Knüpfer, C. B., Yang, Y., & Cowburn, M. (2026). The logic of connective faction: How digitally networked elites and hyper-partisan media radicalize politics. Political Communication, 1–33. https://doi.org/10.1080/10584609.2025.2604708
Summary
This article develops the concept of a “logic of connective faction” to explain how ideologically extreme subgroups within a political party build connective ties to hyper-partisan media ecosystems, strengthening their factional position and driving intra-party radicalization. Extending Bennett and Segerberg’s connective action theory from social movements to formal party organizations, the authors use the sudden emergence of “Critical Race Theory” (CRT) as a right-wing talking point in 2020–2021 as a case study. Drawing on nearly 2 million news articles, congressional newsletters, and tweets, they compare Republicans who adopted the CRT issue with those who did not, finding that “CRT Republicans” are more ideologically extreme, more deeply embedded in right-wing digital networks, and rewarded with greater engagement and partisan media attention. The paper argues that these networks function as quasi-organizational “digital surrogates” that co-produce factional agendas alongside party elites.
Key Contributions
- Introduces the “logic of connective faction” framework, extending connective action theory to intra-party factional dynamics.
- Develops the concept of “network pings” — discursive triggers (like CRT) that activate latent factional ties between elites and digital surrogates.
- Provides an original large-scale, multi-platform dataset (n ≈ 1.94M) combining mainstream and right-wing news, congressional newsletters, and tweets.
- Demonstrates empirically how digital connectivity, ideological extremity, and issue adoption co-occur and reinforce one another within a single party.
- Bridges literatures on factionalism, hybrid media systems, radicalization, and democratic backsliding, offering a tool transferable beyond the US case.
Methods
The study combines theoretical framework development with a mixed empirical dataset spanning January 2019 to January 2022. News articles from mainstream (NYT, CBS, CNN) and right-wing outlets (Fox News, Breitbart, Daily Caller) were collected via NewsWhip; congressional newsletters came from the DC Inbox dataset; and tweets from members of the 116th and 117th Congress were drawn via Twitter’s Academic API. CRT mentions were identified with a dictionary approach and manual validation. The authors compared “CRT Republicans” against “Non-CRT Republicans” using ideological proxies (DW-NOMINATE, CFscores, the anti-McCarthy speaker vote, House Freedom/Liberty Caucus membership) and connectivity measures (references to right-wing outlets, hyperlink-based ideological scaling, Media Bias Fact Check left-right ratios, links to questionable sources, and alt-tech platform links). Statistical analysis used negative binomial regression with member fixed effects and two-tailed t-tests.
Findings
- CRT entered right-wing news sites (notably Breitbart) before mainstream and elite sources, spiking in spring/summer 2021 and around the November 2021 Youngkin campaign.
- Over 100 Republican members of Congress adopted CRT in newsletters; only three Democrats ever used the term.
- CRT Republicans scored significantly higher on all four ideological extremity measures.
- CRT Republicans showed significantly higher engagement with right-wing media across all four connectivity indicators and disproportionately linked from official websites to alt-tech platforms (Rumble, Truth Social, Gettr).
- Tweets mentioning CRT received significantly more retweets, replies, likes, and quotes than the same member’s other tweets.
- Articles about CRT gained more Facebook engagement when published by right-wing sites but less when published by mainstream outlets.
- Right-wing sites mentioned CRT Republicans more often, while mainstream outlets showed the opposite — an attention asymmetry that rewards factional behavior.
Connections
This paper’s account of hyper-partisan right-wing media ecosystems and elite radicalization connects closely to work on partisan news audiences and echo chambers such as Bakshy2015-rn and Gonzalez-Bailon2024-rq, and to studies of the broader right-wing information ecology and questionable-source diffusion like Marwick2025-ov and Marwick2025-vx. Its focus on issue diffusion and diverging engagement dynamics across platforms resonates with computational studies of political discourse and virality such as Tornberg2025-ir and Brady2026-ln. The emphasis on democratic backsliding and alt-tech platform migration also links to work on fringe and alternative platform ecosystems.
Podcast
A research-radio episode discusses this paper: 🎧 MP3 · Spotify · Apple Podcasts