

30 Facebook and Political Participation 30 seeking behavior and a desire to be informed. Out of the twelve activities included in the instrument, the average respondent reported less than three. However, our sample overwhelmingly reported being likely to vote 92% said they were likely or very likely to vote in the election. This could mean that the group anticipates participating in an episodic fashion, i.e., during a national election, but does not participate in other situations, or it could be a result of the increased effort to get out the vote among younger voters during the campaign season. It should also be noted that self-reports of voting suffer from issues of social desirability (Holtgraves, 2004; Silver, Anderson, & Abramson, 1986). We also asked respondents about the political activities on Facebook they had engaged in during the previous week. We found that at least some Facebook users utilize the site for purposes of political self-expression, with 48% of respondents reporting they have used Facebook for at least one of the 14 political activities we asked about. The most common activity was posting a wall comment on a political topic, with 20.4% reporting they have done this in the previous week. At first glance, this seems like a small percentage of users engaged in this type of activity. However, when one considers recent research finding that, among first-year college students, only 33.8% reported discussing politics sometime during the previous year (Pryor et al., 2006), this level of activity seems fairly consistent with general political interest among this population. Most likely, reported rates would have been higher had we asked about behavior over a six-month or one-year timeframe, rather than only asking about a one-week time period. Respondents slightly agreed with the statement that Facebook is an appropriate place to express your politics (M = 3.32, SD = 1.05), suggesting that in general, this behavior was condoned, perhaps because political expression was viewed as a component of self-presentation. Given that in some circles in the U.S., it is considered inappropriate to discuss politics or

31 Facebook and Political Participation 31 religion, we are hopeful that Facebook might provide a low-risk way for students to engage with one another about politics. Additionally, in that this constitutes the first presidential election of the Facebook era, it may be that political behavior on the site has yet to become a norm and that we will see more political expression and discussion within SNSs in the future. Participant responses to an open-ended question asking for additional comments reflect this ambivalence about political uses of the site, suggesting that norms around these activities are still nascent: I don't really use Facebook to answer my political questions at all. I do, however, believe its okay to express your views and opinions in a rational matter. I believe that Facebook may be useful for disseminating information about political rallies, and so on, but not really useful for propaganda, or swaying votes, except insofar as it facilitates communication between friends, and provides adspace on a popular social networking site. I personally don't like Facebook for the purpose of expressing political issues because that was not the intended purpose of Facebook. However, I can definitely understand why it has become a vehicle for political ideas. Conceptions of appropriate behavior will no doubt be contested as the Facebook userbase becomes more heterogeneous (e.g., one person s rational expression of views may be another person s propaganda ). Our third research question asked whether political activity in Facebook was related to political participation. Our model explained approximately 29% of the variance in political participation using demographic variables and three Facebook-related measures: Facebook Intensity, Political Activity on Facebook, and Exposure to Network s Political Activity on Facebook. In other words, general intensity of Facebook use is only mildly associated with political participation, but propensity to use Facebook for political expression and to receive political messages is highly associated with that outcome. One possible explanation for our findings is that those who are politically active in other settings are also politically active on

32 Facebook and Political Participation 32 Facebook, or are more attuned to political messages. Conversely, Facebook may provide an environment for unengaged users to explore political activity, which then translates into real world political participation. Similarly, seeing one s friends engage in political activity may help make these activities more normative, also encouraging previously unengaged users to take action. Our results suggest that Facebook users are likely to see political activity in the News Feed items of their network, especially during political campaign seasons or during politically charged moments. For example, 70% of our participants reported seeing status updates that mentioned politics, 51.2% reported seeing someone join or leave a political group, and 51% reported seeing someone become a fan of a political candidate or group. Further work needs to be done to address the directionality of this relationship, perhaps with longitudinal data. In many ways, Facebook supports political activity through its technical and social affordances. The site enables individuals to find others with shared political beliefs through features such as political Groups and Pages. It includes political affiliation as one of its profile fields, suggesting that this ranks as an important identity marker for the site designers. Through a broad range of public and private communication features, it enables user communication with a large network of friends, giving those with a political message an effective platform for evangelizing. However, our participants seemed less enthusiastic about the possibility of Facebook being used as a persuasion tool. This may be a residual effect of the highly social environment Facebook is associated with our respondents may conceive of Facebook as a purely social space and may resent the intrusion of political pressure, even though messages that mention politics are accepted. Blatant attempts to persuade one s friends to vote for a particular candidate, however, are likely to be met with resistance, as suggested by the fact that the mean score for agreement with the statement, I use Facebook to convince my friends to vote for my

33 Facebook and Political Participation 33 candidate was only 1.88 (SD = 1.13). Limitations This study is limited in its generalizability due to the fact that we only examined members of the Facebook community, and those users were all students at one institution. Additionally, the cross-sectional design prevents us from establishing causality. Another limitation may have been the visit by presidential candidate Barack Obama to the university s campus two weeks before the launch of the survey. This visit may have increased levels of political activity on campus, which then spilled over onto Facebook. To address these concerns, future research should analyze the impact of other SNSs on political participation. A random sample of all users of an SNS would also benefit this area of research. Also, conducting surveys at universities that received visits by presidential candidates, as well as those that did not, would allow researchers to determine if these live appearances by the candidates impacts Facebook political activity. While the timing of this research to coincide with the presidential election is not a limitation, it would be prudent to do research during a time period outside of an election season. Presidential elections are typically times of high political engagement due to their level of importance. Consequently, the levels of political activity reported by our sample are likely to be greater than those in periods between election cycles. Research conducted during non-election periods would add a different dimension to research on SNSs and political participation, complementing the results presented in this paper. Beyond survey research, experimental interventions would also prove useful in moving this area of research forward by providing means to establish causal claims. Conclusion