Tweeting Conventions

Abstract

This study explores the use of Twitter by political reporters and commentators—an understudied population within the rapidly growing literature on digital journalism—covering the 2012 Republican and Democratic conventions. In particular, we want to know if and how the “affordances” of Twitter are shaping the traditional norms and routines of US campaign reporting surrounding objectivity, transparency, gatekeeping, and horse race coverage, and whether Twitter is bursting the “bubble” of insider talk among reporters and the campaigns they cover. A sample derived from all tweets by over 400 political journalists reveals a significant amount of opinion expression in reporters' tweets, but little use of Twitter in ways that improve transparency or disrupt journalists' (and campaigns') role as gatekeepers of campaign news. Overall, particularly when looking at what political journalists retweet and what they link to via Twitter, the campaign “bubble” seems at the moment to have remained largely intact.

KEYWORDS:

• content analysis
• elections/campaigns
• internet/new technology
• journalism
• Twitter

ACKNOWLEDGEMENTS

The authors wish to thank Jose Araiza, Jonathan Lowell, Rachel Mourao, Magdalena Saldaña, and Pei Zheng for their research assistance, and Mitchell Wright for designing and maintaining an innovative data-capture system.

Notes

1. Twitter use is expanding quickly but is much more common among journalists and politicos than among the general public, only a small percentage of whom use it (Pew Research Center 2012; see also Farhi 2009b; Pew Research Center 2011a).

2. Today, reporters are less likely to ride the bus, or the campaign plane, than in Crouse's time, in part because of the expense and partly because of new media tools that make it easier to cover the campaign virtually rather than from inside the “bubble,” as reporters often refer to it (see Farhi 2009a).

3. This lack of transparency about campaigns' often meticulous efforts to manage the news explains the firestorm surrounding the New York Times’ revelation that journalists who covered the 2012 presidential race had allowed campaign officials to vet quotations from the candidates and surrogates before publication—in exchange for reporters' continued access to the candidates (Peters 2012). This so-called “quote approval,” the New York Times reported, “has become accepted in Washington and on the campaign trail” (para. 3). What had not become accepted, it seems, is public knowledge of the practice.

4. As Mitchelstein and Boczkowski (2009, 575) observe, in the emerging research on social media and journalism, “most studies continue to apply existing lenses to look at new phenomena,” although “the potential for theoretical renewal is becoming increasingly evident.” The opposite problem seems also to be the case: that too little effective use is being made of research conducted on mainstream or traditional news by scholars taken with the novelty of new media platforms and tools. Indeed, as Mitchelstein and Boczkowski also observe in their review of that literature, “Historical matters have not figured prominently in the scholarship about online news production [which] runs the risk of overemphasizing novelty and gives a sense of shallowness to the empirical findings and associated theoretical conclusions of many studies” (575).

5. Of course, as Singer (2001) has noted, this gatekeeping function has been eroding since the advent of online news. Debates have continued, however, about the degree to which gatekeeping persists even in seemingly non-editorial processes like citizen bloggers-as-aggregators.

6. The tension between the professional objectivity norm and how reporters might really use social media is evident in a recent version of the Washington Post's social media policy, which states: “nothing we do must call into question the impartiality of our news judgment. We never abandon the guidelines that govern the separation of news from opinion, the importance of fact and objectivity, the appropriate use of language and tone, and other hallmarks of our brand of journalism” (Hohmann 2011, 44).

7. We chose states to include in our sample based on advertising spending at the time the sample was compiled (see Associated Press 2012; New York Times 2012). Pennsylvania was later surpassed by spending in Wisconsin and New Hampshire and was No. 10 in campaign ad spending as of October 23, 2012. Because many state and local outlets have at best one reporter assigned to cover national politics, we chose all reporters who listed politics as a beat who were available in the database within each state.

8. The full coding scheme is available upon request from the authors.

9. The final sample of 1237 tweets did not differ from the main sample of 1629 in terms of distribution across platforms: tweets from Web journalists (37.4 percent), print newspaper and magazine (30.5 percent), cable television (20.1 percent), broadcast television (11.2 percent), radio (0.5 percent), and wire services (0.2 percent); national outlets created the bulk of the tweets (82.1 percent), while the rest came from local or regional news outlets (17.9 percent). Full results of zero-order correlations between news outlet types and tweet characteristics are available from the authors on request.

10. Overall agreement and Krippendorff's alphas by variable: Tweet focus, 89 percent, α=0.84; link source, 94 percent, α=0.91; retweet type, 100 percent, α=1; retweet source, 100 percent, α=1; opinion, 89 percent, α=0.78; information-seeking, 99 percent, α=0.80; job talk, 93 percent, α=0.77; personal information, 99 percent, α=0.66; voter bloc, 99 percent, α=0.94; horse race, 99 percent, α=0.89; policy issues, 99 percent, α=0.96; fact-checking, 100 percent, α=1.

11. Interestingly, 86.3 percent of those retweets contained no additional comment by the journalists passing on someone else's tweet.

12. Coders were instructed to determine the identity of the Twitter user whom the journalist in our sample directly retweeted—not the Twitter user who originally wrote the tweet. While the latter would certainly be a good measure of the degree to which outsider voices can be included in journalists' Twitter conversation, as a practical matter it can be difficult to trace the progeny of retweets, particularly those that are retweeted widely.