President Obama’s executive action on immigration provoked a predictable political storm on Twitter, with politicians, pundits and the perennially opinionated staking out their positions.  A Twitter visualization shows tweets peaked at about 9,500 tweets a minute during the president’s White House speech on Thursday, Nov. 20. But while the social media platform heaved under the weight of immigration-related tweets, immigration journalists were relatively quiet.

How quiet? A survey of 34 national immigration reporters’ Twitter feeds returned a total of 913 tweets in the 24 hours starting at 7 p.m. on Thursday, Nov. 20. This is in comparison to the 384,999 #immigration tweets which were posted in just two hours between 7 p.m. and 9 p.m. on Thursday Nov. 20.

The original list of 34 reporters was compiled with the help of Frank Sharry, the founder and executive director of the immigration rights group America’s Voice and included well-known immigration reporters such as NPR’s Mara Liasson and Julia Preston from The New York Times. But since they didn't tweet much in that time period (Liasson didn't tweet at all) we decided to focus on journalists who tweeted at least 30 times to look for usage patterns and possible trends.

The list originally included pundits such as Byron York from The Washington Examiner and Mickey Kaus from The Daily Caller but we decided to deemphasize primarily opinion journalism in our analysis.

Once we removed these two groups of tweeters, we were down to just eight journalists who were split between primarily digital publications and legacy outlets: Sahil Kapur from Talking Points Memo; Seung Min Kim from Politico; Adrian Carrasquillo from BuzzFeed; Elise Foley from The Huffington Post; Mark Caputo from the Miami Herald; Zeke Miller from Time; Ed O’Keefe from The Washington Post; and Damien Cave from The New York Times.

The story provided a rare opportunity to analyze journalists’ use of Twitter because it was possible to set up searches for individual reporters in advance of the White House speech instead of trying do historical research on the hashtags. Much Twitter research is centered on hashtags, but hashtag searches return so many results that it’s difficult to find any signal in the noise. Also, we found that reporters appeared reluctant to use hashtags, which further reduces the efficacy of such efforts in studying journalists’ use of Twitter.

This study is not intended to be a criticism of any of the journalists surveyed -- it’s more an attempt to understand how individual journalists are integrating Twitter into their daily work and how journalists actually work on the platform. For example, questions arise daily for both journalists and journalism students and it is useful to see how working journalists address those questions.

Questions like:

  • How often did they tweet?
  • What information did they share? Their own? Other journalists or news organizations? Official sources? Their audience? Who did they engage with on Twitter? Their colleagues? Other journalists or news organizations?
  • Did they share share their opinions?

How often?
A trio of male reporters from legacy outlets accounted for 268 tweets of the total 467 Twitter output (or 60 percent), and the leading tweeter from the group was Marc Caputo from the Miami Herald who tweeted 112 times over the 24 hours. He was followed by Time’s Zeke Miller with 93 tweets and Ed O’Keefe from the Washington Post with 63. In comparison, Julia Preston from The New York Times tweeted twice, Dana Bash from CNN tweeted once and NPR’s Mara Liasson did not tweet at all during the same 24 hours. These findings may suggest further research into gender on Twitter to see if those patterns are the same overall.

Overall, these reporters tweeted at almost double the rate of the reporters from primarily digital publications, with 301 tweets against 166 for the latter group.

Sharing and engagement
All the reporters shared information via either retweets or links to their own content or employer as well as to reporters from outside their own news organization. However, while all of the reporters were generally comfortable sharing links to competing news organizations or journalists they overwhelmingly shared content to and from other journalists or news organizations and official sources. There is very little evidence of any interaction with their audience.  

Opinions
In keeping with general journalism practice, none of the tweets surveyed expressed opinions on the issue of immigration, but according to AP, it’s not just the journalists' own opinions that can cause problems. Retweeting others’ opinions can also be problematic. The AP advises journalists to “adorn” retweets by adding some context or explanatory text before the retweeted text,” and offers this example of a “bad” tweet:  “RT @jonescampaign: Smith’s policies would destroy our schools.” The guidelines advise reporters to use the following edit:  “Jones campaign now denouncing Smith on education. RT @jonescampaign: Smith’s policies would destroy our schools.”

AP says this step helps to ensure that the journalist is not associated with the content of the tweet. I use this rule in the classroom because it’s a good way of getting student journalists to slow down and think about what they’re tweeting. It’s good advice, although students often report that it’s hard to put into practice.

And that practice doesn’t seem to be catching on outside of AP. The majority of retweets from the immigration journalists were posted without such explanatory text. For example, of the total 161 RTs and MTs from the immigration reporters only 7 tweets, or 4.3 percent of the sample, used the AP method.

I’m not convinced that the boilerplate “RTs are not endorsements” is enough to protect a journalist’s reputation and the AP advice, if it catches on more widely, would be a good solution to the problem with inadvertently sanctioned opinions.

This sample shows that there is more work to be done on increasing journalists’ engagement with the non-journalists and possibly more debate needed around retweets being perceived as endorsements or opinions.

  • This project was made possible by an academic research grant from Discover Text. which offers a powerful set of tools to collect, sort and analyze Twitter data. This project is only  a small example of what’s possible with the software.