News Feed Filtering: A Threat to Online Political Expression?


In recent years, it has become conventional wisdom that social media sites like Facebook and Twitter serve as democratic platforms for citizens to express their political views and opinions. The key assumption underlying this idea is that these sites operate on a peer-to-peer sharing structure that allows anyone to communicate to a public audience with their posts—provided, of course, that these audience members are connected to them as “Friends,” “Followers,” etc. Therefore, a social media user’s level of reach and influence (including political influence) could be understood as simply a matter of how many connections they have in their networks.

However, as Facebook has moved more and more towards a model of complex algorithmic news feed filtering, the notion that one’s posts will actually be seen by one’s peers is increasingly being called into question. This development, which is likely to become industry standard, may have important consequences for the use of social media as a platform of political expression.

At the moment, these sorts of concerns about Facebook are largely centered on the diminishing “organic reach” of organizational Facebook Pages, such as those of non-profit activist groups (in addition to businesses and brands). What is “organic reach,” you ask? As B. Traven explains in a great piece for ValleyWag,

Put simply, “organic reach” is the number of people who potentially could see any given Facebook post in their newsfeed. Long gone are the days when Facebook would simply show you everything that happened in your network in strict chronological order. Instead, algorithms filter the flood of updates, posts, photos, and stories down to the few that they calculate you would be most interested in. (Many people would agree that these algorithms are not very good, which is why Facebook is putting so much effort into refining them.) This means that even if I have, say, 400 friends, only a dozen or so might actually see any given thing I post. One way to measure your reach, then, is as the percentage of your total followers who (potentially) see each of your posts. This is the ratio that Facebook has more-or-less publicly admitted it is ramping down to a target range of 1-2% for Pages. In other words, even if an organization’s Page has 10,000 followers, any given item they post might only reach 100-200 of them.

As Traven argues, this lowered capacity for political organizations to send messages to their own followers significantly compromises Facebook’s role as an open marketplace of political expression. The only way that organizations can truly bypass these filters is to pay Facebook to promote their posts, and while major commercial brands (and perhaps major party candidates) can afford the price, small non-profits and advocacy groups largely cannot. Thus, Facebook is starting to look less like an even playing field for political communication and more like television, where big-ticket ad buys dominate and the resource-rich get a much louder megaphone than the resource-poor.


However, I can’t help but think that this is just the tip of the iceberg. The impact of news feed filtering will likely extend far beyond the realm of organizational Facebook Pages, which are already experiencing the effects of a diminished “organic reach.” Now that we know that Facebook is making behind-the-scenes decisions about which posts you will see from your connections and which posts will be hidden from you, how else could this model end up skewing the flow of online political discourse?

To explore this question, it is helpful to understand exactly how Facebook currently filters its News Feeds. However, this is not as easy as it sounds. According to TechCrunch, the algorithm that sorts Facebook News Feeds uses over 100,000 different factors to determine the relevance of posts. Among the most important factors are the following (as told to TechCrunch by Facebook News Feed Director of Product Management Will Cathart):

• How popular (Liked, commented on, shared, clicked) are the post creator’s past posts with everyone
• How popular is this post with everyone who has already seen it
• How popular have the post creator’s past posts been with the viewer
• Doe the type of post (status update, photo, video, link) match what types have been popular with the viewer in the past
• How recently was the post published

In addition, Facebook recently announced that it would push to prioritize “high-quality” content in users’ feeds and reduce the amount of spam posts. What constitutes “high quality” in Facebook’s eyes? Here is how the company defines it on its website:

While the goal of News Feed is to show high quality posts to people, we wanted to better understand what high quality means. To do this we decided to develop a new algorithm to factor into News Feed. To develop it, we first surveyed thousands of people to understand what factors make posts from Pages high quality. Some of the questions we asked included:
• Is this timely and relevant content?
• Is this content from a source you would trust?
• Would you share it with friends or recommend it to others?
• Is the content genuinely interesting to you or is it trying to game News Feed distribution? (e.g., asking for people to like the content)
• Would you call this a low quality post or meme?
• Would you complain about seeing this content in your News Feed?

In other words, Facebook is ostensibly adopting a model of ‘giving people what they want,’ rather than making editorial decisions about quality based on their professional judgment. Indeed, in a piece for Fortune, Matthew Ingram writes that “Facebook’s director of news partnerships, Andy Mitchell, rejected the idea that the social platform is some kind of gatekeeper when it comes to the discovery of news, saying Facebook doesn’t control the news-feed — users control it by telling Facebook what they are interested in. In other words, Facebook sees itself as merely reflecting the desires of its users.” Critics, on the other hand, have pointed out that the process is still susceptible to censorship based on various national laws as well as the site’s own community standards.

However, an even bigger potential issue here is how news feed filtering may pave the way for Facebook (and social media sites more generally) to manipulate the flow of news and political expression along ideological lines. Surely, the company would vehemently deny doing this. Yet I couldn’t help but think of what the recent gay rights-celebrating rainbow profile picture campaign suggests about Facebook’s capacity to elevate the flow of posts that display a particular political point of view. Did Facebook actually favor these rainbow picture posts in its News Feed algorithm, above and beyond the criteria outlined above? Honestly, I have no idea. Yet the very fact that the company itself orchestrated this particular meme on its own platform indicates that it is not entirely above the fray of partisan political battles, and furthermore, appears willing to leverage its prominent position in the viral media culture to advance certain political ideas.

Since I strongly support the sentiment behind the rainbow profile pictures, it’s hardly a sticking point for me. However, the campaign seems to complicate Facebook’s insistence on being a fully neutral channel for political discourse with no vested interests beyond serving the most relevant content to its users. Could it be possible that the site would play with its algorithms to filter down the reach of posts that it deems inimical to its worldview, or boost the flow of posts that support issues favored by its leadership? This may sound like conspiracy theory territory, but Facebook has opened Pandora’s Box in a very real sense by moving to a model of news feed manipulation.

What, then, could be a solution to this looming threat? It seems unreasonable to argue that Facebook should simply go back to its original model of showing every piece of content from Friends in the order in which they were posted. As the company often points out, there are so many people and organizations now on Facebook that an average user has approximately 1,500 posts that are eligible to show up in their News Feed every day—i.e. far too many to wade through. Twitter, by contrast, has had a long-standing commitment to showing every tweet from a user’s connections in chronological order, but recently announced via the New York Times that they are “question[ing] our reverse chronological timeline” in order to stay competitive and profitable.

At a time of increasing news feed manipulation, what I’d like to see is not only more transparency about how these feeds are actually filtered (which could alleviate fears over ideological bias), but also more user choice about how the criteria are applied. In other words, Facebook users should be able to decide how much “organic reach” they receive from their own connections. Now, it is true that Facebook already allows users to control their feeds to some extent by hiding posts from specific “Friends.” However, these options can be greatly expanded. Want to see more posts from your political organization Pages than just the default 1-2%? You should be able to choose such an option with a simple click of a button. Want to see more political posts that diverge from your views, rather than just seeing those that the algorithm determines you’d be most likely to “Like?” You should be able to have that choice in the settings. Putting users in more control of their own news feeds would go a long way towards mitigating concerns that social media sites like Facebook are becoming the new political media gatekeepers.

Facebook’s Rainbow Profile Pictures: The ‘Slacktivism’ Debate Rages On


Every time a symbolic online campaign goes viral, a new round of debate ensues over the value of so-called ‘slacktivism,’ and Facebook’s recent gay pride rainbow profile picture craze is no exception. Amidst the flood of think pieces that appeared in response to the rainbow images, two articles in the Washington Post caught my attention in particular. First, there’s Caitlin Dewey’s slice of digital optimism, in which she makes a case for why these symbolic actions may be subtly, yet meaningfully, persuasive. While reading this piece, I heard echoes of my own research on the 2013 red equal sign profile picture campaign, which posited that the peer-to-peer agenda-setting dynamics of these symbolic actions cannot be easily dismissed.  In Dewey’s article, UW professor (and professional colleague) Philip Howard offers the key pull-quote: “Openly declaring your support for gay marriage or dissatisfaction with modern policing [on social media] can create a brief moment where people who see the image have to think about the issue.” Dewey then follows: “And if they don’t have a clear opinion, or see that image many times, it could — potentially! gradually! — shift their social views.”

Indeed, this is precisely the sort of logic that was invoked by a number of interviewees in my red equal sign study, and while the extent of such proposed effects is still very much an open question, it’s nice to see these ideas being given serious consideration in the mainstream media. As I point out in my article, the ‘slacktivism’ discussion has become overly focused on how online symbolic actions may relate to other (typically offline) forms of political participation–either empowering people to do more by boosting their identification, or convincing them to do less because they feel like they’ve already done their part–and this ends up side-stepping the thorny question of viral political persuasion altogether. Personally, I don’t think there are any easy answers to this question, and I balk when my research is used to unequivocally ‘prove’ that symbolic online activism ‘works’ (as it was in this Christian Science Monitor opinion piece from 2013). However, I am pleased to see that the issue of peer persuasion is edging into the public conversation about online political expression, particularly since it has been brushed aside by so many critics (as well as many academics in my field, quite frankly).

Another piece in the Washington Post, an op-ed written by LGBT activist Peter Mosk0witz, has gotten considerably more attention and has seemingly risen to the top of the blogosphere heap. In the piece, Moskowitz repeats the typical ‘slacktivism’ critique that online actions become a weak substitute for ‘real’ activism, citing a 2014 study which found “that people who make these token displays of support often do it simply to boost their own public images without making any real sacrifice to benefit the cause.” This research report, subtitled “How the Social Observability of an Initial Act of Token Support Affects Subsequent Prosocial Action,” is a prime example of what I’m talking about above. Once again, the focus is on how taking a symbolic action online will influence future political behaviors. While this line of research is not without value, it conveniently avoids the issue of what symbolic online action may be doing (or not doing) as its own distinct phenomenon. Even the use of the term “token support” here to describe this symbolic activity suggests that the researchers are unwilling to entertain the notion that it might have real-world consequences in and of itself.

On the other hand, Moskowitz offers a more convincing critique of the Facebook rainbow profile picture campaign when he brings up the issue of political self-segregation on social media, i.e. the fact that most people who post about their political viewpoints are likely to have online connections that already agree with them, and that those who don’t agree with their views are likely to simply block them. This issue of digital ‘information cocoons,’ in the words of Cass Sunstein, is crucially important to explore moving forward. Facebook itself famously tried to mitigate these polarization concerns earlier this year with an internal study purporting to show a diverse and cross-cutting political conversation on the platform, although it has since been raked over the coals as sloppy and self-serving research.

The most controversial and talked-about aspect of Moskovitz’s op-ed, however, is his argument that straight allies should not be waiving the rainbow flag on Facebook because they do not fully understand the LGBT struggle and its history. In his view, the campaign is a form of co-optation that weakens the movement and its symbols: “When millions of people cloak themselves in a symbol without understanding what it means, they dilute that symbol’s power.” Moskovitz’s argument has since been pilloried on the right as an example of victimhood culture run amok, and on the left as an example of how some activists turn away valuable allies over petty squabbles. I have my own reservations about the idea that political symbols should be reserved for some but not others. However, I do find Moskovitz’s discussion about the ’emptying’ of the rainbow symbol to be quite interesting, and I sympathize with his concern that political symbols are often circulated online as trendy memes without much thought about their meaning and history. A major theme of my ongoing work on viral politics is that we need to become more thoughtful and reflective about the symbolic political material that we choose to forward to others, precisely because these symbolic expressions can be consequential. While I disagree with his conclusion that the Facebook rainbow pictures are somehow obscuring the multiple struggles that the LGBT community still faces (in fact, I would argue the opposite), I believe that these kinds of serious conversations about the deeper meaning of viral political content are incredibly important to have.

2014: The Year of the Political Hashtag Campaign


As 2014 draws to a close, it seems destined to go down in history as a breakthrough year for the political hashtag. Of course, we’ve seen numerous examples of this phenomenon in the past, but in 2014, hashtag campaigns dominated political discourse on social media (and beyond) and reached a whole new level of prominence. Last year, it was the profile picture that emerged as a central locus of digital-era political expression, from the red equal signs for marriage equality to the Trayvon Martin-inspired blackout campaign. This year, however, hashtag campaigns were the story for viral politics. I’m not just talking here about the use of hashtags to classify, coordinate, and publicize broader political movements, which has been popular on platforms like Twitter for quite a while now (#OWS, #arabspring, etc.). Rather, what solidified in 2014 was the widespread use of hashtags as political memes in their own right, complete with a set of corresponding actions for each participant to take in the course of spreading endless variations on a theme. In other words, the hashtag emerged not just as a tool for online political mobilization, but also as a distinct advocacy tactic.

To illustrate the delineation I’m making here, let’s consider the most significant viral politics story of the year (at least in the United States): the online-fueled movement to protest the police killings of unarmed black civilians like Michael Brown and Eric Garner. Over the course of the year, many different hashtags were connected to this movement, from the catch-all #blacklivesmatter (which came to function as both the movement’s de facto name and its key rhetorical rallying point), to emotionally-charged references to the Brown case (#handsupdontshoot) and the Garner case (#icantbreathe). As they trended, these hashtags were mostly used within the movement as rather traditional political slogans, originating on Twitter and migrating to offline venues of protest like hand-held signs and T-shirts. However, it was another hashtag, #iftheygunnedmedown, which highlighted the use of the hahtag for distinctly participatory digital advocacy. The behavior, or instructions, for the #iftheygunnedmedown meme was quite straightforward: to create their own versions, young blacks juxtaposed two photographs of themselves, one fitting the media stereotype of the “thug” and the other showing a conventionally “proper” appearance in the eyes of mainstream society (in graduation robes, military uniforms, etc.). The campaign thus critiqued the way in which black victims of police violence like Michael Brown are negatively depicted in the news media, and furthermore worked to expand and reframe the visual representation of young black men and women in the public sphere. #iftheygunnedmedown was only a small piece of the broader social media activity surrounding the #blacklivesmatter movement, yet it most acutely demonstrates how networked digital platforms in fostering unique and innovative modes of grassroots political expression.


This sort of hashtag-as-participatory meme popped time and time again in the online political discourse of 2014. Another example used in conjunction with #blacklivesmatter, #crimingwhilewhite (in which whites offered testimonials of their own privileged treatment when dealing with police), proved to be a source of controversy within the movement. The tactic was also employed repeatedly in the course of feminist online activism, another major viral politics story of 2014. In particular #yesallwomen and #whyistayed both worked to incorporate women’s personal testimonies into larger feminist advocacy efforts, epitomizing the emergent hashtag-as-meme trend. #yesallwomen was actually a response to another hashtag, #notallmen, which has been used to critique feminist arguments about the widespread nature of misogyny in society; both appeared in the aftermath of the May 2014 mass shooting in La Isla, California, perpetrated by a young man who was apparently inspired by his hatred towards women. #yesallwomen became a venue for women to share personal stories about their experiences with sexism, misogyny and the threat of gender-based violence, bringing visibility to the issue one tweet at a time. In a similar fashion, #whyistayed grew out of another high-profile violent incident in the news (involving professional football player Ray Rice’s assault on his wife), and provided an opportunity for women to collectively share stories about domestic violence in order to shine a public light on this often-ignored problem.


What I find particularly interesting about #yesallwomen and #whyistayed (as well as #iftheygunnedmedown and #crimingwhilewhite) is how they use major news stories as springboards for addressing how the larger issues at stake affect the lives of individuals throughout the society. Rather than just spreading awareness or showing support or solidarity for a cause, they seek to expand the narratives and conversations surrounding political issues and provide ways of systematically linking them to the experiences of everyday citizens. This seems to me to be one of the most important aspects of viral politics more broadly: how social media and participatory culture can be used to make the personal political—and the political personal. Indeed, the many hashtag campaigns popularized in 2014 look to be a major step in this direction.

Political Viral Videos, Citizen Journalism, and the Question of Partisanship: Interview on WWOR 9 “New Jersey Now”

New Jersey Now Joel Penney Interview

Recently, I was interviewed on “New Jersey Now,” a public affairs TV program that airs locally on WWOR 9. Fortunately, they also put the shows up online, so you can check out the interview here. In the clip, I discuss the future of political viral videos, citizen journalism and partisanship, and a number of other social media-related topics with host Jim McQueeny. We also talk about a lighthearted YouTube video put out by New Jersey Governor Chris Christie’s office entitled “No Pain, No Gain.” As you’ll see, I found the satirical tone of the clip to be a bit confusing in this specific context, although it’s safe to say that this kind of internet-style humor is the new normal for political advertising.

Black Twitter and the Rise of Viral Pressure Groups

Black Twitter

A few weeks ago, the Washington Post published an excellent piece about the phenomenon of Black Twitter, explaining how this community has used peer-to-peer social software to mobilize around numerous race-related political causes. Recently, Black Twitter activists have enjoyed a number of high-profile successes, such as pressuring a book publisher to drop a deal with a Trayvon Martin murder trial juror, pressuring InterActive Corp to admonish (and eventually fire) a PR executive who tweeted a racist joke, and pressuring singer Ani DiFranco to cancel an event at a former slave plantation. As these examples demonstrate, the political power of this form of activism rests in its high-profile application of public pressure. While tweeting responses to troubling news of racism and racial insensitivity may not “do” anything political in and of itself, it can put a media spotlight on an issue that may in turn lead to real change.

For many decades, professional watchdog organizations (AKA “pressure groups”) like the NAACP, GLAAD, and NOW  have engaged in this form of activism to varying degrees of success. Now, it appears that decentralized groups of citizens are taking it upon themselves to band together around common issues and draw the public’s attention to them through strategic media interventions. The fact that Black Twitter has no organizational center, but is rather an open-ended community of like-minded citizens who find one another via popular hashtags like #PoliticosBlackIntellectuals #solidarityisforwhitewomen, is quite significant. It reminds me a lot of W. Lance Bennett and Alexendra Segerberg’s point that the traditional collective action of  organized social movements is giving way to “connective action,” which describes a more diffuse and personalized style of public engagement that is powered by digital peer-to-peer networks.

That being said, there does appear to be a certain cohesiveness to virtual communities like Black Twitter, even though they are strongly decentralized. It is important to keep this point in mind when thinking about how political causes “go viral” on Twitter and social media platforms in contemporary times. When we say that a political news story, photo, or video “goes viral,” it conjures up images of widespread social popularity that does little to specify the investments and agendas of particular groups that have strategically contributed to the peer-to-peer spread of content. Communities like Black Twitter are essentially issue publics that draw upon their ranks to deliberately make a story like the Ani DiFranco concert or the PR exec’s racist tweet go viral – in other words, their viral popularity doesn’t just materialize out of thin air (i.e. from the aggregated effect of isolated, individual shares), but is rather the result of a concerted effort on the part of groups of invested citizens who wish to make an impact on the public sphere. Clearly, the agenda-setting labor of these viral pressure groups is a crucial object of study for scholars who seek to understand the emerging shape of digital activism.

Obamacare Stories: User Reviews Meets Political Marketing

Obamacare Success Stories Facebook PAge

Now that the Affordable Care Act has gone into effect, Republicans and Democrats have been waging an intensive marketing battle to frame its impact on the country. Quite interestingly, user-generated content has become a primary focus of both parties’ communication strategies. Using social media outreach, the parties have been collecting personal stories of Obamacare successes and (alleged) horrors to use as rhetorical ammunition in the war of public opinion. As CBS news explains, the House Republican website now has a section soliciting “Your Story” in an attempt to find people who are upset about losing their healthcare plans and are eager to speak about it. They even ask for a YouTube video link on the story submission form, signaling how they plan to use this user-generated material in their media marketing. The Democrats have responded with a similar strategy, using the Facebook page Affordable Care Act Success Stories to help build a more positive narrative about the program.

Obamacare Supporter

While the personal story has long been a staple of political rhetoric (think of all of those people that Bill Clinton would single out in his speeches), these latest developments indicate how political marketing is being thoroughly overhauled in the social networking age. As any marketing professional these days will tell you, peer recommendations that circulate online are far more trusted than traditional top-down persuasive communication, and political marketers appear to be adapting accordingly. The Obamacare stories in particular signal a ‘user review’-style approach to political communication that is fairly new. Instead of telling citizens about the benefits or drawbacks of the program themselves, they would rather you hear from ‘everyday people’ whose experiences are inherently more valued and trusted. Of course, social media facilitate this process to an immense degree, as anyone can now broadcast their ‘user review’ of Obamacare with a few clicks of a button.

In an editorial for the Guardian this week, Ana Marie Cox decries this sort of strategy as ‘anecdata,’ noting that statistical information is far more credible for making a case about a government program’s effects than a handful of individual stories that may not be representative of the broader population. Cox also makes the point that a lot of these social media-generated stories turn out to be false, such as a Facebook note written by Ashley Dionne (“this law has raped my future”) that went viral despite containing misleading information about the plans she is eligible for. Obviously, both parties should be wary of running with user-generated stories that haven’t been checked out, as this kind of internet hoax is becoming all too common. However, the emotional appeal of personal narratives can’t be denied, and I would expect that political marketing will continue to move towards these sorts of participatory formats that exploit the trustworthiness of ‘everyday people’ for rhetorical gain.

Creepy Uncle Sam: Republicans Try Viral Video Strategy to Fight Obamacare

creepy uncle sam

With the Affordable Care Act (AKA Obamacare) going into effect on October 1st, Republicans have ramped up their attempts to undermine the legislation. In addition to Congress forcing a government shutdown, conservatives have turned to viral marketing strategies to try to shift public opinion against Obamacare. The most high-profile of these efforts involves web video ads released online by a group called Generation Opportunity, funded by the Koch brothers. As explained in this piece for Time, the ads are intended to convince young adults to opt out of Obamacare so that the program will lose out on a crucial source of funding and thus fail in the long term. However, the ads having been getting the most attention for their outrageous gross-out humor, which is, of course, the lingua franca of the YouTube generation.

The ads, featuring a masked “Creepy Uncle Sam” character who threatens unsuspecting young patients with comically horrifying gynecological and rectal exams, mark an important shift in tone for American political advertising. In an earlier post, I discussed how the 2012 Obama campaign used internet-style pop culture humor to get their message across to a new generation of voters who tend to tune out more formal types of campaign messaging. Of course, the Republicans don’t want to be left behind, and the “Creepy Uncle Sam” ads signal how they too are now trying to latch onto the zeitgeist and create digital content that is entertaining and novel enough to inspire social sharing. So far, it looks like Generation Opportunity has succeeded in ‘going viral,’ with each of its video ads topping a million views on YouTube. In a sure sign of viral success, the liberal group The Other 98% has responded with an online parody that asks viewers to “opt out of Koch propaganda, not Obamacare.”

What I find particularly interesting about all of this is how the internet has freed the agents of political communication to try things that were seemingly never possible in broadcast television or other media. To put it simply, these ads are pushing the envelope of taste. I’ll leave it to other commentators to deconstruct the potential homophobic (and even racist) undertones of ads that compare Obama to a rectal-probing rapist, but suffice it to say, this is the kind of humor that one would expect from South Park rather than from a policy-minded political group. However, as political advertising moves online more and more, we can expect that its creators will do whatever it takes to capture public attention an inspire viral shares on Facebook, Twitter, and other social platforms. The fact that these videos may be too risque for prime time TV is precisely the point – in a Wild West web environment where content is almost completely unregulated, shock is the rule rather than the exception, and audiences expect to see something that they can’t and won’t see anywhere else. I wonder, though, if there are any limits as to how far online political advertising will go in terms of skirting the lines of taste, or if these sorts of strategies could potentially engender a backlash. For now, it seems that all previous limits are off the table, and that the political classes are willing to be gross, weird, or just plain ridiculous in order to get jaded young voters to listen to their message.