Page of



(p.381) 14 Conclusion
Network Propaganda
Yochai BenklerRobert FarisHal Roberts
Oxford University Press

Abstract and Keywords

This book has examined how the American political media ecosystem figures in discourses on national politics in general and on presidential politics in particular. It has shown that the internet has no single effect on democracy, news media, or people’s ability to distinguish truth from fiction. Instead, “the internet” is really an integral part of two very different media ecosystems, one of which conforms to the very worst fears of those critical of the effects of the internet on democracy and the other combines attention paid to professional media still pursuing norm-constrained journalism with diverse outlets for mobilization, challenging agenda setting and questioning the mainstream media narrative. These findings suggest that the very introduction of the internet and social media does not itself put pressure on democracy as such, but they also imply that there is no easy fix for epistemic crisis in countries where a hyperpartisan, propaganda-rich environment exists.

Keywords:   media ecosystem, national politics, presidential politics, internet, democracy, news media, journalism, mainstream media, social media, propaganda

THE SECOND DECADE of the twenty-first century has seen dramatic new strains on the democratic project. The 1970s, ’80s, and ’90s saw the demise of the last authoritarian regimes in Western Europe followed by Latin America, Central and Eastern Europe, and to a degree East and Southeast Asia and Africa. By 2018 trends appear to have reversed, and illiberal majoritarian parties or authoritarian regimes are asserting themselves across the globe. As governments, civil society organizations, academics, and media tried to understand what was driving this global change, many focused on technology and technological change. Processes of technological progress that were out of human control were overwhelming our capacity to make sense of the world and govern ourselves as reasonable democracies. The most optimistic feature of our work is that the culprit was not technology.

Technology allowed us to analyze millions of stories published over a three-year period. Technology allowed us to analyze millions of tweets and links, and hundreds of millions of Facebook shares and words to make sense of these stories. And yet, all this technology-enabled research has led us away from technology as the primary explanatory variable of our present epistemic crisis.

Technology is not destiny. Technology interacts with institutions and ideology to shape how we make meaning, how we organize our affairs across economic, political, and personal domains, and how we make our culture and identity.

In the United States, a set of technological innovations rolled out since the 1970s, from FM radio, satellites, and cable, through the personal computer, the internet, and social media, have been adopted and adapted by two fundamentally different epistemic communities in two radically different ways. The civil rights revolution, the women’s movement, and the (p.382) emergence of self-actualization and individual liberty at the heart of both the New Left and the neoliberal moment created a new political reality and a new bifurcation of American polity in terms of how meaning and identity were defined and how these translated into practical political ambitions. On the left, a loose coalition of civil rights advocates, feminists, consumer advocates, environmentalists, and vestiges of organized labor reorganized to challenge an establishment seen as too white, male, and corporate. Some of this reorientation focused on identity; much of it was focused on gaining specific programmatic victories effected through law and policies aimed at achieving better practical outcomes for members of the coalition. On the right, starting in the South, white-identity backlash against the civil rights movement made common cause with the religious right, which itself emerged as a backlash against both the women’s movement’s challenge to the patriarchal family and the secularizing force of putting individual choice at the normative core of markets and morality.1 These two pillars of the emerging conservative coalition were fundamentally about meaning and identity, rather than about programmatic achievement. The pro-business pillar of the emerging conservative coalition was very much about programmatic achievement, particularly deregulation and reduction of taxes, but found ways to translate its programmatic goals into articles of faith—most visibly in Grover Norquist’s tax pledge. That pledge, initiated in 1986, commits politicians who take it to oppose “any and all” tax increases or reduction or elimination of credits, without qualification. Functionally, it has become a precondition to running for office as a Republican. And it transformed the most basic tool for economic and social policy at the heart of the power of the legislative branch into an article of faith divorced from practical considerations and immune to evidence of changed conditions.

The deregulation of cable and elimination of the fairness doctrine in broadcast in the 1970s and 1980s created the institutional conditions for divergent organizational strategies to explore the markets for listeners and viewers. The increased channel capacity provided first by AM radio, then cable, and finally the internet meant that strategies focused on intense engagement of large but still minority audiences became a viable market strategy. Audiences who made up the emerging conservative movement proved a lucrative market for merchants of angry, ideologically pure messages that expressed a shared sense of outrage and loss in the fast-moving, fast-changing world. Rush Limbaugh was the first major commercial success at selling this kind of sentiment since Father Coughlin was forced off the air at the beginning of World War II. Fox News followed, and deregulation allowed (p.383) Clear Channel to consolidate both radio stations and right-wing outrage content into a seamless distribution network for these messages, around the clock, in every corner of the country, to tens of millions of listeners. These outlets offered an ideological coordination point, a cathartic experience of shared anger, and a platform for disciplining political elites who did not hew to the pure message produced by these new right-wing media. Democrats, by contrast, never developed a sufficiently large and homogenous group to form the basis of a similarly successful strategy focused on ideological purity. The coalition the party represented was too diverse to support a single entity like Fox, or Rush Limbaugh in the late 1980s to mid-1990s. Its constituents spread their attention across too many outlets to sustain efforts, like Air America or MSNBC, to replicate the strategy that had succeeded so well on the right.

Thirty years of divergent organizational practices and market dynamics turn out to produce quite different markets, offerings, and consumption habits. When the market in question is the market in informing political beliefs and news, it turns out that this divergence produces very different views of what is going on in the world and who to believe about what is going on.

Our study offers a large-scale observation, sustained over a substantial period of time, of how the American political media ecosystem discusses presidential politics specifically and, by extension, national politics. We found that there is no single effect that the internet has on democracy, or on news media, or on people’s ability to tell truth from fiction. In America, “the internet” is really two very different media ecosystems. One conforms to the very worst fears of those critical of the effects of the internet on democracy. It exhibits all the characteristics of an echo chamber that radicalizes its inhabitants, destabilizes their ability to tell truth from fiction, and undermines their confidence in institutions. The other is closer to the model of the networked public sphere. It combines distinct attention to professional media still pursuing norm-constrained journalism with diverse outlets for mobilizing activists, challenging agenda setting, and questioning mainstream media narratives. This larger part of the American media ecosystem certainly has its own share of propagators of disinformation and commercial clickbait, but these operate in an environment that contains their dissemination and limits their impact.

The differences between the two media ecosystems are palpable. Despite extensive efforts, we were unable to find an example of disinformation or commercial clickbait started on the left, or aimed from abroad at the left, that took hold and became widely reported and believed in the broader network that stretches from the center to the left for any meaningful stretch of time. (p.384) By contrast, as this book demonstrates amply, we found such instances repeatedly succeeding in the right-wing media ecosystem, with pervasive exposure and lasting effects on the beliefs reported by listeners, readers, and viewers within that network. Few if any moments more clearly capture the sheer dehumanization that such a dynamic can birth than the sight of the host of the fourth-most watched television news show in America, Laura Ingraham, tweeting to her two million followers, publicly mocking a 17-year-old high school senior, David Hogg, for not getting into the colleges of his choice. Hogg’s crime, by which he merited public bullying by a Fox News host for the entertainment of her followers, was that he had shown remarkable leadership in the high school students protest movement in favor of gun control in the wake of the Marjory Stoneman Douglas High School shooting.

For those not focused purely on the American public sphere, our study suggests that we should focus on the structural, not the novel; on the long-term dynamic between institutions, culture, and technology, not only the disruptive technological moment; and on the interaction between the different media and technologies that make up a society’s media ecosystem, not on a single medium, like the internet, much less a single platform like Facebook or Twitter. The stark differences we observe between the insular right-wing media ecosystem and the majority of the American media environment, and the ways in which open web publications, social media, television, and radio all interacted to produce these differences, suggest that the narrower focus will lead to systematically erroneous predictions and diagnoses. It is critical not to confound what is easy to measure (Twitter) with what is significantly effective in shaping beliefs and politically actionable knowledge in society. We are only at the very beginning of the ability to create the capacity to engage in such broad, cross-platform research. Television archival data is becoming available; talk radio is still observable only through sporadic transcripts. Facebook data in nonpublic- pages is largely inaccessible. Political ads are not available, but that is changing either through voluntary efforts of companies or through changes in campaign law. We need broad, publicly accessible databases for all these different media so that we can begin to apply the emerging array of data science techniques to a sufficiently broad and diverse set of media to actually represent how people get their news and how people come to learn about the world and understand it.

It is equally critical not to confound whether a phenomenon is observable and whether it actually has an impact. Throughout this book we have argued that many of the concerns over information disorder or the post-truth moment discussed in the past two years focus on what is observable (p.385) and novel, without accounting for how impactful it has in fact been. Russian information operations penetrated American social media in ways that would have been considered the stuff of spy movies in 2012. The evidence supporting the existence of these efforts is real and persuasive, and the recent origin of the organizations perpetrating them suggest that these efforts are in fact novel. The fact that we observe these efforts strongly suggests we need to build capabilities to identify them in real time, and interdict and counter them as they occur. But simply because these efforts are observable and require a response does not mean that they actually had an impact. As we explained in Chapter 8, our own findings, when compared to the publicly disclosed accusations of interference, suggest that these efforts sought to build on and widen already existing fissures in American society but that they appear to have mostly amounted to little more than jumping on a bandwagon already well underway. Similarly, we do not challenge the veracity of some of the excellent journalistic reporting, particularly by Craig Silverman at BuzzFeed news, on the hyperpartisan commercial clickbait. These too are very real and do in fact pollute the environment. We have no systematic reason to think that they could not grow to become an even bigger problem, but our observations at present, and the work of others, suggest that these played a role not fundamentally dissimilar to that of the supermarket checkout-counter tabloid—a side entertainment, not a driver of discourse, opinion, or changes in beliefs at a population level.

Unique in our view among the putative novel suspected contributors to disinformation dissemination is targeted advertising. Here, as is true of the other major suspects, the effect sizes reported in the published science suggest that at least psychographically microtargeted advertising of the Cambridge Analytica variety is highly unlikely to have made a difference in the 2016 campaign. We have no similar evidence with which to confirm or refute the claims that Facebook put Donald Trump in office by selling his campaign the best of their microtargeted advertising capabilities. Nonetheless, the overall trajectory of the entire market in personal data collection and algorithmically delivered, personalized advertising across the online economy suggests that even the small documented effects may well amount to significant effects in the near future. Because effects will be extremely hard to measure, however, both candidate and issue campaigns are likely to invest heavily in these microtargeting techniques, and we will almost certainly not be able to measure their impact until it is too late. For this reason we dedicated a good bit of our solutions chapter to the regulation and transparency of political advertising broadly, not only during electoral campaigns, and believe that this (p.386) is an area where many ounces of prevention are necessary and justified, even in the absence of proof of impact.

Our study is, ironically, both optimistic and pessimistic about the possibilities for democracy in an age of ubiquitously networked communications. It is optimistic because it suggests that the introduction of the internet and social media does not itself put pressure on democracy as such. Different countries, with different histories, institutional structures, and cultural practices of collective sense-making need not fear the internet’s effects. There is no echo chamber or filter-bubble effect that will inexorably take a society with a well-functioning public sphere and turn it into a shambles simply because the internet comes to town. The American online public sphere is a shambles because it was grafted onto a television and radio public sphere that was already deeply broken. Even here, those parts of the American public sphere that were not already in the grip of a propaganda feedback loop and under the influence of hyperpartisan media dedicated to a propagandist project did not develop such a structure as a result of the internet’s development. In fact, after the election of Donald Trump, online communications outside of the right-wing media ecosystem became more mainstream-focused, not less; and this was true even on Facebook, the most polluted of online platforms. These observations suggest that professional journalism continues to play a critical role in anchoring public debate in facts and evidence-based norms and that it functions within a vibrant network of nontraditional sites that constitute a more decentralized, participatory networked public sphere that can work around and through its interactions with the mainstream to diversify expression, counter some of the failure modes of the mainstream, and make mobilization more democratic.

But that happy image goes only as far as a country’s politics and institutional history permit it. The pessimistic lesson of our work is that there is no easy fix for epistemic crises in countries where a politically significant portion of the population does occupy a hyperpartisan, propaganda rich environment. Regulation of, or self-regulation by, platforms can help deal with some of the commercial pollution effects. National security identification of foreign propaganda campaigns may be able to help cleanse public debate from them, although at the cost of substantial surveillance of the major platforms of public discourse. If, however, the origins of crisis are primarily domestic and comprised of intentional manipulation of large parts of the population by its own political leaders and preferred media, there is little that technocratic solutions can do consistent with a commitment to free expression. Here, the hard work of containing disinformation and hate-provoking messages falls (p.387) heavily on those media most trusted by people who are still tuned in to the majority of the network but are at risk of falling into the disinformation vortex. Emphasizing truth seeking rather than neutrality in journalism may help. Regulators might seek to limit the market reach of any given platform, to avoid making any platform a single point of failure that can be used by a disinformation campaign. Public funding for reliable professional media and an educated public may help as well. We embrace these diverse measures because some of them may, in fact, begin to roll back some of the uncertainty and distrust. But if the fundamental problem has deep political roots and takes a political shape, it is hard to imagine that it will be solved by technocratic rather than political and cultural means.

Breathing new life into the truth-seeking institutions that operate on reason and evidence would require a revival of the idea that science, scholarship, journalism, law, and professionalism more generally offer real constraints on what one can say and do, and that they are not all simply modes of legitimating power. This is unimaginable without an underlying shift in political culture, a shift that would require either a successful internal effort from leaders on the right to extricate their party and its base from the propaganda feedback loop or a series of electoral defeats that would force such a transformation. The former is unlikely without the latter. These political and cultural developments will have to overcome not only right-wing propaganda, but also decades of left-wing criticism of objectivity and truth-seeking institutions. Developing such a framework without falling into high modernist nostalgia is the real answer to the threat of a post-truth era. (p.388)


(1.) Yochai Benkler, “Network Pragmatism: Towards an Open Social Economy,” in Towards a Participatory Society: New Roads to Social and Cultural Integration, ed. M. Archer, P. Donati, and M. Sánchez Sorondo, Proceedings of the Pontifical Academy of Social Sciences, Acta 21, Vatican City, 2017.

Copyright © 2021. All rights reserved.
Privacy Policy and Legal Notice