How are social media platforms and technology companies changing the future of journalism? Companies like Facebook, Google and Twitter, which started out as distribution channels, are now edging out news organizations in producing fast and shareable content. This trend raises questions about how to preserve high-quality journalism and infuse greater transparency and accountability into platform companies, particularly after the “fake news” revelations of the 2016 election. Taylor Owen and Emily Bell tackle these issues in their report, “The Platform Press: How Silicon Valley Reengineered Journalism.”
Owen and Bell focus on the convergence between publishers and platforms, arguing that technology companies have conquered the markets for attention and advertising. To test this theory, they conducted more than 70 interviews, performed an extensive content analysis and hosted two private roundtables. Based on this research, they found that low-quality content is consistently beating out journalism with high civic value. The Tow team argued that this progression of publishers into the social world has created two types of news organizations: one that maintains and develops its own presence and one where publishing is no longer the primary activity used to support journalism.
Owen and Bell’s research is part of the Platforms and Publishers research project at the Tow Center for Digital Journalism at Columbia Journalism School. Owen is a Senior Fellow at Columbia and an Assistant Professor of Digital Media and Global Affairs at the University of British Columbia. He was previously the Research Director of the Tow Center, where he led a program studying the impact of digital technology on the practice of journalism. His work can be found at www.taylorowen.com and @taylor_owen.
In what ways did the 2016 election shape your interest in and approach to this project?
Emily Bell and the research team for the Tow Center’s Platforms and Publishers research project started studying the relationship between journalism publishers and social media platforms in 2015, well before Donald Trump had a hope of the nomination. This early research began to establish the argument that core functions of the publishing process were being absorbed into a small number of platform companies. Emily and I began working on the broader report in the fall of 2016, at the height of the panic in the news industry over the rise of disinformation on social platforms. The election was very much on our minds; the scholarly and public discourse over media and the election were evolving so rapidly that we repeatedly delayed putting pen to paper.
The election also shaped the report in a more fundamental way, however. We decided to focus on Facebook, particularly the iterative changes in its advertising model and newsfeed algorithm over the year leading up to the election. We found a pretty clear correlation between changes to the make-up of Facebook as a platform and the increasingly problem of misinformation. It became clear to us that the challenges posed to journalism were not a distinct problem that could be addressed with better journalistic practice or more sustainable business models. They were structural problems with the digital information ecosystem itself, particularly the pernicious effect of the advertising technology industry and the opaqueness of the algorithms determining content. The report became a study not just of the changing nature of journalism publishing, but of also of the health of the broader civic divorces in which journalism is situated.
What recent changes in the news ecosystem encourage the spread of low-quality content? What are some of the unique challenges this poses to journalists?
First, it should be noted that there has always been a large proportion of low-quality information in our information diets. Journalists like to believe high-quality accountability journalism was primarily consumed, when really it was often tabloid content. The Internet just allowed us to measure this trend far more accurately.
However, the past year has revealed a central failure in the social media ecosystem, particularly on Facebook, the dominant online content distribution platform. The problem is that anyone – lobbyists, politicians, propagandists, non-profits, businesses or journalists – can directly reach highly specified audiences via social platform targeting.
In the report, we identify three key variables that have enabled this particular phenomenon. First, the evolution of Facebook ad products and the highly detailed personal data targeting audiences has created an ecosystem ripe for abuse. The widespread surveillance of user activity has been combined with the power to influence what media consumption patterns. Second, while access to audience can be purchased, exactly who sees what is governed by opaque algorithms. The nature of the information ecosystem on Facebook is derived from a complex and ever changing set of rules. The biases and incentives that are embedded in this computational system are unknown, as they are hidden from view, but their very nature is often unknowable even to those that originally program them. Third, the way in which information is displayed on Facebook creates a relativism between categorically different content types. The design of the platform displays a false story with the exact same feel of authority as an article by a legitimate news organization. This makes it very hard for the audience to make qualitative distinctions between information types.
What are some of the long-term implications of this “third wave” of journalism for local news organizations?
Local news organizations face a particularly daunting challenge in the current digital ecosystem. First, both the legacy organizations that have managed to successfully transition to the platform economy, as well as the digitally native organizations that have built a model from the ground up, do so at scale. They leverage social platforms to boost their brands, to drive traffic and build ad revenue. It is unclear to me how this model can be leveraged by local news organizations who by definition can’t scale. Second, as we show in the report, engaging on a wide range of platforms is highly labor intensive. A small local organization simply cannot produce custom content for a dozen platforms with a single social media staff.
What is next on your research agenda?
I am particularly interested in three research questions, all of which build from the Platform Press project. First: what are the wider implications of this new platform ecosystem on civic discourse, and on our democratic institutions? While it is important that this ecosystem is causing legitimate challenges for the news industry, the structural problems that we identify also have far wider social consequence. The rise of relativism in our civic discourse is a fundamental challenge to the information needs that sit at the foundation of democracy. Second: I am exploring the algorithmic and increasingly Artificial Intelligence driven computational systems that drive both our media ecosystems and increasingly determine a wide range of public policy. These computational systems sit uncomfortably in our systems of governance, which put a premium on human agency. Finally, I think we need to explore creative and potentially uncomfortable policy solutions for these twin problems – the decline of civic discourse and the dominance of algorithmic agency – and hope that research exploring potential governance paths could help create solutions.