Algorithmic inference, political interest, and exposure to news and politics on Facebook

ABSTRACT

The visibility of news and politics in a Facebook newsfeed depends on the actions of a diverse set of actors: users, their friends, content publishers such as news organizations, advertisers, and algorithms. The focus of this paper is on untangling the role of this last actor from the others. We ask, how does Facebook algorithmically infer what users are interested in, and how do interest inferences shape news exposure? We weave together survey data and interest categorization data from participants’ Facebook accounts to audit the algorithmic interest classification system on Facebook. These data allow us to model the role of algorithmic inference in shaping content exposure. We show that algorithmic ‘sorting out’ of users has consequences for who is exposed to news and politics on Facebook. People who are algorithmically categorized as interested in news or politics are more likely to attract this kind of content into their feeds – above and beyond their self-reported interest in civic content.

KEYWORDS:

• Social media
• political communication
• customization
• algorithmic inference
• political exposure

Acknowledgements

We thank the anonymous reviewers for their feedback, as well as colleagues at the University of Michigan and at the GESIS-Leibniz Institute for the Social Sciences, where this work was presented in its early stages. We are grateful to the participants in this study who generously shared their Facebook data with us.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Kjerstin Thorson is an associate professor in the College of Communication Arts & Sciences at Michigan State University. Her research focuses on how people use digital and social media to learn about and participate in politics, especially youth and young adults. She also explores how political messages spread across digital media, and how the social sharing of media messages is shaping the way we see the political world.

Kelley Cotter is a PhD student in Information and Media at Michigan State University. Her research focuses on how users become aware of and knowledgeable about algorithms applied in sociotechnical systems and resultant implications for the accumulation of different forms of capital. She also explores the intersection between algorithmic (il)literacy and digital inequalities.

Mel Medeiros is a PhD student in Information and Media at Michigan State University. Her research focuses on political communication, digital politics, social media, and the platformization of politics.

Chankyung Pak is a postdoctoral researcher in Amsterdam School of Communication Research at Universiteit van Amsterdam. His research interests lie in computational social sciences as applied to the study of transforming Internet media ecosystems and their impact on formation of public opinion. Recently, he investigates the economic base of extreme ideas or disinformation.

Notes

1 These topics do not represent ads individuals have seen, but rather topics they may be interested in. Facebook does not provide any information about ad or newsfeed content exposure, but does provide information about ads users have interacted with previously in “Ads History” and in a list of “Advertisers you’ve interacted with.”

2 The number of liked pages related to news and politics is highly correlated with the number of ad topics related to news and politics assigned to each user. However, the name of the actual pages liked and what is listed in the ad categories are often not the same — algorithmic translations occur frequently.

3 This assumption of control is implicit, suggested through the ways researchers measure selective exposure: the studies cited above and many others rely on measures of clicks, time spent reading, selecting articles they wish to read, listing sources commonly used, and engagement on articles. These measures are based on the premise that selection and exposure occur simultaneously, indicating that through their selection choices, individuals directly control their media exposure.