Is Social Media Changing How We Understand Political Engagement? An Analysis of Facebook and the 2008 Presidential Election

Abstract

This research conceptualizes political engagement in Facebook and examines the political activity of Facebook users during the 2008 presidential primary (T1) and general election (T2). Using a resource model, we test whether factors helpful in understanding offline political participation also explain political participation in Facebook. We consider resources (socioeconomic status [SES]) and political interest and also test whether network size works to increase political activity. We find that individual political activity in Facebook is not as extensive as popular accounts suggest. Moreover, the predictors associated with the resource model and Putnam’s theory of social capital do not hold true in Facebook.

Keywords

Facebook political participation 2008 election social media social network sites

Introduction

There is little doubt that social networking sites (SNS) such as Facebook have raised the attention of scholars interested in social media’s effects on the political landscape. In October 2007, roughly one year before the 2008 U.S. presidential election, Facebook touted more than fifty million active users with a majority logging onto the system at least once a day (Facebook 2009b), and by August 2008, a little over one month before Election Day, Facebook reached one hundred million active members (Facebook 2010). Social media also played a major role in the 2012 U.S. presidential election with Facebook citing nine million of its users voting in the November election (Facebook 2012). Moreover, Facebook and other SNS have been used to mobilize individuals to participate in protests around the globe such as the London youth riots in the summer of 2011 and the 2009 Iranian protest against the reelection of Mahmoud Ahmadinejad and most especially those in the Middle East, collectively referred to as the Arab Spring. Despite popular accounts illustrating the ability of social media to mobilize users for political activity, little empirical work in the academy has measured the nature of political engagement occurring within these sites. The 2008 U.S. presidential campaign, often regarded as the first Facebook election, offers an excellent opportunity to address this topic.

While the Internet is not a new player in American campaigns and elections, the 2008 U.S. presidential campaign stands out in terms of the prominent role that SNS, in particular Facebook, played. Facebook, for example, cosponsored with ABC News a presidential debate on January 5, 2008, and provided users the ability to become actively involved before, during, and after the debate in its “U.S. Politics” application (Facebook 2009a). As part of the “application,” users were able to give live feedback, take part in debate groups, and see behind-the-scenes postings from ABC News reporters at the debate, add support for their favorite candidate, and even register to vote. Inasmuch as user engagement during the presidential debate is unique, our article seeks to understand how SNS, such as Facebook, are changing the nature of political participation. We consider whether the same antecedents that predict offline participation also predict online participation. Finally, we consider the extent to which users engaged themselves over the course of the 2008 primary and general elections rather than during the one, isolated event. Using data compiled from student questionnaires, school records, and the students’ Facebook user profiles (text, images, applications, correspondence, etc.) throughout the 2008 election cycle, we assess the level of political participation demonstrated by college undergraduates and recent graduates in Facebook and factors that influence that participation.

Our research uses a resource model to test whether the same factors helpful in understanding offline political participation are also useful in explaining online political participation, especially in the context of Facebook. A resource model hypothesizes that resources, political engagement, and recruitment facilitate political activity. Our model focuses on the first two components of the resource model—resources (socioeconomic status [SES]) and political engagement (specifically, political interest). In addition, we test whether network size works as Putnam (2000) would predict—to increase one’s level of political activity.

The Internet and Politics

Scholarly research on the role the Internet has played in civic and political life has mostly followed two trajectories. The first considers the manner in which the Internet as a communication tool is used by campaigns, candidates, and causes (Bimber and Davis 2003; Foot and Schneider 2006). The second trajectory seeks to understand and explain the effects of new media on individual civic and political behavior (Bimber 2003; Drew and Weaver 2006; Jennings and Zeitner 2003; Johnson and Kaye 2003; Katz and Rice 2002; Shah et al. 2005). Understanding how the Internet, in a general sense, affects political and civic engagement is indeed important and much has been learned from such research. Nevertheless, there is a gap in our understanding of how online social networks (e.g., Facebook) foster political engagement and activity. This research seeks to fill this gap.

Research on traditional offline social networks suggests the important role an individual’s social network plays in facilitating political engagement. Putnam (1993, 1995a, 2000) describes how an individual’s membership in civic organizations such as the Elk’s, bowling leagues, and Rotary serve as conduits to promote political participation. As the story goes, social networks, such as a bowling league, help to foster interpersonal trust and cooperation that spreads between and among individuals in these informal social networks and from which spring the potentiality of civic and political engagement that serves community and democracy at large in the real world.

Bimber (2003, 199) suggests that throughout U.S. political history, consecutive information revolutions have led to a larger percentage of the voting population being disengaged from the political process. He notes that engaging in the political process over the years has become less compelling thereby disaffecting voters. The question of whether the Internet revolution carries the same consequences for political participation as previous information revolutions has become the focus of much scholarly research. Putnam (2000) and others (Best and Krueger 2005; Bimber 2001, 2003; Johnson and Kaye 2003; Katz and Rice 2002; Nisbet and Scheufele 2004; Shah et al. 2007; Xenos and Moy 2007) find that the Internet (when measured as frequency of Internet use) has had either a negative or insignificant relationship with engagement. In some cases (Bimber 2001; Katz and Rice 2002), research demonstrates mixed results where different scales of engagement, measures of Internet use, or control variables yield different, positive, and perhaps significant results.

Political communication scholars suggest that the contemporary, new media era lowers the cost of accessibility to political information thereby making it more likely that people are willing and able to invest themselves politically. Some argue that while evidence shows that the Internet facilitates political engagement, such engagement is demonstrated by those already engaged. That is, the Internet has only perpetuated and reinforced existing disparities—such as those associated with gender, SES, race, and age (Bimber and Davis 2003; Davis 1999; Hill and Hughes 1998; Jennings and Zeitner 2003)—in political engagement, doing little to engage those who are typically disengaged (Bimber 1999; Norris 1999). These studies suggest that “[t]he new information environment has not changed levels of engagement in any substantial way” (Bimber 2003, 24).

Another tack in the accessibility of information argument is that the Internet can help convert or mobilize the politically disengaged to become politically engaged. The Internet offers convenience and accessibility for a larger swath of citizens, increased access to information, online opportunities for political expression and political action, identification and affiliation with like-minded citizens, and “the convenience (or novelty) of online engagement may draw in those disillusioned with traditional modes of political participation” (Boulianne 2009, 3). Although research on the effect of the Internet as a tool to organize for political participation is inchoate, several findings suggest that the Internet can propel individuals into political life especially in terms of allowing them to gather political information, connect with others, mobilize, and recruit individuals to causes and actions because the Internet can significantly reduce the costs of participating (Bonchek 1995, 1997; Johnson and Kaye 2003; Leizerov 2000; Norris 2000, 2004; Resnick 2004; Shah, Kwak, and Holbert 2001; Tolbert and McNeal 2003). Clearly, it is less time-consuming to sign an online petition or donate money with the click of a mouse than it is to do so offline.

SNS and Political Engagement

While research discussed above contributes to our understanding of the role the Internet has had with regard to political engagement, the principal challenge to the above conclusions is that in much of the research the Internet is viewed as a single, monolithic technology and considering the Internet or Internet use in this broad and general perspective is limiting. Indeed, a component of the Internet that has gained significant attention by users and scholars alike is social media or SNS, such as Facebook. Whereas the findings above suggest that Internet use has had little if any impact on political engagement, research of online communities and social networks (Rheingold 2000, 2002; Schuler and Day 2004; Smith and Kollock 1999; Wellman and Haythornthwaite 2002) clearly demonstrates the ability of the Internet to support collective civic and political actions at local, national, and global levels. Indeed, while SNS made their initial appearance during the 2008 elections, according to a recent Pew (2011) study, “22% of online adults used Twitter or social networking sites such as Facebook or MySpace in the months leading up to the November, 2010 elections to connect to the campaign or the election itself.”

An important step logically here is to move beyond generic Internet use and information accessibility to understand social media as the locus of online public life and therefore the organizing environment within which political engagement is substantiated. To do this, we draw from the understanding that the Internet reduces the barriers to participation and thus reduces social inequality that exists in public life (Bonchek 1995, 1997; Johnson and Kaye 2003; Leizerov 2000; Norris 2000, 2004; Resnick 2004; Shah, Kwak, and Holbert 2001; Tolbert and McNeal 2003). We also consider Putnam’s conception of social capital. In the spirit of Putnam, Bourdieu, and Wacquant (1992, 14), we consider social capital as “the sum of the resources, actual or virtual, that accrue to an individual or a group by virtue of possessing a durable network of more or less institutionalized relationships of mutual acquaintance and recognition” (cited by Ellison, Steinfield, and Lampe 2007). In terms of political discourse, social capital is the consequence of a robust social life of, “—networks, norms, and trust—enable[ing] participants to act together more effectively to pursue shared objectives” (Putnam 1995b, 664–65). Therefore, we expect social capital to have a positive effect in its ability to foster greater commitment to active political participation both offline as well as within online social networks.

Conceptually, there is a difference with the first generation of Internet and politics research cited and research that examines the role of social media directly. Notably, first generation research examined the medium of the Internet as a singular technology with the potential to influence the passive political actor’s behavior and knowledge. We see social media as having the same influence but taking it one step further to provide the individual a set of unique tools that enables greater flexibility to actively engage in the public sphere. Benkler (2006) discusses how the emergence of the networked public sphere allows individuals to take advantage of capabilities that make them greater participants in the conversation. We believe that social media technology is one of these capabilities that has fundamentally changed the landscape in which the user/political actor is engaged and that social media technologies such as Facebook through its infrastructure provides tools to facilitate engagement. First, as (Bond and Fariss et al. 2012) have found, the technology allows individuals the means not only to build a network of connections but also to be influenced and to influence that network exponentially. Second, as Benkler (2006) and Jenkins (2006) allude, social media eliminates the costs and boundaries for the individual online political actor to create and share content within a networked environment. We also believe Facebook itself is uniquely positioned to facilitate online engagement because its feature-set (e.g., the “newsfeed” and user “wall”) acts as mechanisms to support the individual’s voice in broadcasting political content to a networked audience or online public sphere.

Facebook-specific research has examined the use of the network in the development of social capital and online community and found positive effects (Ellison, Steinfield, and Lampe 2006, 2007; Vanden Boogart 2006). For example, Ellison, Steinfield, and Lampe (2006) examine Facebook usage intensity against satisfaction with college life across three types of social capital: bridging, bonding, and high school (maintained). They find that Facebook plays a significant role in reducing barriers to participation in college life for low self-esteem students and more so than it does for their high self-esteem peers. Ellison, Steinfield, and Lampe (2006, 32) also make a significant theoretical finding that Facebook users, “are using the online channel less to meet new people than to intensify and solidify relationships that started offline.”

Extant research in offline social networks also suggests that network size can increase the likelihood that an individual will come in contact with other politically active individuals, thereby having a positive effect on that person’s likelihood to participate (Kwak et al. 2005; Leighley 1990; Verba, Schlozman, and Brady 1995). If Facebook can foster social capital and larger networks foster political participation, it seems likely that Facebook, due to the ability of users to easily expand and maintain large networks, will facilitate participation for users with larger networks than those with smaller networks.

A frequent criticism leveled at youth is that they are increasingly and disproportionately disengaged from politics as a result of their position in the life cycle (Loader 2007; Strate et al. 1989; Verba and Nie 1987; Zukin et al. 2006). Yet, with young Americans more quick to embrace the advances in Internet technologies than any other cohort and the overwhelming propensity, 70 percent of eighteen- to twenty-five-year-olds perceiving the Internet as a “useful source of political and issue information” versus the 48 percent of those over twenty-five (Delli Carpini 2000; see also Pew 2011¹) demonstrates the need to investigate youth, new media, and the effects on political participation. Increasing usage rates within SNS among the young and the perception that social media were instrumental in engaging youth during the 2008 election, our research has practical relevance and implications for political mobilization and engagement in American politics.

Finally, scholars have long considered elections and the stimuli they provide as important to voter mobilization. For example, Campbell’s (1960) hypothesis of surge and decline suggests that the short-term force of information that flows in the context of an election influences voter turnout. Due to the fact that presidential elections produce a greater deal of new information in addition to the overall symbolic importance of a presidential election, we suspect that Facebook is no different insofar as mobilization or activity is concerned. However, we are curious as to what particular forms of participation experience the largest response to the increasing intensity of the campaign.

A unique aspect to our study is that while many scholars have suggested that the Internet fosters offline political engagement by reducing the associated cost, we suggest that the Internet, social media, more specifically, can also foster online political engagement by reducing the cost to participate and, perhaps, even more than it would for offline engagement, given the unique features SNS such as Facebook provide users. It is one thing for an individual to use the Internet to gather information about an issue but another to then use that information to fuel offline engagement. The cost of gathering the information is reduced but the cost of the offline behavior resulting from that information is unchanged. However, if one were to use the Internet to gather information and then by a mere click of a mouse, join an online protest, sign an online petition, recruit a friend to join a cause via his or her network connections, or donate money via online contribution form, the cost of participating is significantly reduced and it is now far less expensive for the individual to engage in online political activity than it is for the offline equivalent. By reducing the cost on both sides of the equation, in addition to the extra boost generated by the social capital created by using social media, it seems possible that Facebook specifically can nurture [online] political engagement, bridge the participation gap by mobilizing those who find the cost of offline participation too high, and stimulate interest in specific political activities that perhaps are more effective in an online environment than they would be in an offline environment. Thus, our study seeks to explore the following research questions:

Research Question 1: Which forms of online political participation offered by Facebook register the largest response to the increasing intensity of the 2008 U.S. presidential campaign?

Research Question 2: How do traditional predictors of participation such as income, race, and gender work with regard to political participation in Facebook?

Research Question 3: How does a Facebook user’s network size (number of friends and number of groups) relate to that user’s political participation?

Our measure of online activity considers both new and old modes of engagement but it considers them in an online environment. We do consider aspects of offline engagement, such as voting or attending a meeting or rally, for example, but we focus on the nature of online political engagement, the correlates of online political engagement, and whether online engagement is merely offline engagement in a digital space or is online engagement altogether different. We suspect that there are new or evolving forms of political engagement that have adapted to a digital environment and if this is the case, these new forms that are showing up in a digital spaces, such as Facebook, have value in and of themselves, and are worthy of investigation.

Data and Measures

The data for this study come from two sources. First, data were collected from student surveys and school records of University of California undergraduates enrolled in sixteen real-time general education courses between September 2003 and June 2007 (N = 1,014). Second, we expand the above data set with additional data measuring political engagement collected via content analysis of Facebook user profiles of those participants in our sample who have “open” (or public) profiles. With participant permission, we matched students surveyed to their Facebook profiles and noted those who had Facebook profiles and also noted whether their profile was open (public) or closed (private). In an attempt to understand the effects of the 2008 presidential campaign on political engagement, this research uses an experimental design to measure the change in online political engagement, via Facebook, over the course of the 2008 presidential primary and general elections. Therefore, we collect and content analyze Facebook profile data at two points in time: (1) in January 2008 prior to Super Tuesday (February 5, 2008), and (2) immediately prior to and after the general election (November 4, 2008).² Of the 1,014 students who participated in the initial survey, approximately 460 maintained open Facebook profiles at T1 and approximately 326 at T2. Obviously, we encountered participant decay with a little over 100 participants securing or altogether deleting or closing their Facebook profiles. With regard to our relevant measures, however, we determined via t-tests on a variety of demographic variables the only significant difference between those with open Facebook profiles and those with closed profiles is with regard to the number of friends. Students with closed profiles, on average, had fewer friends (M = 2.46, SD = 0.99) than those with open profiles (M =2.71, SD = 1.17), t(709) = 2.295, p = .025. Our initial hunch was that many of those who closed their profiles were students who had graduated, were entering the workforce, and, in an effort to keep that which happens in Facebook, they might have altogether closed their more “unprofessional” student profiles. However, tests revealed no significant effect due to year in school.

Content analysis was carried out by two undergraduate students who were trained and provided with a codebook³ and archived copies of participant Facebook profiles. Intercoder reliability⁴ was established to be nearly 90 percent. Facebook profiles were coded for the presence or number of just under fifty (mostly political) characteristics. As a result of the near impossibility to predict the totality of images, groups, words, and terms that could be coded into our categories, we implemented a general rule to code text and/or images as belonging to a particular category (Nachmias and Nachmias 1987).⁵

Dependent Variables

For our dependent variables, we create a simple additive index of political activity (0–12) constructed from thirteen dichotomous variables that indicate whether the participant engaged in a particular political activity. A high score indicates the individual is more politically active. The political activity index includes a variety of political activities⁶ occurring within the Facebook environment, including political discussion (as indicated by number of politically oriented posts on the user’s Wall), number of political events posted and/or attended (e.g., political speech, debate, rally, protest, fund-raiser), whether an individual publicizes the intention to vote or that he or she actually did vote, and also for which candidate the user voted. We also include in the index variables that measure the presence of Facebook political applications such as whether the user supports any political cause(s), whether the user has recruited a friend to join a particular cause, and whether the user has displayed any political bumper stickers, political buttons, and so forth, on his or her Facebook profile. Please see the index for a breakdown of the individual measures included in the analyses. The reliability (Cronbach’s α) of the political activity index is .70.

Independent Variables

We consider variables⁷ common to research regarding political participation including those associated with the resource model, including parental income,⁸ political interest,⁹ GPA, and SAT. To test Putnam’s hypothesis, we include network size (number of Facebook friends and number of group memberships). We also include a variety of control measures, including age, sex, ethnicity, and political ideology. Scholars of political behavior have long recognized the above variables as standard explanatory variables related to political engagement (Abramson, Aldrich, and Rohde 2006; Brady, Verba, and Schlozman 1995; Hess and Torney 1967; Jennings 1993; Luskin 1990; Niemi and Junn 1998; Putnam 1993, 1995a, 2000; Schlozman et al. 1995; Wolfinger and Rosenstone 1980).

With regard to network size, there are differences in the manner in which scholars have operationalized it from number of “discussant generators” (Huckfeldt and Sprague 1995) or the number of people with whom an individual discusses politics (Kwak et al. 2005). For our measure of network size, we include a raw count of the number of Facebook “friends” a particular user has and the number of groups to which the user belongs.

Limitations

There are a few limitations to our study worth discussing. First, our sample was drawn from a population of students at a large, public university in California. While this stage of the sample is most likely representative of college students in general, the second stage of our sample includes only those students from the first stage who maintain a public profile on Facebook. As a result, there could very well be an issue of bias due to self-selection insofar as the students who maintain a public Facebook profile might differ significantly from those who either do not have a Facebook profile or have secured their profile. Generalizations, therefore, should be made with care. A second limitation is that our sample comprises students who completed a lower division undergraduate course offered by one of four specific academic departments: computer science, psychology, history, and/or writing, which some might argue yield an unrepresentative sample. However, these are large, introductory classes that likely capture a cross section of the student population. Finally, one of the joyously frustrating aspects of conducting live web and Facebook research is that it is essentially trying to shoot the proverbial moving target. Over the course of the year between T1 and T2 when we captured Facebook data, Facebook underwent a significant two-step interface modification, the first of which occurred in July 2008 when Facebook launched the optional beta interface for “the new” Facebook, and in September when all users were officially migrated to the new interface. While we were able to adjust our archival technique to mitigate any threats to our data and what we were able to collect, we cannot guarantee that there was no effect on what we captured and thus coded. The nature of data collection between T1 and T2 as referred to above can also account for participant profile modifications where content deletion, editing, and securing one’s profile become commonplace in particular as Facebook’s interface changes were highly publicized and profile security was highlighted in the media.

Results

The purpose of our study is to understand whether Facebook users are politically active in Facebook and, if so, to identify the antecedent variables of that political activity, including those associated with Putnam’s theory of network size as well as those associated with the resource model. Also, we assess the forms of online political participation in Facebook that yield the largest response to the increasing intensity of the campaign from the primary to general election. To begin with this part of the analysis, we first consider the nature of political activity in Facebook. We assume that Facebook users will be more active politically at T2 (the general election) than at T1 (the primary election). Indeed, according to the results of our independent-samples t-test, presented in Table 1, there is a statistically reliable difference in the level of political activity for the primary election (M = 0.24, SD = 0.7056) and the general election (M = 1.399, SD = 1.48), t(752) = −7.149, p = .000.¹⁰

Table 1.

Political Activity on Facebook Primary and General Elections.

	Primary election M or % (n)	General election M or % (n)
Political Activity Index^a (α = .70)	0.24 (450)	1.48 (304)
Profile picture	2% (9)	0.3% (1)^b
Political events	2% (7)	8% (25)^b
Political notes	0.7% (3)	1.0% (3)
Posted items	1% (6)	0.7% (2)^b
Political discussion (on the wall)	3% (13)	18% (54)^b
Political status update	2% (9)	14% (42)^b
Voted or intends to vote	0.4% (2)	15% (45)^b
Candidate support	2% (8)	12% (35)^b
Political applications	5% (23)	8% (24)^b
Political “causes”	4% (19)	7% (20)
Donate to political cause(s)	2% (9)	0.7% (2)
Recruit other to join political cause	2% (7)	0.3% (1)
	α = .58	α = .72

t(411) = −7.149, p = .000.

Indicates significant difference between primary and general elections for each individual activity according to independent-samples t-test. Results of individual t-tests are available upon request.

Moreover, our t-tests reveal statistically significant differences for particular types of political participation between the two elections. Table 1 also contains the overall level of participation for each election as well as the level of change among specific political activities that comprise the political activity index. For the primary election, political activity was not all that extraordinary. The specific political activities in which a larger proportion of users were engaged include political applications, political causes, and political discussion with 5, 4, and 3 percent of our sample, respectively, engaging in these activities. Only 0.4 percent of our sample voted or intended to vote during the primary election according to their Facebook profiles. In contrast, the specific political activities during the general election in which more users participated include political discussion, political status updates, votes or intends to vote, and supports a candidate. The proportion of our sample who participated in political discussion and political status updates both jumped more than 15 percent from the primary to the general election. The increase in candidate support and political events was less at 10 and 6 percent, respectively. The smallest increase was among those who joined a “political cause,” showing only a 2-percent increase from the primary to the general election. For a few activities, participation declined from the primary to the general election. Donating money to and recruiting others to join a political cause both declined over the course of the election, which is counterintuitive to offline political engagement. However, the difference for these two activities was not statistically significant, according to our t-tests. Only politically oriented profile pictures and posted items proved to show a statistically significant decrease between the primary and general elections.

What the reader should note, however, is that political activity during either election was not altogether remarkable. As Figure 1 shows, Facebook during the 2008 presidential election was not the hotbed of political activity that popular accounts may have us believe. During both the primary and general elections, users were more likely to be nonactive politically than they were to be politically active. Nevertheless, we do see that several types of political activities increased over the course of the election. This finding conforms to our expectation that an election, in particular the general election with its impressive symbolic importance and attention, is more likely to attract the attention of individuals and mobilize those individuals to partake in political action especially with regard to political discussion and status updates in Facebook.

Figure 1.

Frequency of political activity during the primary and general elections.

Turning to our regression analysis, using standard ordinary least squares (OLS), we regress a set of standard predictors, including those associated with two theories of participation, on our political activity index (Table 2) and on a select group of specific activities (Table 3), including voting, donations, wall discussion, and time-based¹¹ acts (our activity index minus the preceding three acts). Table 2 presents two different models, one for the primary election and one for the general election to compare differences that the predictors might have for the different elections. Again, it is our purpose to assess the nature of the impact that standard predictors associated with participation have with regard to participation in Facebook. Specifically, we compare those associated with the resource model against those of network size à la Putman. The results demonstrate that by and large many of the variables considered usual suspects and likely to contribute to political participation do not carry the same relationship to political activity in a Facebook environment. In fact, in Table 2, two of the “big three” standard predictors of political activity (sex and parental income) do not prove to be significantly related to political activity in Facebook and race is only significant for the primary election. Overall, our model explains approximately 47 and 63 percent of the variance in political activity during the 2008 primary and general elections, respectively.

Table 2.

Regression Model of Political Participation among Facebook Users during 2008 Election.

	Election
	Primary	General
Variable	Coefficient (SE)	Coefficient (SE)
Constant	0.67 (1.18)	0.24 (4.09)
Sex (male)	0.09 (0.21)	−1.31 (0.72)
Age	−0.09 (0.10)	0.36 (0.36)
Parental income (low to high)	0.05 (0.07)	0.25 (0.28)
Race (white)	−0.45* (0.21)	−0.40 (0.77)
Political ideology (conservative high)	0.29 (0.16)	−0.60 (0.78)
Political interest (interest high)	0.86*** (0.20)	3.45* (1.36)
GPA	−0.05 (0.28)	0.26 (0.89)
SAT (verbal)	0.09 (0.08)	0.24 (0.38)
SAT (math)	−0.18** (0.07)	−0.12 (0.38)
Number of friends	−.005 (0.12)	−0.17 (0.37)
Group membership	−0.36* (0.18)	−0.26 (0.59)
	R² = .47	R² = .63

p ≤ .05. **p ≤ .01. ***p ≤ .001.

Table 3.

Regression Model of Political Activities in Facebook during 2008 Election (Pooled).

	Time-based acts	Voted	Political contributions	Political discussion (wall posts)
Variable	Coefficient (SE)	Coefficient (SE)	Coefficient (SE)	Coefficient (SE)
Constant	−0.71 (1.11)	−0.28 (0.30)	−68.53 (328.43)	0.20 (1.16)
Sex (male)	0.50* (0.21)	0.10 (0.06)	−62.80 (62.83)	0.37 (0.22)
Age	0.09 (0.10)	0.01 (0.03)	31.19 (29.43)	0.11 (0.10)
Parental income (low to high)	0.17* (0.07)	0.01 (0.02)	23.30 (21.81)	−0.10 (0.08)
Race (white)	−0.33 (0.21)	−0.06 (0.06)	−97.06 (61.55)	−0.04 (0.22)
Political ideology (conservative high)	−0.01 (0.17)	0.00 (0.05)	49.41 (49.70)	0.02 (0.18)
Political interest (interest high)	0.76*** (0.23)	0.05 (0.06)	−8.85 (67.69)	0.90*** (0.24)
GPA	−0.20 (0.25)	−0.01 (0.07)	9.65 (74.44)	−0.41 (0.26)
SAT (verbal)	0.04 (0.08)	−0.01 (0.02)	31.13 (24.82)	0.02 (0.09)
SAT (math)	0.02 (0.08)	0.01 (0.02)	−38.76 (23.14)	0.08 (0.08)
Number of friends	0.06 (0.11)	0.04 (0.03)	−6.43 (30.93)	0.05 (0.11)
Group membership	0.15 (0.14)	0.05 (0.04)	−19.41 (41.31)	0.25 (0.15)
	R² = .35	R² = .13	R² = .13	R² = .36

p ≤ .05. **p ≤ .01. ***p ≤ .001.

The resource model of political participation predicts that income and interest drive political participation. Our findings are mixed. In Table 2, parental income, while in the expected direction, lacks statistical significance in our models. Nevertheless, the impact of political interest proves positive and significant so that those Facebook users who are more interested in politics are more likely to participate via Facebook during the primary and general elections (β = .86 and p = .000, β = 3.45 and p = .02, respectively). Overall, political interest has the strongest impact of all the predictors. In addition, it is worth including race with our discussion of the resource model as race is often correlated with (SES) factors that drive participation. Our results demonstrate that for the primary election, nonwhites are more likely to be politically active on Facebook than are whites (β = −.45 and p = .04). The effect of race is insignificant for the general election, however.

In considering Putnam’s model, we predict a positive and significant relationship between network size (number of friends and group membership) and political participation. We find that the number of Facebook “friends” a user has is not significantly related to political participation, contradicting our expectation on the effect of network size. Moreover, our results demonstrate the direction of the relationship to be opposite of our expectations so that those with more friends are less likely to be politically active. This suggests that collecting friends or building one’s social network in Facebook is an independent activity undertaken by users who are less inclined to be politically engaged with that network. While group membership fails to reach statistical significance for the general election, it does prove statistically significant for the primary election. However, the direction of the effect is also opposite of what we predicted. That is, the results demonstrate that the more Facebook groups to which a member belongs, the less likely the person is to participate politically in Facebook. SNS “friends” and network do not seem to offer the same sorts of benefits that real-life friends do in terms of developing the type of social capital needed to nurture political engagement. Our results demonstrate that perhaps SNS such as Facebook, in line with the findings of Ellison, Steinfield, and Lampe (2007), generate bridging capital rather than the bonding capital explained in Putnam’s theory of social capital. Like resource theory, our support for Putnam’s theory is mixed.

The final variable that proves significant is SAT (math) and we see that for the primary election the overall impact is slight and the direction is negative thereby demonstrating that those who perform better in math standardized tests are less likely to be politically active. We attribute this finding to the fact that SAT math (and verbal), like college major, offers a quick and dirty categorization into which we can divide college students, which, as a qualitative difference, might accompany different majors (Jennings 1993). In line with Jennings’ findings, we infer that those who earned higher scores on SAT (m) are generally less politically active and therefore less likely interested in and mobilized by the primary election campaign as born out by our results.

In Table 3, we disaggregate our political activity index to include time-based activities, voting, political donations, and political discussion. Using standard OLS regression and pooled data, we find two things worthy of note. First, in terms of voting and political contributions, none of the predictors demonstrate statistical significance. While this is not altogether surprising in light of the results presented in Table 2 and discussed above, we had expected at least political interest to prove significant. We do not know whether lack of significance is the result of Facebook equalizing the participation playing field or due to the low number of users who actually participated in either activity. Second, we find that overall political interest is significantly associated with time-based activities and political discussion (β = .76 and p = .001, β = .90 and p = .000), respectively. As well, sex is significantly and positively related to time-based activities so that males are more likely to participate in time-based activities than are females (β = .50 and p = .02). While sex fails to reach significance for the other political activities, it just barely misses for voting and political discussion. We assume, based on the civic volunteerism model, a positive relationship between parental income and political donations, and while the relationship is positive, it fails to reach statistical significance. However, parental income is positively and significantly related to time-based activities (β = .17 and p = .02). It seems likely the benefits of parental income bestowed to one’s offspring during the formative years carries over into college with respect to political participation, in a more general sense. However, mom and dad’s money is their money and it seems that what is theirs has no significant influence on what their college-aged offspring has and can do with regard to political donations. While group membership is in the positive direction for time-based activities, voting, and political discussion, it fails to reach statistical significance in all three, although only barely misses for political discussion. In terms of the four models, the R² = .35, .13, .13, and .36 for time-based activities, voting, political donations, and political discussion, respectively.

Conclusion

For several years, scholars have attempted to better understand the relationship between online communication and political engagement. A corollary of the extant research is better understanding how contemporary new media technologies are appropriated by individuals for political purposes. Our research adds to this work in that we seek to identify how individuals politically engage themselves with social media, and specifically, Facebook. We seek to understand whether individuals engage in politics in unique ways in Facebook and/or if the correlates of traditional offline political engagement stand true in this new environment.

The 2008 presidential election provides the unique opportunity to empirically investigate the nature of political engagement in Facebook. First, because the 2008 election was the first election in which Facebook had existed as a thoroughly diffuse cultural medium. And second, by the 2008 presidential campaign, with more than one hundred million active users, Facebook was generating national attention from both major political parties, candidates, and national media outlets as a tool to mobilize active political engagement. What we have found, however, is that despite the enthusiasm surrounding Facebook, individuals in general engaged in limited political activity via Facebook during the 2008 presidential campaign cycle. Despite limited overall engagement, we confirm the general election acted as the driving force for individual engagement within Facebook, especially with regard to particular behaviors such as political discussion, political status updates, advertising one’s intention to vote or that they voted, and support for a candidate. Finally, we find that one’s political interests play a significant role in determining whether an individual is more or less engaged in Facebook during both the primary and general elections. The significance and strength of the effect of political interest is in line with previous research both in the area of traditional offline political activity and online behavior, in general. While many have considered whether the Internet can equalize access to information and politics, we find, as have others, that interest propels action. Those who are more interested are those who are more likely to be engaged and politically active.

Interestingly, our findings suggest that Facebook political engagement may be unique in that very few of the traditional predictors of offline political engagement instantiate themselves within the SNS space, with two notable exceptions. First, our analyses do show that the number of Facebook friends one has bears no relation on political participation, which in essence contradicts Putnam’s social capital premise, yet there may be reason to believe there are both practical and conceptual differences between “Facebook” friends and traditional friends. However, while the number of groups to which a member belongs does yield significance, its direction is contrary of what we expected. The second exception is that minorities, at least in the primary election, were more likely to participate politically than were whites. In addition, our results demonstrate that individuals with higher SAT math aptitude tend to be engaged significantly less. In and of itself, the finding might not be earth shattering, but when compared with Carlisle and Patton’s (2009) analyses of who maintains a Facebook profile and what Hillygus (2005) finds with regard to SAT scores, social science curriculum, and political engagement, we extrapolate that our findings support her civic education hypothesis, insofar as political participants who maintain Facebook user profiles exhibit higher verbal SAT aptitude and backgrounds in the social sciences.

Finally, and the most important takeaway, while many tend to view the Internet and new technologies as tools that can easily enrich and nourish civic and political life, research has mainly revealed that existing inequalities in real life are translated and carried over into online life creating a digital divide that exists as a result of the interplay of national, institutional, and individual characteristics. Our research reveals that some of the traditional predictors that create differentials in political engagement, most notably parental income, sex, and race/ethnicity, do not appear relevant in the Facebook context. And, if they do (e.g., race), they tend to benefit those who are generally less likely to participate (e.g., minorities). We find this to indicate that perhaps Facebook is leveling the playing field and allowing those who might lack the resources to participate in a conventional sense, the ability to participate in a digital sense. Nevertheless, we are slightly concerned that there exist significant demographic and individual-level differences between those who use Facebook and those who do not (especially with regard to parental income, ethnicity, and SAT (verbal); Carlisle and Patton 2009). It is important to note that while there does exist a slight divide related to parental income, it does not negatively affect those in the lowest parental income brackets, but rather those students in the middle-income bracket. Still, we see that in general more students from the highest parental income category are Facebook users. This leveling of the playing field in Facebook political engagement may likely be the result of two factors. The first is that the university provides each registered student with an e-mail address and Internet access, and second, that Facebook is a free networking service where the only barrier to access is an e-mail address and Internet access.

It is our belief that continued research into the political life of Facebook users is relevant and important, especially as the technology evolves and more users join Facebook, which currently reports more than one billion active users (Facebook 2013). As citizens, campaign strategists and candidates become more familiar with how to use Facebook as a tool for political engagement, it is likely that Facebook will become more embedded into the political landscape. This is clearly the case as seen since the 2008 presidential election, Facebook has continued to be a presence for political activity in the 2010 midterm and 2012 presidential elections. If and as the ranks of active political Facebook users become more established, in the long term, we firmly anticipate Facebook interaction to be a standard measure of online political activity. To what extent citizens will ultimately engage with each other using Facebook or another new technology and how that differs from their offline engagement and the political system is an open question for future research.

Footnotes

Acknowledgements

We would like to acknowledge Drs. Bruce Bimber, Richard Mayer, Kevin Almeroth, and Dorothy Chun at the University of California, Santa Barbara, for graciously providing access to the data for this research—data that were derived from their Andrew W. Mellon Foundation funded research grant titled “Assessing the Pedagogic Implications of Technology in the Classroom.” We are also grateful for the assistance of Whitney Richardson and Ryan Sargent who carried out the content analysis component of our project. Thank you to Eric R. A. N. Smith, M. Kent Jennings, Bruce Bimber, and Anand Sokhey for providing feedback and assistance with previous drafts.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Financial support for the most recent data collection used here was funded by a seed grant from National Science Foundation (NSF) SBE-0620073, ADVANCE grant awarded to Idaho State University.

Notes

References

Abramson

Paul

Aldrich

John

Rohde

David

. 2006. Change and Continuity in the 2004 Elections. Washington, DC: Congressional Quarterly Press.

Benkler

Yochai

. 2006. The Wealth of Networks: How Social Production Transforms Markets and Freedom. Boston: Yale University Press.

Best

Samuel J.

Krueger

Brian S.

2005. “Analyzing the Representativeness of Internet Political Participation.” Political Behavior 27 (2): 183–216.

Bimber

Bruce

. 1999. “The Internet and Citizen Communication with Government: Does the Medium Matter?” Political Communication 16 (4): 409–29.

Bimber

Bruce

. 2001. “Information and Political Engagement in America: The Search for Effects of Information Technology at the Individual Level.” Political Research Quarterly 54 (1): 53–67.

Bimber

Bruce

. 2003. Information and American Democracy: Technology in the Evolution of Political Power. Cambridge: Cambridge University Press.

Bimber

Bruce

Davis

Richard

. 2003. Campaigning Online: The Internet in U.S. Elections. New York: Oxford University Press.

Bonchek

Mark

. 1995. “Grassroots in Cyberspace: Recruiting Members on the Internet or Do Computer Networks Facilitate Collective Action? A Transaction Cost Approach.” Paper presented at the 53rd Annual Meeting of the Midwest Political Science Association, Chicago, IL, April 6–8.

Bonchek

Mark

. 1997. From Broadcast to Netcast: The Internet and the Flow of Political Information, PhD diss., Harvard University.

10.

Bond

Robert M.

Fariss

Christopher J.

Jones

Jason J.

Kramer

Adam D.I.

Marlow

Cameron

Settle

Jaime E.

Fowler

James H.

2012. “A 61-million-person experiment in social influence and political mobilization.” Nature 489:295-298.

11.

Boulianne

Shelley

. 2009. “Does Internet Use Affect Engagement? A Meta-analysis of Research.” Political Communication 26 (2): 193–211.

12.

Bourdieu

Pierre

Wacquant

Loïc

. 1992. An Invitation to Reflexive Sociology. Chicago: University of Chicago Press.

13.

Brady

Henry E.

Verba

Sidney

Schlozman

Kay L.

1995. “Beyond SES: A Resource Model of Political Participation.” The American Political Science Review 89 (2): 271–94.

14.

Campbell

Angus

. 1960. “Surge and Decline: A Study of Electoral Change.” In Elections and the Political Order, edited by Campbell

Angus

Converse

Philip

Miller, and

Warren

Stokes

Donald

, 40–62. New York: Wiley.

15.

Carlisle

Juliet E.

Patton

Rob

. 2009. “Facebook and Political Engagement During the 2008 Presidential Campaign.” Paper delivered at the 104th Annual Meeting of the American Political Science Association. Toronto, Ontario, Canada, September 3–6.

16.

Davis

Richard

. 1999. The Web of Politics. Oxford: Oxford University Press.

17.

Delli Carpini

Michael

. 2000. “Gen.com: Youth, Civic Engagement, and the New Information Environment.” Political Communication 17 (4): 341–49.

18.

Drew

Dan

Weaver

David

. 2006. “Voter Learning in The 2004 Presidential Election: Did The Media Matter?” Journalism & Mass Communication Quarterly 83 (1): 25–42.

19.

Ellison

Nicole

Steinfield

Charles

Lampe

Cliff

. 2006. “Spatially Bounded Online Social Networks and Social Capital: The Role of Facebook.” Paper presented at the Annual Meeting of the International Communication Association, Dresden, Germany, 2006. https://www.msu.edu/~nellison/Facebook_ICA_2006.pdf.

20.

Ellison

Nicole

Steinfield

Charles

Lampe

Cliff

. 2007. “The Benefits of Facebook ‘Friends’: Social Capital and College Students’ Use of Online Social Network Sites.” Journal of Computer-Mediated Communication 12 (4): Article 1. http://jcmc.indiana.edu/vol12/issue4/ellison.html.

21.

Facebook. 2009a. “Facebook Announcement: Facebook/ABC News Election 08.” http://www.facebook.com/press/releases.php?p=12926.

22.

Facebook. 2009b. “Facebook Timeline.” http://newsroom.fb.com/Timeline.

23.

Facebook. 2010. “Facebook Timeline.” http://newsroom.fb.com/Timeline.

24.

Facebook. 2012. “The 2012 Election Day through the Facebook Lens.” http://www.facebook.com/notes/facebook-data-science/the-2012-election-day-through-the-facebook-lens/10151181043778859.

25.

Facebook. 2013. “Facebook Timeline.” http://newsroom.fb.com/Timeline.

26.

Foot

Kristen

Schneider

Steven

. 2006. Web Campaigning. Cambridge: MIT Press.

27.

Hess

Robert

Torney

Judith

. 1967. The Development of Political Attitudes in Children. Chicago: Aldine.

28.

Hill

Kevin

Hughes

John

. 1998. Cyberpolitics: Citizen Activism in the Age of the Internet. Lanham: Rowman & Littlefield.

29.

Hillygus

D.Sunshine

. 2005. “The Missing Link: Exploring the Relationship between Higher Education and Political Engagement.” Political Behavior 27 (1): 25–47.

30.

Huckfeldt

Robert

Sprague

John

. 1995. Citizens, Politics and Social Communication: Information and Influence in an Election Campaign. New York: Cambridge University Press.

31.

Jenkins

Henry

. 2006 Convergence Culture: Where Old and New Media Collide. Cambridge: MIT Press.

32.

Jennings

M.Kent

. 1993. “Education and Political Development among Young Adults.” Politics and the Individual 3 (2): 1–24.

33.

Jennings

M.Kent

Zeitner

Vicki

. 2003. “Internet Use and Civic Engagement: A Longitudinal Analysis.” Public Opinion Quarterly 67 (3): 311–34.

34.

Johnson

Thomas J.

Kaye

Barbara K.

2003. “A Boost or a Bust for Democracy? How the Web Influenced Political Attitudes and Behaviors in the 1996 and 2000 Presidential Elections.” Harvard International Journal of Press/Politics 8 (3): 9–34.

35.

Katz

James E.

Rice

Ronald

. 2002. Social Consequences of Internet Use: Access, Involvement and Interaction. Cambridge: MIT Press.

36.

Kwak

Nojin

Williams

Ann E.

Wang

Xiarou

Lee

Hoon

. 2005. “Talking Politics and Engaging Politics: An Examination of the Interactive Relationships between Structural Features of Political Talk and Discussion Engagement.” Communication Research 32 (1): 88–111.

37.

Leighley

Jan

. 1990. “Social Interaction and Contextual Influences on Political Participation.” American Politics Quarterly 18:459–75.

38.

Leizerov

Sagi

. 2000. “Privacy Advocacy Groups Versus Intel: A Case Study of How Social Movements are Tactically Using the Internet to Fight Corporations.” Social Science Computer Review 18 (4): 461–83.

39.

Loader

Brian

. 2007. Young Citizens in the Digital Age: Political Engagement, Young People and New Media. London: Routledge.

40.

Luskin

Robert C.

1990. “Explaining Political Sophistication.” Political Behavior 12 (4): 331–61.

41.

Nachmias

David

Nachmias

Chava

. 1987. Research Methods in the Social Sciences. 3rd ed. New York: St. Martin’s Press.

42.

Neuendorf

Kimberly

. 2002. The Content Analysis Guidebook. Thousand Oaks: SAGE.

43.

Niemi

Richard

Junn

Jane

. 1998. Civic Education: What Makes Students Learn. New Haven: Yale University Press.

44.

Nisbet

Matthew C.

Scheufele

Dietram A.

2004. “Political Talk as a Catalyst for Online Citizenship.” Journalism & Mass Communication Quarterly 81 (4): 877–96.

45.

Norris

Pippa

. 1999. “Who Surfs? New Technology, Old Voters and Virtual Democracy.” In democracy.com , edited by Kamarck

E. C.

Nye

J. S.

, 71–94. New Hampshire: Hollis.

46.

Norris

Pippa

. 2000. A Virtuous Circle: Political Communications in Postindustrial Societies. Cambridge: Cambridge University Press.

47.

Norris

Pippa

. 2004. “The Bridging and Bonding Role of Online Communities.” In Society Online: The Internet in Context, edited by Howard

P. N.

Jones

, 31–41. Thousand Oaks: SAGE.

48.

Pew. 2011. “22% of Online Americans Used Social Networking or Twitter for Politics in 2010 Campaign.” http://pewinternet.org/~/media//Files/Reports/2011/PIP-Social-Media-and-2010-Election.pdf.

49.

Putnam

Robert

. 1993. Making Democracy Work: Civic Traditions in Modern Italy. Princeton: Princeton University Press.

50.

Putnam

Robert

. 1995a. “Bowling Alone: America’s Declining Social Capital.” Journal of Democracy 6 (1): 65–78.

51.

Putnam

Robert

. 1995b. “Tuning In, Tuning Out: The Strange Disappearance of Social Capital in America.” Political Science and Politics 28 (4): 664–83.

52.

Putnam

Robert

. 2000. Bowling Alone: The Collapse and Revival of American Community. New York: Simon & Schuster.

53.

Resnick

Paul

. 2004. “Impersonal Sociotechnical Capital, ICTs, and Collective Action among Strangers.” In Transforming Enterprise, edited by Dutton

Kahin

O’Callaghan, and

Wyckoff

, 486–99. Boston: MIT Press.

54.

Rheingold

Howard

. 2000. The Virtual Community: Homesteading on the Electronic Frontier. Rev. ed. Cambridge: MIT Press.

55.

Rheingold

Howard

. 2002. Smart Mobs: The Next Social Revolution. Cambridge: Perseus Publishing.

56.

Schlozman

Kay L.

Burns

Nancy

Verba

Sidney

Donahue

Jesse

. 1995. “Gender and Citizen Participation: Is There a Different Voice?” American Journal of Political Science 39 (2): 267–93.

57.

Schuler

Douglas

Day

Peter

. 2004. Shaping the Network Society: The New Role of Civil Society in Cyberspace. Cambridge: MIT Press.

58.

Shah

Dhavan V.

Cho

Jaeho

Eveland

William P.

Kwak

Nojin

. 2005. “Information and Expression in a Digital Age: Modeling Internet Effects on Civic Participation.” Communication Research 32 (5): 531–65.

59.

Shah

Dhavan V.

Cho

Jaeho

Nah

Seungahn

Gotlieb

Melissa R.

Hwang

Hyunseo

Lee

Nam-Jin

School

Rosanne M.

McLeod

Douglas M.

2007. “Campaign Ads, Online Messaging, and Participation: Extending the Communication Mediation Model.” Journal of Communication 57 (4): 676–703.

60.

Shah

Dhavan V.

Kwak

Nojin

Holbert

R. Lance

. 2001. “Connecting and Disconnecting with Civic Life: Patterns of Internet Use and the Production of Social Capital.” Political Communication 18:141–62.

61.

Smith

Mark

Kollock

Peter

. eds. 1999. Communities in Cyberspace: Perspectives on New Forms of Social Organization. London: Routledge.

62.

Strate

John M.

Parrish

Charles J.

Elder

Charles D.

Ford

I. Coit

1989. “Life Span Civic Development and Voting Participation.” American Political Science Review 83:443–64.

63.

Tolbert

Caroline

McNeal

Ramona

. 2003. “Unraveling the Effects of the Internet on Political Participation.” Political Research Quarterly 56 (2): 175–85.

64.

Vanden Boogart

Matthew

. 2006. Uncovering the Social Impacts of Facebook on a College Campus, Master’s thesis, Kansas State University. http://krex.k-state.edu/dspace/handle/2097/181.

65.

Verba

Sidney

Nie

Norman

. 1987. Participation in America: Political Democracy and Social Equality. Chicago: The University of Chicago Press.

66.

Verba

Sidney

Schlozman

Kay

Brady

Henry

. 1995. Voice and Equality: Civic Voluntarism in American Politics. Cambridge: Harvard University Press.

67.

Wellman

Barry

Haythornthwaite

Caroline

. eds. 2002. The Internet in Everyday Life. Oxford: Blackwell Publishing.

68.

Wolfinger

Raymond

Rosenstone

Steven

. 1980. Who Votes? New Haven: Yale University Press.

69.

Xenos

Micahel

Moy

Patricia

. 2007. “Direct and Differential Effects of the Internet on Political and Civic Engagement.” Journal of Communication 57 (4): 704–18.

70.

Zukin

Cliff

Keeter

Scott

Andolina

Molly

Jenkins

Krista

Carpini

Michael Delli

. 2006. A New Engagement?: Political Participation, Civic Life, and the Changing American Citizen. New York: Oxford University Press.