Political Facebook unfriending

The academic way to introduce the topic of politically-motivated unfriending on Facebook would be to talk about the gap in the literature that the study of it fills. And it does fill a gap: despite an enormous body of literature that endeavors to explain what the internet has done and is doing to politics (and, to a lesser extent, what politics does to the internet), and despite another enormous body of literature that talks about the effects of media selectivity, we really know virtually nothing about the act of unfriending a Facebook friend on political grounds. To the extent that social media are becoming an increasingly important arena for political discussion, we should probably try to learn more about unfriending, given that one of its consequences is a more homogenous Facebook feed.

Another academic way to start talking about unfriending would be to highlight the importance of understanding disconnectivity as an integral part of any attempt to understand connectivity. Important steps in this direction have been made by people like Ben Light, Laura Portwood-Stacer, and Tero Karppi, but even they would agree there is still much we don’t know. It’s clear, though, that unfriending is becoming a political tool: there was a flurry of unfriending in Germany when people were urged to unfriend friends who had liked the page of a racist, anti-immigrants organization; there would appear to have been some unfriending among white people in response to the grand jury’s decision against indicting the white policeman who had shot Michael Brown in Ferguson. All we have here, though, is anecdotes. Some research is clearly needed.

But this is not why I decided to study politically-motivated Facebook unfriending. In the summer of 2014 the protracted conflict between the Israelis and the Palestinians flared up, with six weeks of combat between the Israeli army and Hamas in the Gaza Strip. Unsurprisingly, social media went into overdrive. The Facebook feeds of Israeli Facebook users (of whom I am one) were dominated by the fighting. It was actually considered bad taste to post anything unrelated to the war (though it was permissible to that on Instagram). While people were writing about their experiences, posting images, sharing news reports, commenting, liking, and the rest, there were also lively discussions about what one could or couldn’t say about the combat. These discussions were about the limits of legitimate discourse. One response to Facebook users who were perceived to have overstepped those limits and expressed themselves in an illegitimate fashion was to unfriend them. Judging from the Israeli media, this seemed to be a pretty prevalent practice (among Jewish Israelis) at the time.

But how prevalent actually was it? Who was unfriending whom, and why? Was unfriending more characteristic of people with certain political leanings? How should we understand the act of unfriending in this kind of context? Would there be as much of it in other political contexts that are less extreme than war, such as election campaigns?

After being politely told by a Facebook research department still reeling from the emotional contagion experiment fallout that, no, they wouldn’t be able to collaborate with me on this (at least that’s my explanation), I realized I would have to collect the data myself. And so, one week after the final ceasefire, I surveyed 1,013 Jewish-Israeli Facebook users and asked them whether they had unfriended or unfollowed anyone during the period of fighting for reasons that were to do with the fighting and the politics around it. To the best of my knowledge (and I would honestly be delighted to be put right about this), this is the first time Facebook unfriending behaviors in relation to a given event have been studied.

Once the results were in, I sat down with my collaborator, Dr. Shira Dvir-Gvirsman, to analyze them. The headline finding was that 16.1% of Jewish-Israeli Facebook users (yes, I think my sample was quite representative of that population) had unfriended or unfollowed someone during the fighting. In other words, one in six of the respondents had carried out an act of disconnectivity that impacted on what showed up on their Facebook feed. They had decided to silence other Facebook users, to mute them on their feed.

Did these unfrienders and unfollowers have any distinguishing traits? I won’t reproduce the output of the logistic regression here, but we did find that you were more likely to have unfriended or unfollowed a Facebook friend the younger you were, the more friends you had, and the more politically active you were on Facebook. Interestingly, the political views of Facebook users did not influence unfriending, but the extremity of those views did; that is, the further respondents positioned themselves from the center on a political scale of left to right wing, the more likely they were to unfriend or unfollow. What emerges is an image of the Facebook user who silences others on his feed as young, politically active, and with relatively extreme political views (this is as opposed to centrist views; I do not mean to imply that they are extremists in the normal sense of the word).


Whom were the unfrienders unfriending? Simply put: weak ties. Two-thirds of those who unfriended said they had disconnected from someone they are not close to; only 2% said they had unfriended someone from their family. This makes sense: there is a cost to face in unfriending, and this cost may be too high vis-à-vis family members or close friends. We may also be willing to put up with family members because we love them, and not only because we daren’t unfriend them. Furthermore, we tend not to know the political views of our weaker ties, and hence may be unpleasantly surprised when they are expressed-and during the fighting last summer, politics was pretty much all that Israelis were discussing on Facebook, providing plenty of opportunities to learn about the political views of people whose attitudes had up until then remained unknown.


The two main reasons given by respondents for unfriending during the fighting were encountering content they disagreed with and encountering posts that offended them. Being argued with was not in itself a reason for unfriending (only 7% of respondents said that had been a factor). Also, Jewish-Israeli Facebook users tended not to unfriend other users for posting too much about the war, which is notable given that in the US, and in admittedly more routine times, posting too much about politics is the top cause of politically-motivated Facebook unfriending.

Not long after the war, Prime Minister Netanyahu dissolved the Israeli parliament (the Knesset) and called general elections, providing the electorate a chance to go to the polls, and providing me with another opportunity to measure some more political Facebook unfriending. Dr. Dvir-Gvirsman and I are still poring over the election data, but one or two things jumped out at us immediately. (I don’t want to go into too much methodological detail. Suffice to say that we returned to the same respondents as before, and 737 responded (out of the original 1,013), which is a pretty decent response rate for the second wave of a longitudinal panel survey.) Perhaps unsurprisingly, there was less unfriending around the elections than there was around the war, but there was quite a lot more unfollowing, and thus more disconnectivity in general.


By and large, the same variables that predicted unfriending and unfollowing during the war predicted unfriending and unfollowing in the elections as well. (Experts in Israeli politics might be interested to learn that left-wingers were more unlikely to remove other Facebook users from their feed than right-wingers during the elections, but this is not the time or place to go into this.)

However, what we find particularly intriguing is that whether or not you unfriended or unfollowed during the summer’s fighting is a far stronger predictor of doing the same during the elections than whether or not you had unfriended before the war of 2014. This can be seen in the probability tree: if you had never unfriended in the past, but did so during the war, the chance that you would unfriend in the elections was much higher than if you had unfriended in the past, but didn’t during the war. This suggests that political unfriending leads to more political unfriending. The act of political unfriending or unfollowing-controlling the political slant of your Facebook feed-would appear to be something people add to their digital repertoire. Unfriending for any reason at all at some time in the past is related to unfriending in the recent Israeli general elections, but unfriending during last summer’s war is a far more powerful predictor.

john-fb4Unfriending and unfollowing allow us to shape the politics of our Facebook feed at a remarkable level of granularity. But the effects of this may go beyond the filtering and muting that we do ourselves. As we know, Facebook doesn’t show us everything that’s going on in our network; it is selective, feeding multiple cues into its algorithm. We can only speculate as to the effects of unfriending and unfollowing on the political content of our feed. In the meantime, I hope that others will be tempted to survey Facebook users’ disconnective practices in relation to elections, protests, wars, and other potentially divisive events.

If you’d like to hear more about this, Shira and I will be presenting at the ICA in San Juan, so come along, and/or get in touch for a chat.

Postscript: A couple of small things. First, you may be wondering where Twitter is in this story. For various reasons, Twitter is not popular in Israel, with around 155,000 users, compared to 4m Facebook users (out of a population of 8m). Second, you may be thinking that this research would be richer if I had interviewed some unfrienders as well. I have, and I’m still in the process of doing so, so I can’t really report on that part of the research yet.

Update: You can read a published version of parts of this research in the Journal of Communication, or ask Nicholas for a copy.