I'm constantly watching the impact of hashtags on politics and advocacy - from the explosion of #BlackLivesMatter and #YesAllWomen to the way candidates try to establish their own campaign hashtags (think #FeelTheBern and #ImWithHer). As I prepare to give a presentation on the topic at the upcoming #HashtagPolitics event at Baruch College in New York City, I'm looking for new ways to deconstruct the use of political hashtags. As part of my research for this event, I've been exploring various tags used by, and for, the top presidential candidates using a free tool called Hashtagify.me. For any hashtag you enter, Hashtagify.me will tell you which other hashtags it's most commonly paired with.
For today's analysis, let's compare the #NeverTrump and #NeverHillary hashtags. Hashtagify.me provides visual representation of both the overall use of these tags, based on the size of the circle, and what tags are most likely paired with them.
As you can see in images below, #NeverTrump is more commonly used than #NeverHillary. This is not surprising - #NeverTrump is a more deliberate campaign, with several prominent Republicans driving it (see Matt Bai's critical analysis of the #NeverTrump effort for more insights).
#NeverHillary is a derivative hashtag, inspired by #NeverTrump, but without a formal effort behind it. The fact that it's garnered such a large number of uses is a good indication of how an effort like #NeverTrump can spur additional hashtags related to other candidates.
#NeverCruz, #Never Kasich and #NeverSanders hashtags have also popped up - #NeverCruz is used about as often as #NeverHillary, while the others are used far less, with #NeverSanders used the least (similarly, #NeverBernie also ranks at the bottom of this list).
Looking back at the "#NeverFrontrunner" data, it's interesting to note which hashtags are most often paired with the two #Never hashtags.
With #NeverTrump, we see it most commonly paired with #Trump and #DonaldTrump. This helps to ensure that those favorable to Trump will see the #NeverTrump messages. It's also frequently paired with #UnitedWeCruz, #ChooseCruz and #CruzCrew, suggesting that Cruz supporters are using the #NeverTrump prolifically to take down the frontrunner and elevate their preferred candidate.
Other commonly paired hashtags with #NeverTrump are #TCOT, #PJNET and #GOPdebate. #TCOT and #PJNET are two very large conservative Twitter conversation communities - their use of #NeverTrump suggests a strong opposition to The Donald from the more diehard conservatives. The #GOPdebate hashtag is far more mainstream, as it's commonly used by everyone live tweeting the debates. It's hard to draw any implications here without knowing who among the live tweeters are also using #NeverTrump (for this analysis, I'm only using the freely available data from the site, as most people using it would).
Note, #NeverTrump doesn't appear to be used in conjunction with Democratic or progressive hashtags. For now, it seems that the left is sitting back and watching the Republicans battle among themselves.
As for #NeverHillary, we see quite a different pattern. This hashtag seems to be used predominantly by Bernie Sanders supporters, Donald Trump supporters and #NeverTrump supporters. In other words, the #NeverHillary effort, albeit smaller than the #NeverTrump campaign, is hitting Clinton from both the left and the right. By contrast, the #NeverTrump campaign seems to be coming at Trump predominantly from the right, though that "right" may include more mainstream Republicans who see Trump as a RINO (Republican in Name Only).
Beyond the specific observations made here about these hashtags, I want to stress the usefulness of Hashtagify.me for helping us better understand the context of hashtags, especially political hashtags. When viewed alone, a hashtag doesn't always provide enough context to really understand what is going on. Many people jump into using a hashtag because they agree with the main thrust of the conversation, but many others will use it to parry that thrust. Knowing what hashtags are being paired with the core one being observed gives us greater insight into the context (without having to spend big bucks on content analysis).