Samuel Woolley, Manufacturing Consensus: Understanding Propaganda in the Era of Automation and Anonymity, New Haven, CT: Yale University Press, 2023, 232 pp., $38.00 (hardcover).

 

Reviewed by

Chang Sup Park

The University of Oklahoma

 

Samuel Woolley’s (2023) Manufacturing Consensus: Understanding Propaganda in the Era of Automation and Anonymity serves as a seminal ethnographic exploration of the digital tools—specifically social bots, sockpuppets, and coordinated algorithmic campaigns—that have fundamentally restructured the landscape of political influence. Drawing on over eight years of fieldwork, interviews, and direct observations of “bot builders” and political strategists, Woolley shifts the scholarly focus away from merely “counting” bots toward understanding the human agents who leverage technology to distort public discourse. His core thesis is that the era of automation and anonymity enabled by the Internet and social media has ushered in a new, pervasive mode of manipulation: “manufacturing consensus.”

 

Woolley positions his work as both a tribute to and an evolution of the propaganda theories established by 20th-century scholars such as Walter Lippmann (1922/1997), Edward Bernays (1928/2005), and later, Edward Herman and Noam Chomsky. The title itself is a direct riff on Herman and Chomsky’s (1988) famous work, Manufacturing Consent. However, Woolley argues that the digital era requires a significant shift in terminology and conceptualization.

 

In the broadcast era, “manufacturing consent” described the top-down engineering of public approval by elite-controlled mass media. Propaganda was often viewed by scholars (e.g., Walter Lippmann, Edward Bernays) as a necessary, if elitist, tool for managing collective attitudes in a functional democracy. In contrast, the Internet fosters a “many-to-many” ecosystem where the power to spread propaganda has been “democratized.” This democratization does not imply equality; rather, it means that state actors, mercenary PR firms, and ordinary individuals can now utilize sophisticated tools to amplify their own messages while suppressing opponents.

 

Woolley portrays contemporary “computational propaganda” (the use of algorithms and automation over social media) as inherently illiberal and nefarious in its effects. While 20th-century propaganda sought to build genuine approval, “manufacturing consensus” involves creating an “illusion of popularity” to trigger bandwagon effects. By making fringe or manufactured ideas appear as the dominant public will, propagandists shape opinions and behaviors through perceived social pressure.

 

A cornerstone of Woolley’s contribution is his amendment of Herman and Chomsky’s (1988) “propaganda model.” The original model proposed five filters (media ownership, advertising, sourcing, “flak,” anticommunism/fear) that constrained the news media. Woolley introduces a crucial sixth filter: the automated and anonymous manipulation of social media. This new filter accounts for how information is “laundered” through bots and sockpuppets to game the algorithms of platforms such as Facebook, X, and Telegram. These tools do more than just spread misinformation; they create “informational chaos.” By flooding the digital square with noise and polarized content, propagandists make it difficult for citizens to distinguish organic public opinion from manufactured trends.

 

Woolley identifies three interconnected levels through which this consensus is fabricated, forming what Woolley metaphorically describes as an “ouroboros” of manipulative information. The first level is tool-based manipulation, which involves the direct use of computational tools, such as political bots (automated profiles) and sockpuppets (fake profiles managed by humans), to mimic authentic human interaction. These tools are used to artificially inflate metrics (e.g., likes, shares, follows), to lend a candidate or cause the illusion of large-scale support.

 

His second level of manipulation is algorithm-based manipulation, which refers to the tactical gaming of platform architectures. For example, “growth hackers” (people who specialize in exploiting social media platforms’ technical and algorithmic systems to artificially amplify political content and create the illusion of popularity) may use “phone farms” to bypass bot detection and force specific hashtags into “trending” sections. This forces the platform’s recommendation engine to show the content to real users who were not part of the initial campaign. Woolley’s last level of manipulation is media-based legitimation, which describes the process by which traditional news organizations inadvertently legitimize propaganda. When journalists report on trending topics or use bot-driven metrics as a proxy for public sentiment, they provide “oxygen” to the manufactured narrative, spreading it further into the broader media ecosystem.

 

What distinguishes Manufacturing Consensus from purely quantitative studies is Woolley’s commitment to “reflexive” ethnography. He provides vivid portraits of the professionalized industry where influence is a commodity. Readers are introduced to notable figures, including Carlos Merlo, the “King of Mexican Fake News,” and “Hernan,” a growth hacker who rents out the profiles of rising influencers to political campaigns.

 

Woolley also draws from his experience at “Jigsaw” (Google’s incubator) to analyze the tech industry’s internal struggles. He observes that while tech companies often view propaganda as a technical “bug” to be fixed with better code, it is actually a sociotechnical phenomenon deeply embedded in the platforms’ advertising-driven business models. Tech giants are often “treading water,” reacting to evolving tactics involving partisan nanoinfluencers rather than addressing the structural affordances of anonymity that enable them.

 

Woolley argues that the success of modern propaganda should not be measured solely by whether it changes a person’s vote. Often, the goal is “apathy” or “societal discontent.” By creating an environment where lies and conspiracy spread faster than truth, propagandists today undermine the shared reality necessary for democratic function.

The book cites harrowing examples of manufactured consensus as a tool of state-sponsored harassment and real-world violence. This includes the use of Facebook to incite genocide in Myanmar, the information chaos surrounding the 2016 and 2020 U.S. elections, and the role of encrypted messaging apps in inciting violence in India. In these contexts, manufactured consensus is not a victimless marketing trick; it is a direct threat to human rights and democratic resilience.

 

The book’s chapters build progressively to provide a comprehensive view of the field. Chapter 1 introduces the intersections of propaganda, social media, and political bots. Chapter 2 develops the core concept of manufactured consensus. Chapter 3 examines state-sponsored computational propaganda. Chapter 4 explores the rise of automated political influencers. Chapter 5 discusses the commercial propaganda firms and the tech industry’s internal dynamics and responses. Chapter 6 focuses on journalism and political bots, analyzing how reporters and news organizations interact with and sometimes amplify automated manipulation. The Conclusion offers provocations, provisional solutions, and reflections on the future of countering manufactured consensus.

 

This work is particularly useful for interdisciplinary audiences, including scholars in communication, political science, and information science. Policy makers and activists will find value in the pragmatic discussions on platform governance and media literacy. While the academic tone may require familiarity with certain concepts like “astroturfing,” the accessible storytelling makes it valuable for any reader interested in social media’s societal impact.

 

While Woolley’s ethnographic approach is a powerful antidote to “number-crunching” studies, it does present certain limitations. His refusal to focus on quantifying bot activity sometimes makes it difficult to assess the relative scale of these campaigns compared with organic political movements. Propaganda’s effectiveness often stems from sheer amplification, and some readers may find that vivid anecdotes could benefit from triangulation with quantitative bot-detection data to strengthen generalizability. Furthermore, the book focuses primarily on established platforms like Facebook and X. There is comparatively little discussion of emerging platforms such as TikTok, the expanding influence of generative AI and deepfakes, and the complexities of encrypted messaging apps, though Woolley does briefly acknowledge them.

 

The solutions chapter remains necessarily provisional, given the evolving nature of computational propaganda. Woolley underscores that digital tools have dramatically increased the volume and potency of propagandistic activity, arguing that analyzing the quality of such production is just as crucial as attempts to measure its quantity.

 

Overall, Manufacturing Consensus is a rigorous and timely provocation that bridges propaganda’s historical roots with its digital mutations. It reminds us that technology is not a neutral force; it amplifies human intent, often for the sake of profit or power. By humanizing the actors behind the automation, Woolley provides an unsettling but illuminating look at the “primordial ooze” of the Internet. The book ultimately argues that while technology facilitates these harms, the solutions are not purely technical. It calls for a unified understanding of modern propaganda that considers the people, organizations, and industries involved. For those grappling with the Internet’s dual promise and peril for public discourse, Woolley’s work is useful reading.

 

References

 

Bernays, E. L. (2005). Propaganda. New York, NY: Ig Publishing. (Original work published 1928)

 

Herman, E. S., & Chomsky, N. (1988). Manufacturing consent: The political economy of the mass media. New York, NY: Pantheon.

 

Lippmann, W. (1997). Public opinion. New York, NY: Free Press. (Original work published 1922)

 

 

 

Copyright © 2026 (Chang Sup Park, [email protected]). Licensed under the Creative Commons Attribution Non-commercial No Derivatives (by-nc-nd).

 

https://doi.org/10.65476/dznjz583