The following is an essay written by a SSRC Social Data Research Fellow, based on their project. The program was generously funded by the Omidiyar Network.
Are we using the right categories to think about social media? Can we use the same categories to think about all social media, or does each platform require distinct categorizations?
Even framing the question this way makes it clear that the answer to both questions is no, but there are reasons why we might think otherwise. For pretty much all adults today, broadcast media continues to define the concepts through which we perceive new media, such as the social internet. However, if your primary understanding of YouTube is as a repository for uploaded Tucker Carlson clips, you’re missing the big part of the iceberg. The problem is especially bad for a platform like YouTube, which combines some elements that seem familiar, and others that are entirely novel. A content creator sitting alone and looking directly into the camera seems similar to a news anchor. However, unlike television, the video itself competes for the viewer’s attention with an interactive comment section in which communities form, and a massive repository of other videos to watch.
To understand how YouTube fits into our broader media environment, then, requires a willingness to accept that “YouTube” is not a single, coherent phenomenon. There are too many kinds of actors for them to all be understood through just one framework. Therefore, it is important to consider each YouTube channel on two dimensions: are they personality-driven—i.e., run by a single person—or are they affiliated with a larger organization, like a television channel?
Absent these distinctions, most public discourse and indeed even academic research about YouTube is confused. This has important ramifications when we consider how to make sense of the rise of far-right media on the platform. The presence of decently produced videos from a constellation of political agitators for the creation of a white ethnostate—and evidence of an active, engaged audience for those videos—shocked many people across the mainstream ideological spectrum. Even the platform they appeared on seemed surprising: YouTube? The place where I watch recaps of Saturday Night Live? Although there has been plenty of news media coverage since that initial burst of attention, I think that most people who prefer to get their news in written form—which likely includes most academics, journalists, and readers of MediaWell posts—still don’t grasp what’s at stake for the way that political ideas take form and disperse on the platform.
To understand these dynamics, we need to take a few steps back in order to understand the unique constellation of actors on the platform itself. There is some ambiguity in terms that makes it difficult to talk about “YouTube Politics.” Most YouTube users never encounter political influencers; in fact, a significant number of YouTube users only watch music videos and sports highlights, and never even watch the videos of any person who creates content specifically for YouTube. Someone who watches those clips might have said that they used the platform pretty often, but they might have no idea about what other people use it for.
In our large-scale, longitudinal analysis of YouTube politics, my co-authors and I aim to help researchers, journalists, politicians and the average non-YouTuber understand what is going on on the red app. One aspect of our research is to formalize the distinction between “inner” and “outer” YouTube proposed by NYT reporter Kevin Roose. “Outer” YouTube comprises content that was created for television and is uploaded to YouTube to garner more viewers. “Inner” YouTube, however, is where the parasocial relationships between audience and creator are cultivated and communities of fans hang out. This has implications for how political ideas and ideologies circulate on the platform and gain adherents.
In the context of our larger project to understand supply and demand for political information on YouTube, we created a typology of YouTube channels. On both the supply and demand side, these channels differ on two key dimensions: whether they are institutionally affiliated or not, and whether they are personality-driven or not. The economics of video production, revenue generation and ideological flexibility vary on both of these dimensions. This lens allows us to investigate both the various techniques these channels use to actually generate revenue and the position these channels occupy in the larger digital media ecosystem.
YouTube politics is divided along multiple correlated dimensions. The ideology of the channel, whether the channel is personality-driven, and the institutional affiliation of the channel are independently important, but in ways that are mutually reinforcing. Crucial for our analysis is the fact that there exists a cluster of right-wing channels that also tend to be personality-driven and unaffiliated with an institution: analyzing only one of these dimensions fails to capture the full dynamic. We therefore classify YouTube channels along these three, interrelated dimensions:
Ideology: Channels were classified into an intuitive left-center-right ideological spectrum. In total, our dataset contains 939 left channels (32%), 726 center channels (25%), and 1,267 right channels (43%). This imbalance reflects the overall distribution of political content on YouTube as a whole; it was even more right-leaning a few years ago.
Personality: We classified channels as personality-driven if they feature “a recurring host and orient their channel’s content around the host’s personality.” In total, 2,729 channels (93%) were personality-driven in some form, and 205 (7%) were not.
Organizational Ties: Channels were classified as affiliated with an organization if the host had either well-known or explicit ties in the past to legacy media outlets, advocacy groups, university lecture series, and/or party organizations, or if the channel itself represents such an organization. In total, 573 channels (20%) were coded as organizationally affiliated and 2,361 (80%) were not.
With this typology in hand, we can explore how these different types of channels engage with different media outlets and social media platforms and what that means for the larger media ecosystem. We analyzed the transcripts of millions of videos, searching for keywords that referenced other actors on the platform. Mapping these references showed us that the divide in YouTube politics reflects the larger divide in political media as a whole, with independent, personality-driven channels referring to hyper-partisan media more often. In the current environment, right-wing channels account for most mentions of these hyperpartisan news outlets. Affiliated channels are better integrated with the mainstream media, where you are less likely to see many references to hyperpartisan news sites, regardless of ideological leaning. We believe that this is not a coincidence: the aesthetic of the charismatic, personality-driven YouTube influencer is more aligned with the vibe of hyperpartisan news sites than with the staid style of established media. In addition, affiliated channels seek to grow their audiences by avoiding hyper-partisanship, just like ad revenue-supported legacy media
Moving to other parts of the media ecosystem, we also searched for references to other social networking sites. Personality-driven channels make more references to social media, like Twitter and Facebook, than non-personality-driven channels—they are better integrated into the overall social media environment. Furthermore, for many creators, YouTube serves as the “top of the funnel” for connecting with new viewers. YouTube is an effective gateway to a larger political, social media ecosystem.
We argue that the specific characteristics of unaffiliated channels shows their importance in shaping political ideas on the platform. These unaffiliated YouTube channels are forming close ties to newer, less-established social media sites. YouTube is still the best place to find an audience, but these YouTubers are also paying attention to Twitch, Steam, Reddit and Telegram. These other platforms are where their communities of fans gather, and where affiliated channels maximize their number of deeply invested fans. By selling merchandise, offering Patreon subscriptions, and accepting live donations, these channels can extract significantly more money from their most devoted viewers.
There is simply no good analogy to the concepts of broadcast media that can help us understand the phenomenon of “YouTube Politics.” Breaking YouTube down according to categories that emerge organically from how the platform is actually used, however, demonstrates how distinct parts of YouTube can be understood in the context of the media environment of 2022. We believe that this confluence of existing media institutions, personal audience recruitment and retention methods, and novel monetization models, produces a new kind of politics. We are still developing our understanding of exactly how this new “YouTube Politics” works, which we believe is essential for understanding American politics more broadly in the decade to come.