Abstract
There is much concern about the long-tail of online political information that now competes with standard news content. Yet we know little about how successfully it does so, and why. From the 2008 US presidential campaign, I develop a rich, original meta-database to map the relationships between the top viral election videos (278 million views) and the known-universe of blogs that linked to them (13,000 links). Through a mixed-method analysis of both content and web metrics, I find that news videos compete with political, entertainment and citizen videos; journalism blogs also compete with political, entertainment, and lifestyle blogs to propagate these new genres of political information. By further analyzing the gatekeeping mechanisms between video and blog categories, I find that online audiences’ civic information preferences help explain the rise of these new participatory political information genres, as well as the ways in which legacy media are struggling to adapt.
Introduction
A decade of news research indicates that mainstream media teeter on the edge of relevance to average American citizens. Some even hold the news media accountable, blaming them for alienating citizens with a ‘monologue that speaks in the ostensibly factual voice of the real world’ (Soffer, 2009: 473). Meanwhile, the digital media landscape has evolved, introducing new formats of political information. With a greater variety of available content, citizens are no longer forced to rely on news media to learn about public life. The growth of online communities of activists and citizens are important sources of these shifts. As a result of growing competition in the information market, journalists no longer have a monopoly in producing or brokering civic content. Furthermore, looking at the online media ecology, genres of civic and political information have expanded well beyond the domain of what is traditionally seen as the news.
One example of this is the phenomenal growth in online viral political information. This important trend is visible in both domestic (e.g. during major political events, such as elections) and international (e.g. during major political crises, such as protests) contexts. The most common format of viral political information propagation is through videos that reach millions of viewers. Sometimes the total viewership is comparable to or surpasses that of network news shows. For example, by 4 November 2008, the ‘Yes We Can’ music video had millions more total visitors than the daily listenership/viewership of NPR’s Morning Edition, ABC’s World News Tonight, and Fox News’ O’Reilly Factor (Pew Research Center, 2006). Unfortunately, many news organizations cover them after the fact as novelty items, instead of as significant political content that has interested audiences. Understanding this category of political information entails asking several broad research questions: What are the most popular genres of viral political content? What role do journalists play in mediating these new types of public information? How might answering these questions help us understand online audiences’ information needs and preferences? Addressing some of these questions might help us identify gaps between audiences’ civic information preferences and the journalistic values of the press.
I seek to address these questions in the context of the 2008 presidential elections. Online viral videos were a novel feature during the campaign season, and both major candidates’ campaign videos generated over 180 million views on YouTube alone. In addition to campaign and news videos, this study also examines video content generated by citizen and entertainment groups. By identifying the new genres of online political information, I compare the major differences between these and traditional news content. Furthermore, it is not enough to examine these videos alone. To explore the relationship between the availability of distinct information genres and audiences’ unique information preferences, I also analyze which types of blogs linked to which types of videos. Doing so, I identify the civic information preferences of unique online publics. I find important differences between journalists’ blogs, and political, lifestyle, and entertainment bloggers. To make sense of these findings, I draw on literature from citizenship studies to argue that the mainstream news media are out of touch with the contemporary information needs of their online audiences. To further illustrate this analytic framework, it is useful to first understand how conventional news habits and information environments have evolved over the past decade.
Declining trust, credibility, and use of traditional news
The public now relies less on traditional news than ever before. Only 35 percent of Americans read the news daily, with younger cohorts the least likely to do so (only 16% for 18 to 30 year-olds, and less than 10% for 12 to 17 year-olds). Television and radio news use is similarly spiraling downward (Patterson, 2007), and nearly 50 percent of 19 to 39 year-olds would not care if their local paper closed (Kohut and Remez, 2009). Perceived levels of trust and credibility in news also parallel the general decline in news use, where only 18 percent of Americans believe that the press deals fairly with all sides, and only 29 percent think journalists get their facts straight (Kohut, 2009). Thus, the crisis in journalism today has at its core an audience problem: citizens are turning away from journalists and looking elsewhere to get political information.
But some also wax optimistic about the growth in online news use. Americans do increasingly turn to the internet for news (Patterson, 2007), with 40 percent of Americans now going online regularly for news (Pew Research Center, 2009). The internet has also bypassed newspapers as the second most important source of national and international news (Kohut and Remez, 2008). But hopes that the internet might reinvigorate audience interest in news seem misplaced at best. Some or much of this growth in online news use may only be complementary to the already news-reading publics’ offline consumption habits (Ahlers, 2006; Dutta-Bergman, 2004). Indeed, today’s participatory news consumers get most of their information from no less than four to six media platforms a day, with only 7 percent of Americans still receiving news from a single platform (Purcell et al., 2010). Additionally, the problems of trust and credibility continue to persist online, with many Americans feeling that online news is no more informative (53%), reliable (77%), or trustworthy (84%) than offline equivalents (Ahlers, 2006). Given their growing options and user-selectivity in online news consumption, audiences are less likely to choose public affairs content when given the choice (Prior, 2005; Tewksbury, 2003; Tremayne et al., 2007).
If citizens are increasingly turning away from standard news content, which genres of online political information might they find more engaging or trustworthy? Who is creating these new genres of public information? How are they being propagated, and who are their gatekeepers?
New genres of political information
Whereas traditional news audiences continue to dwindle, demand for unfiltered content like campaign speeches and transcripts has grown rapidly, especially for campaigns’ online video content. Online video sharing is a key venue for accessing and interacting with news and political information (Madden, 2007). In 2006, 30 percent of internet users viewed online videos, doubling to over 60 percent within one year (Madden, 2009). During the 2008 presidential elections, 23 percent of Americans received campaign information directly from candidates’ emails, 35 percent from online campaign videos, and 39 percent directly accessed primary documents and materials, such as position papers and speech transcripts (Smith and Rainie, 2008).
For more evidence of the importance of direct-to-citizen communication by campaigns and political elites, consider this anecdote: direct video usage by citizens in the 2008 presidential elections was vital to the Obama campaign’s online strategy, which posted 1800+ separate video clips on YouTube and generated over one billion minutes of total viewership by 4 November 2008 (Delany, 2009). Much of this video coverage included full-length speeches uninterrupted by journalists or media pundits, and many campaign videos went viral, only to be picked up by mainstream media after their popularity had already peaked. Direct and unfiltered political information from campaigns and candidates, then, were an important and engaging category of online political information for many, especially during the 2008 presidential campaign.
Another sign of the changing information landscape is the growth and attraction of political satire and comedic content, for which audience sizes have also begun to compete with the mainstream news media (Baum, 2003; Prior, 2003, 2005). For example, 21 percent of 18 to 30 year-olds learned about candidates in 2004 via Saturday Night Live and The Daily Show programming (Kohut, 2004). Soft news and infotainment tend to also be loaded with political information (Niven et al., 2003), and in 2004 the quantity of substantive information in The Daily Show with Jon Stewart and broadcast network news was surprisingly equal (Fox et al., 2007).
Interest in alternative genres of political information was especially visible during the election seasons of 2004, 2006, and 2008. Digital media provided audiences with content beyond news and entertainment organizations, incorporating official political campaigns, and even amateur user-generated productions, among others. YouTube and similar video-sharing sites were prime locations to observe the influx of new information genres in election campaigns because (a) they were important sites for producing and distributing grassroots electoral content, (b) they functioned as an ‘archive where amateur curators scan the media environment, searching for meaningful bits of content’, to bring to a larger public, and (c) they functioned in relation to a host of other popular social networks constituting the networked public sphere (Jenkins, 2006: 274–275).
During the 2008 election, viral videos were heavily discussed and shared throughout the election blogosphere, accumulating millions of views. These videos brought a broader range of everyday audience members into the fold of formal politics. They did so by introducing newer genres of pro-political content that mixed entertainment, comedy, and parody through media manipulation and remixing (e.g. mashups). Viral videos also included vast amounts of direct and unmediated content from the campaign channels discussed earlier. Election viral videos, then, provide a fertile setting to explore these new genres of political information.
New categories of information distributors
The ability of alternative online actors to co-produce and co-distribute content alongside journalists has probably diminished the monopoly of the press over brokering political information. This is especially important in the context of election viral videos because it increases the capacity for alternative actors (e.g. candidates, entertainers, and even citizens) to successfully compete with news media. Benkler (2006) credits the decreased costs of producing and exchanging information over distributed networks, allowing non-market information producers to participate more fully.
Personalized communication logics in these public networks seem related to scholarship that points to changing norms of civic participation in modern democracies that favor individualized styles of political engagement (Bennett, 2008; Coleman and Blumler, 2009; Zukin et al., 2006). Newer generations of citizens are more inclined to emphasize individual voice as a key preference for engagement in public life, making innovative uses of social media technologies to do so (e.g. blogs, open-source software, wikis, social networking services, podcasts, social bookmarking, digital petitions, and even hactivism campaigns) (Levine, 2007: 94–95). Dahlgren (2009) refers to this as the ‘individualization of civic cultures’, while Shirky (2008) calls it the ‘mass amateurization’ of the public sphere. Regardless of terminological differences, these perspectives concede that the networked public sphere has diversified beyond the once powerful role of the press in gatekeeping public information. It further suggests that digital media environments are more receptive to nurturing citizens’ personalized engagement styles than mass media.
Recent US elections exemplify how this process is being supported by the arrival of newer genres of information distributors. In 2004, online networks mobilized through the Howard Dean campaign’s use of meetup.com and the political blogosphere showcased the capacity of citizens to self-organize quickly and loosely. The 2006 congressional elections saw the influx of online political videos as a defining feature of modern campaigning in the United States, particularly George Allen’s videotaped ‘macaca moment’. 1 The 2008 presidential election brought these spheres together in the form of hundreds of viral election videos produced by political campaigns (e.g. barackobama.com’s ‘A More Perfect Union’), news organizations (e.g. Al Jazeera’s ‘Misconceptions of Obama fuel Republican campaign’), entertainment companies (e.g. barelypolitical.com’s ‘I’ve Got a Crush on Obama’), as well as grassroots citizens (e.g. Iraq War veteran Joe Cook’s ‘Dear Mr. Obama’). The election blogosphere helped incorporate multiple domains of actors from proto-political, journalism, advocacy, activist, and e-government sectors into the mix (Dahlgren, 2009: 166–168), many of who were responsible for producing and distributing viral videos.
The election blogosphere also includes journalists. Many news organizations incorporate journalism blogs as an important component of their online repertoire (Tuchman and Ostertag, 2008). These blogs allow readers to interact directly with journalists, and even post their own views and experiences on official news sites. Thus, journalism blogs hosted on news organizations’ sites can alter the relationship between journalists and readers (Robinson, 2006). Robinson’s documentation of journalism blogs (i.e. blogs officially sanctioned by news organizations) finds a notably different narrative form and evolving set of values when compared to most news productions. It is perhaps one of the few areas in which journalism organizations are actively renegotiating their norms and repositioning themselves within the networked public sphere. This evolution could have a lasting impact on the future orientation of traditional news organizations. Observing them interact with newer information genres, such as viral videos, might reveal their informational orientations, and whether these orientations are significantly different from other types of online publics.
Political blogs have also received significant attention, often described as a ‘fifth column’ acting as watchdogs of politicians and the elite media (Farrell and Drezner, 2008). While independent from journalistic organizations, they seem to be in a symbiotic relationship with mainstream news media. Some political bloggers are so popular that they’ve been invited to mainstream news outlets or frequently appear there as commentators, such as Andrew Sullivan (The Daily Dish), Glenn Reynolds (InstaPundit), and Joshua Marshall (Talking Points Memo). Messner and DiStaso (2008) have described this as a news source cycle, or the passing back and forth of information between each other. Reese et al. (2007) find similar: political blogs and traditional journalists had a complementary relationship and linked to each other heavily. Kenix’s (2009) study also supports this relationship, finding that elite political blogs rely heavily on mainstream news media. Although elite political blogs are distinct from journalism blogs, these investigations suggest that they have a more traditional connection to the mainstream news media than previously assumed.
Beyond journalism and elite political blogs, scholarship on alternative gatekeepers (actors that now have the power to influence which content receives attention) and distributors is relatively nascent. Much of the research on blogging and politics has examined the most powerful gatekeepers (Adamic and Glance, 2005). Unfortunately, non-elite blogs run by everyday citizens, both political and general, have not been explored systematically. Little is known about the culture and informational preferences of this long-tail of the blogosphere. It is true that most online content is non-political (Hindman, 2009), but looking beyond elite journalism and political blogs is important because digital media is a critical space for nurturing new forms of political information and engagement preferences (Bennett et al., 2009). So the domain of what is political and not political may be changing as citizens redefine politics. This long-tail of the blogosphere could differ from the practices of elite political blogs and official journalism blogs. Therefore, looking at the long-tail of blogs represents an important uninvestigated space for further research.
The area of blogging poses an interesting site to study the shifts of information norms precisely because it straddles the line between old journalistic news values and new citizen information preferences. By using the 2008 presidential election as a case study, observing how distinct sectors of the election blogosphere responded to different types of digital election content can illustrate the new sources and brokers of information that journalists now compete with. Therefore, I present the following research questions:
What genres of civic content were introduced through election viral videos? Although we know that there were a large number of heavily viewed election videos in 2008, we don’t have a scheme to classify them.
Which categories of blogs linked to and distributed election viral videos? Scholars have observed that elite political blogs and journalism blogs play an active role in sharing such content with the online public. However, studies have yet to show the overall distribution of other actors engaging with online political information distribution.
Which categories of blogs preferred what genres of viral election videos? It is possible, for example, that journalism blogs linked to information genres that most resemble their produce: news videos. Perhaps other categories of blogs favored content that reflected and satisfied other preferences. Investigating these dynamics is especially important for us to better understand the understudied distribution of non-elite long-tail blogs.
Collectively addressing these research questions allow us to map the ecology of viral election videos, their engagement by different categories of blogs linking to them, and to identify potential gaps between journalists and their audiences. By doing so, we might better understand the shifting information norms and preferences of online citizens.
Method
Not every online video is necessarily viral, even if it spreads rapidly through a small sociotechnical system; neither is this necessarily the case if a video receives millions of views, but only with the help of broadcast media (Nahon et al., 2011). For the purposes of this study, I define an ‘election viral video’ as a highly viewed online video with content related to the 2008 election, uploaded on a participatory video platform (e.g. YouTube), and distributed heavily through social media (e.g. blogs, SNS). To approach this empirically, this study draws its data from an Election 2008 metadata archive constructed through relational datasets combining a sample of the top viral election videos (n = 120) and the known-universe of blogs (n = 9556) linking to these videos during the 2008 election season.
Viral videos were selected through a sampling strategy employing viralvideochart.com to identify all election-related videos between January 2008 and January 2009 (ending with the inauguration of Barak Obama). Viral Video Chart offered many unique advantages over other similar services. It used a sophisticated tracking system to cover all known platforms of online video sharing during the election season. The tracking system was operational since the beginning of the campaign season, and allowed for longitudinal consistency. It also provided daily lists of viral videos in multiple categories culled from scraping multiple spheres of online political activity (blogs, social networking sites, and video-sharing platforms). Real-time information about videos was also tracked and analyzed. Other services, such as break.com, collegehumor.com, failblog.com, funnyordie.com, focused on specific types of viral content, none of which covered election information in a consistent and reliable manner. The final video archive of 120 election videos was started on 20 January 2009 (inauguration day) to include viral videos that were active through the election day of 4 November 2008. Drawing on the top 100 viral videos over the preceding year from three categories (political, election, and general), duplicate videos were identified and removed. Out of a total 300 top videos, 120 were election related and covered the full year of viral video activity between January 2008 and 2009.
Additionally, I constructed a separate dataset tracking all known blogs linking to these 120 videos. The most popular YouTube URL for each of the 120 viral videos was identified to track blog links because the most popular links to all videos originated from YouTube.com. Technorati and Google Blog Search were used to reverse-crawl videos being linked by blogs. To do so, scripts were written to automatically harvest all blog posts linking to each viral video on any given day. The data were cleaned to remove duplicate links, and to identify the blogs where the posts originated. This separate dataset resulted in 13,173 unique blog posts originating from 9556 unique blogs to at least one of the 120 top-election viral videos.
Lastly, because blog data and video data were stored in separate datasets, I connected them using a relational database to construct the Election 2008 metadata archive. Some blogs have multiple links to many videos, while the rest have at least a minimum of one link to at least one video, so the relational database distinguishes between blogs, and their links to election viral videos. The final database for this study constitutes various levels of metadata: the most important election videos between January 2008 and January 2009 (n = 120), their cumulative views recorded as of June 2009, the URLs of blog posts which linked to each of these videos, and the number of links from each blog to each video. Finally, all available video content of the 120 videos was also archived. I employ various elements of these metadata and content to address the research questions throughout the analyses.
The Election 2008 meta-database is in quantitative and code form of links and views, but my research questions required a more detailed understanding of the qualitative aspects of the blogs and videos. I first needed to develop categories and definitions to describe the video content and blog genres. I developed grounded categories and definitions by analyzing a random subset of videos and blogs. Four categories of blogs (journalism, political, lifestyle, and entertainment) and four categories of videos (news, campaign, citizen, and entertainment) were inductively identified. I also kept track of non-English and deleted blogs and videos. After video and blog categories were identified, a set of definitions was developed from concrete observations made during the grounded phase, along with a robust set of codes to identify nuances and variations within and across categories. Using the grounded codes, I trained six coders to categorize all viral videos (n = 120), and 10 percent of all blogs (n = 1000). Coders were trained to categorize according to ‘best fit’, and any blogs or videos that fit multiple categories were marked as ‘other’ with detailed notations collected by the coders. The final percent agreement between all six coders for the video coding was 91.7 percent, and 83.3 percent for the blog coding.
Results
Genres of viral political information
Out of 300 top videos covering the full year of viral video activity between January 2008 and 2009, 120 were election related – i.e. 40 percent of the most popular viral videos had political content about the election. It is within this context that we seek to understand new genres of political information that have become important sources of civic information for online viewers. The content distribution shows that the most popular election videos look significantly different from standard news content. Traditional news videos were the smallest in terms of the number of videos and cumulative views. Campaign, citizen, and entertainment videos were far more common and popular. Of the three popular genres, entertainment videos were most popular (in cumulative views), and campaign videos were most common (in total number of videos). I now describe these new genres of online political information in detail and argue that the information landscape may be expanding the definitions of news for online citizens (see Table 1).
Election 2008 viral video content distribution by genre
Campaign viral videos were the largest category (31%; n = 37) produced officially by campaigns, and presented candidates’ full-length speeches, uninterrupted televised election debates, and in-depth online campaign documentaries or traditional political advertisements. Examples include: ‘President-Elect Barack Obama in Chicago’ by the Obama YouTube channel, a full-length campaign speech by then-candidate Obama (5.4 million views); ‘Full Vice Presidential Debate with Gov. Palin and Sen. Biden’ by the CSPAN YouTube channel, a 1.5-hour footage of the vice-presidential debates (1.9 million views); ‘Keating Economics: John McCain & The Making of a Financial Crisis’ by the Obama YouTube channel, a 13-minute documentary attacking the McCain campaign (1.7 million views); and ‘Democrats Praising McCain’ by the McCain YouTube channel, a traditional attack advertisement about their opponents (0.3 million views). Overall, this genre of videos indicates that campaigns found online video to be an important venue for reaching constituents, and that citizens willingly sought out such content.
Citizen viral videos were the second largest group (24%; n = 29), and didn’t have institutional endorsements from political campaigns, the news media, or entertainment organizations (e.g. Saturday Night Live). Instead, these videos communicated viewpoints or concerns by unaffiliated but politically interested citizens about the election. Examples include: ‘McCain’s YouTube Problem Just Became a Nightmare’, a video juxtaposing clips of John McCain contradicting himself throughout various interviews (9.0 million views); ‘Vote Different’, a home-made anti-Hillary Clinton advertisement (6.1 million views); ‘Obama Citizenship’, an amateur documentary about then-candidate Obama’s citizenship status (4.7 million views); and ‘Dear Mr. Obama’, by an Iraq War veteran expressing concerns to then-candidate Obama (14 million views). These videos illustrate that online citizens were interactively engaging content that was produced by citizens like them, or what some may call a meta-deliberation taking place within a networked public sphere. These had strong ideological valence, and reflect a fertile market for citizen engagement often overlooked by news media.
Entertainment videos had entertainment-based values as the main focus, such as humor, popular culture, and satire (18%; n = 22). They generally made light of campaign coverage and electoral politics through musical, satirical, and comedic skits, and also featured home-made mashups of pop-cultural content (songs, films, etc.). Examples include: ‘I’m Voting Republican’, an ironic video with reasons not to vote for Republicans (4.7 million views); ‘Wassup 2008’, a popular beer commercial remake critiquing the Bush administration (8.2 million views); ‘Dear Mr. Obama – Barack Obama Music Video’, a music video produced by a popular musician with direct samples of then-candidate Obama’s official speeches (21 million views); and ‘Barack Obama Hillary Clinton – Umbrella’, a Saturday Night Live skit of the Democratic candidates (96.6 million views). Though entertainment in politics is not especially novel, these entertainment viral videos were unique in their diverse production values. They included high-end productions from major entertainment companies, and home-made amateur productions by everyday citizens. Their immense popularity illustrates that many citizens find value in light-hearted coverage of complex issues that could also provide rich analytic perspectives via creative rhetorical moves.
News videos constituted unedited clips from cable television or network news and represented the smallest category (15%; n = 18). This content included reports by journalists, and analysis and commentary on major news shows. Examples include: ‘CNN Laughs It Up Over Sarah Palin Interview’, an unedited clip from CNN coverage of the election (5.3 million views); ‘Misconceptions of Obama fuel Republican campaign’, an Al Jazeera English segment (2.4 million views); ‘Tucker Bounds on Palin’s National Security Experience’, an interview clip with journalist Campbell Brown (0.7 million views); and ‘Keith Olbermann: Special Comment on Hillary Clinton’, an MSNBC election segment (0.2 million views). These news video snippets were directly cut from cable and network news shows, and redistributed through YouTube. News videos tended to have the lowest distribution of cumulative views, unlike campaign, citizen, and entertainment videos. News videos were the least popular genre of viral political information.
Looking at the overall distribution of videos that fit each category is one way of measuring the importance of each genre of information. But this doesn’t account for variations in average cumulative views per video by genre. Most online content popularity is distributed non-parametrically, so differences in average popularity are also important. It is possible that a smaller category of videos might actually be more popular than a larger category. For example, relative to entertainment videos, citizen and campaign videos generated lower average cumulative views: 29 citizen videos accumulated 63 million views, and 37 campaign videos accumulated 42 million views. Even considering average views per video, both citizen and campaign videos were more popular than news videos: news videos were least present and popular (18 news videos generated a dismal 28 million views).
In regard to the landscape of online political information, the online public did not find news videos to be an engaging source of political information. It is true that people had the option of watching the news videos on television. However, the point remains that in the rapidly growing online information system, other genres, like entertainment, campaign, and citizen content, far outpace interest in standard news coverage.
Categories of blogs linking to viral election videos
I have illustrated the relative lack of online interest in viewing traditional news content, in part a result of significant differences and variations in the availability of newer genres of political information. An alternative explanation could be less content specific and more structural: it is possible that viral news videos were less popular because journalists simply do not enjoy the gatekeeping power to promote online content. With increasing competition in the online environment, other online distributors may be responsible for promoting alternative political information genres over standard news content. To explore this, I examined one of the most influential information distributors during elections: bloggers. Most studies so far only investigate the top bloggers exerting the most influence. However, the long-tail of non-elite bloggers together can outweigh the influence of a few elite bloggers. Therefore, I coded for and identified all blogs that linked to these election viral videos (i.e. the ‘known-universe’ of election blogs). From this, I derive four mutually exclusive categories of blogs, which account for 74 percent of the blog population, and find that the election blogosphere wasn’t dominated by journalism blogs, whose presence was minuscule in comparison (only 5%). It was dominated by the long-tail of non-elite political and lifestyle blogs. Furthermore, entertainment blogs, though only 16 percent of the election blogosphere, linked significantly more to viral videos by frequency than any other genre of blogs. In the following, I describe these information gates in sequence to argue that the mainstream press is neither present nor active enough in the blogosphere to effectively mediate online political information (see Table 2).
Election 2008 viral video blogosphere categories
Political blogs were the most common category of blog linking to viral election videos (29%). This isn’t entirely surprising because they reflect blogs run by individuals or groups with overt electoral and political interests. These bloggers frequently comment on and engage topics relating to governmental affairs and public policy. Because blogs were sampled on the criteria that they linked to election viral videos, it is reasonable that many of them had an interest in politics. So it is not surprising that political blogs had the largest presence in linking to viral election videos. Lifestyle blogs were the second most common category of blogs linking to viral election videos (24%). Paralleling political blogs, lifestyle blogs are also importantly different. They belonged to individuals or families, and are often written in the form of online journals. They focus on the daily non-political happenings of authors’ personal lives, and rarely refer to overt political issues and mainstream news media. That these blogs account for almost one in four of all blogs linking to viral election videos indicates that they engage a vast cross-section of the online public in ways that standard news information probably doesn’t.
Entertainment blogs (16%) were a significantly smaller category compared to political or lifestyle blogs. They were managed by organizations or individuals belonging to popular cultural industries, and frequently had overt interests in music, film, or celebrity news and information. They tended to be current or former media workers, with experiences in marketing and public relations industries. Though being only 16 percent of all blogs, that many of these entertainment blogs were present in the election blogosphere is novel and echoes findings from recent scholarship on the significance of satire and infotainment in mediating public affairs today. The presence of entertainment blogs also suggests that popular entertainment venues, such as The Daily Show or Saturday Night Live, may not be the only location where entertainment values now influence politics.
Journalism blogs only represented 5 percent of the election blogosphere. These were run by individuals who identified as practicing journalists, or professionals who worked within the news industry. That journalism blogs were a minute segment of the election blogosphere was both surprising and alarming. Could it be that most online political videos were mediated not by journalism blogs, but by political and non-political blogs or entertainment-oriented blogs? To investigate this possibility, I employ more robust statistical techniques. Table 3 lists average number of links per category of blogs to the viral video collection. On average, journalism blogs and political blogs seem to link more to viral videos than entertainment and lifestyle blogs. But this is misleading. Like most internet data, these links are distributed in a non-parametric fashion, and standard statistics such as means and standard deviations do not tell an accurate story. Applying the appropriate non-parametric significance test reveals a far different narrative: entertainment blogs linked significantly more to the viral video population than other categories of blogs (regardless of sample size). Of six possible relationships between four categories of blogs, Man-Whitney U tests illustrate that only entertainment blogs linked more to viral videos than all categories of blogs. Entertainment blogs (mean rank = 124.05) had more links to viral videos than journalism blogs (mean rank = 104.02); z = 2.167, p = 0.030. Entertainment blogs (mean rank = 322.40) had more links to viral videos than political blogs (mean rank = 278.65); z = 2.962, p = 0.003. Entertainment blogs (mean rank = 212.26) also had more links to viral videos than lifestyle blogs (mean rank = 192.33), though this relationship was only significant at the 90th percent interval level; z = 1.768; p = 0.077.
Blogs-linking frequency to viral video population by blog categories
The important story about the gateways mediating online political information is that entertainment bloggers were most active in promoting and distributing online viral videos. Journalism blogs were almost negligible in their comparative presence and linking frequencies, and weren’t able to compete with the collective presence of entertainment, political, and lifestyle blogs promoting viral content. In general, there’s little doubt that the information gateways for online political information have expanded beyond what we tend to think of as the job of traditional journalistic gatekeepers. At the very least, journalists bloggers weren’t able to compete with non-journalist bloggers to promote and distribute viral election videos.
Linking preferences between blogs and videos
Mapping both the genres of political information (categories of viral videos) and distribution gateways to them (categories of election blogs) has identified that news content is least popular and journalism blogs are the least influential brokers of viral information. But beyond this, what are the content preferences of various bloggers sharing online viral videos? Do certain blogs privilege certain video genres over others? These questions allow us to test assumptions about what journalists tend to think of as useful civic information, versus what the newer gateways might be prioritizing. Figure 1 illustrates the overall links from four categories of blogs going to four genres of viral election videos.
Linking preferences to viral video content genres
Even the few journalism blogs avoided linking to traditional news content, and instead linked to campaign content (33%). Political blogs linked most to citizen content (43%), and linked least to news content (10%). Entertainment blogs linked most to entertainment content (40%), and also linked least to news content (9%). Lifestyle blogs also linked most to entertainment content (32%), and again linked least to news content (16%). News videos were again consistently least promoted of viral election videos by all categories of blogs (including journalism blogs).
It is also generally the case that each category of blogs linked to viral election content based on their unique information preferences.
First, citizen content was linked most by political blogs. This suggests that citizens with an overt interest in mainstream politics seem to value content produced by fellow citizens. This supports the idea that online citizens prefer political information that nurtures feelings of co-presence, downplays the importance of institutions, and provides for a conversation between citizens (Coleman and Blumler, 2009). Many citizen videos were produced by unaffiliated citizens speaking directly each other, and sometimes speaking directly to elites, as well. Citizens who care about politics, then, tend to prefer political information that promotes active dialogue.
Second, entertainment and lifestyle blogs seem to value entertainment content over all else. This is not surprising but interesting, nonetheless, that entertainment bloggers were an influential source of gatekeepers with a specific content bias. That lifestyle blogs also favor entertainment content is interesting for a related reason. If non-political citizens favor entertaining, satirical, and comedic coverage of mainstream politics, it becomes very important to evaluate the effects of consuming such political content. Though this trend is consistent with recent scholarship highlighting the growth of soft news and infotainment, we do not know if the end result of consuming this content is further disengagement from mainstream politics and increased feelings of political distrust and cynicism. On the other hand, satire and creative communication can be engaging and refreshing to audiences in ways that traditional journalists find challenging to provide. Satirical personalities like Jon Stewart and Stephen Colbert frequently speak truth to power with less risk than mainstream journalists of losing political capital and access to sources (Gray et al., 2009).
Third, it is perhaps most significant that all categories of blogs consistently linked to campaign videos. Campaign videos ranked first or second highest in being linked by all blogs: 33 percent of journalism links, 30 percent of political links, 29 percent of entertainment links, and 27 percent of lifestyle links. So it seems there is a generally high level of interest in direct-to-citizen content which bypasses journalists and news workers. Instead, online audiences seem more interested in tuning in to hear campaign elites directly.
Conclusion
Today, 69 percent of internet users report watching or downloading online videos (Purcell, 2010). Since 2007, online video content consumption has grown rapidly for comedic or humorous videos (from 31 to 50%), and political videos (15 to 30%). These videos commonly spread through social networking sites like MySpace and Facebook (52%), video-sharing sites like YouTube or Google Video (49%), and news websites and blogs. Online videos distributed outside the domain of broadcast media, then, are indeed having a daily and lasting impact on citizens. This study of viral election videos and their engagement within the blogosphere during the 2008 presidential campaign has identified and illustrated two key developments and challenges for the future of news and public information in post-broadcast democracies. These findings speak to the enduring changes taking place in the online landscape of civic information, and highlight important gaps that exist between journalists and their audiences.
First, judging from the distribution of viral election video content, the landscape of online political information has expanded well beyond standard news content. Newer genres of political information include direct-to-citizen campaign videos, citizen-produced videos, and entertaining and satirical videos. Standard news videos were least common and least popularly viewed. They were also distributed least by all categories of blogs. Entertainment videos, created by both elite productions and amateur citizen productions, were most popularly viewed. In fact, the top three videos of the viral video population were all entertainment videos, and accounted for 25 percent (70 million views) of all cumulative views. The total sample of viral entertainment videos (22 videos) accounted for 40 percent (118 million views) of all viral election video views. The immense popularity of entertaining and satirical content reflects a general skepticism to formal and detached elite coverage provided by traditional news media.
Second, the online gatekeeping mechanisms that mediate and distribute political content have expanded beyond journalists. At least three new categories of gatekeepers competed with journalists (5%) in the election blogosphere. For example, citizen blogs, including both non-elite political and lifestyle bloggers, made up over 50 percent of election bloggers. These trends suggest that journalists are increasingly outnumbered by networked gateways to political information. Furthermore, the gateways occupied by professional journalists (journalism blogs) are generally ignoring newer genres of political information relevant to today’s publics. Even entertainment bloggers were much more likely to link to newer genres of political information than journalist bloggers.
These trends (the diversification of information genres, and information gateways to access them) indicate that the networked public sphere nurtures new and distinct information preferences. Journalists and the press do not seem well positioned to accommodate either shift. With more civil society actors participating in the public sphere, the diversification of political information may be good for democratic life. But the breakdown of barriers to produce and vet public information by non-journalists may also be problematic. It is important to adjust our media research agenda to explore what the effects of new genres may be, and how they engage with new styles of citizenship and participation.
Footnotes
Acknowledgements
The author gratefully acknowledges support for this study from the Google Research Award funded Virality of Information project (www.retroV.org), the Center for Communication and Civic Engagement (www.engagedcitizen.org), the Department of Communication, and the Information School, at the University of Washington, Seattle. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of supporting organizations. Replication data used in this research are available at: 