Section III: The Role of Media, News, and Propaganda

If partisanship is the disease, the modern American media environment is the vector. Two structural facts are worth keeping in view as we go through the specifics. First, the boundary between news and opinion has eroded almost everywhere — not just at the outlets you dislike. Second, the trust collapse is symmetric in one important sense: Americans across the political spectrum have lost faith in the press, even if for different reasons.

The Associated Press and United Press International, in the mid twentieth century, were the closest thing the United States had to a shared editorial baseline — wire services that supplied member papers across the political spectrum with reporting that aspired to be usable on the editorial page of the Manchester Union-Leader and the Berkeley Barb on the same day. That baseline has frayed. Critics on the left accuse the AP of "both-sides" framing on issues like climate and election integrity; critics on the right accuse it of soft-pedaling Democratic stories and adopting activist-preferred vocabulary on race and gender. That the AP draws fire from both directions is probably evidence that the wire service is closer to neutral than its critics on either side believe — but the trust gap is real on both ends.

The Bias Landscape

Two organizations have done the most serious public-facing work to map American media bias: AllSides and Ad Fontes Media. Both should be cited together because they use different methods and produce overlapping but not identical results — exactly what reasonable replication looks like.

AllSides (founded in 2012 by John Gable, a former Republican congressional staffer, and Joan Blades, co-founder of MoveOn.org) rates outlets on a five-point scale from Left through Lean Left, Center, Lean Right, and Right. Its methodology combines blind bias surveys — panels of Americans across the spectrum read content with branding removed and rate the bias — with third-party reviews, editorial reviews, and community feedback (AllSides, 2024). The AllSides Media Bias Chart Version 9 (released 2024) places MSNBC, Mother Jones, and The Nation on the Left; The New York Times news, The Washington Post news, CNN, and NPR on Lean Left; Reuters, the Associated Press, and The Christian Science Monitor in the Center; The Wall Street Journal news and Newsmax on Lean Right; and Fox News opinion, Breitbart, Daily Wire, and The Federalist on Right. Critically, AllSides distinguishes between news and opinion divisions of the same outlet — Wall Street Journal news ranks center to lean-right, while WSJ opinion ranks right.

Ad Fontes Media (founded in 2018 by patent attorney Vanessa Otero) uses a two-dimensional chart that rates both bias on a left–right axis and reliability from "Original Fact Reporting" at the top to "Contains Inaccurate/Fabricated Info" at the bottom. The methodology employs panels of analysts who score articles on multiple criteria, with each article scored by analysts from across the political spectrum. The Ad Fontes Interactive Media Bias Chart (Otero, 2024) places Reuters, AP, and BBC at the top of the reliability axis; outlets like The Atlantic and National Review in the upper-middle band with moderate ideological lean; and outlets like Daily Kos, Breitbart, OAN, and Palmer Report in the lower band marked by more opinion content and lower factual reliability. Both organizations explicitly acknowledge that bias is not the same as inaccuracy — a true story can be told from a biased angle, and a centrist outlet can publish errors. The right way to read these charts is as a starting point, not a verdict.

MSNBC's prime-time lineup is opinion programming, not news, and the network does not pretend otherwise — but its tone has shifted notably leftward since 2016. Rachel Maddow, whose show ran weeknights from 2008 to 2022 and then weekly afterward, framed her commentary as investigative analysis with an unmistakable progressive editorial perspective; Chris Hayes, Lawrence O'Donnell, and Joy Reid filled out a lineup whose evening output was consistently critical of Republican policies and personalities. AllSides rates MSNBC as Left — its strongest leftward designation — and Pew Research has consistently found that MSNBC's audience skews Democratic by a 70-to-20 margin or more (Pew Research Center, 2020).

Fox News's prime-time lineup is opinion programming, not news, and the network's defense in the Dominion lawsuit explicitly argued that no reasonable viewer would mistake Tucker Carlson, Sean Hannity, or Laura Ingraham for news anchors. The Carlson era (2016–2023) defined the network's prime-time identity; his replacement Jesse Watters has continued the same posture. AllSides rates Fox News opinion as Right — its strongest rightward designation — and Pew Research has consistently found that Fox's audience skews Republican by a 70-to-20 margin or more, the mirror image of MSNBC's audience (Pew Research Center, 2020).

The Fox News–Dominion Voting Systems settlement is the single most consequential media-bias evidence ever obtained in a U.S. courtroom. In April 2023, on the morning that opening statements were to begin in a Delaware courtroom before Judge Eric Davis, Fox News agreed to pay Dominion $787.5 million — roughly half of Dominion's $1.6 billion ask, and the largest known defamation settlement in U.S. media history (Folkenflik, 2023; Grynbaum & Mullin, 2023). Pre-trial discovery had produced internal Fox texts and emails showing that hosts and executives — including Tucker Carlson, Sean Hannity, Laura Ingraham, and Rupert Murdoch himself — privately did not believe the 2020-election-fraud claims they were broadcasting. Carlson texted on November 10, 2020: "Sidney Powell is lying." On January 4, 2021, he texted that he hated Trump "passionately." Murdoch acknowledged in deposition that he could have stopped his hosts from broadcasting election-fraud claims and chose not to. Fox issued a statement acknowledging "certain claims about Dominion to be false" but declined to apologize or air a retraction. There is no comparable evidence on any left-leaning outlet of comparable magnitude and specificity, and a conservative who cares about being honest has to say so.

CNN's editorial drift between 2016 and 2020 became a topic of internal debate at the network itself. CNN ran BREAKING NEWS banners for hours at a stretch on stories of varying urgency, and ran chyrons that fact-checked Trump's statements in real time on screen — a practice that supporters defended as accuracy-promoting and critics described as editorializing in the news ticker. Chris Licht, who took over as CNN chief executive in 2022, said publicly that the network had become too partisan during the Trump years and tried to course-correct; he was pushed out in 2023 after the Trump town hall and tumbling ratings (Smith & Mullin, 2023). The internal acknowledgment that the network's coverage had drifted is significant.

One America News Network (OAN) and Newsmax pushed election-fraud claims further than Fox and have faced their own legal exposure. Newsmax settled with Smartmatic for an undisclosed sum in September 2024; OAN settled with Smartmatic in April 2024, also undisclosed. OAN's role as an open propaganda outlet — its founder Robert Herring directed editorial coverage favoring Trump — has been documented by The New York Times, Reuters, and former employees.

The 2020 internal revolt at The New York Times over Senator Tom Cotton's June 3, 2020 op-ed "Send in the Troops," urging deployment of active-duty military to control unrest after George Floyd's death, is a documented inflection point. The piece had cleared editorial review. After publication, hundreds of Times staff tweeted that publishing the column "puts Black @nytimes staff in danger." Within days, opinion editor James Bennet was forced to resign (Tracy, 2020). The Times appended a 317-word editor's note saying the op-ed "fell short of our standards and should not have been published." Critics across the spectrum — including liberal columnists like Jonathan Chait — argued that the staff revolt represented a troubling new norm: ideologically congenial viewpoints would be published; ideologically uncongenial viewpoints, even from sitting U.S. senators, would not.

Trust data is grim and symmetric in one important sense. The Reuters Institute Digital News Report 2024 found that just 32 percent of Americans say they trust most news most of the time — the lowest figure among major democratic countries surveyed (Newman, Fletcher, Robertson, Eddy, & Nielsen, 2024). For comparison: Finland 69 percent, Denmark 56 percent, Germany 50 percent, the United Kingdom 36 percent. Pew Research data on the partisan structure of the trust collapse is even more striking. In 2003, Pew's media-trust questions showed Republicans and Democrats reported broadly similar levels of confidence in major news outlets — both in the 60–70 percent range for major networks. By 2024, Democrats reported roughly 70 percent confidence in national news media, Republicans about 14 percent, and independents about 27 percent (Pew Research Center, 2024). That is one of the largest partisan gaps in any social-trust measure Pew has tracked.

The Consequences

When the media environment runs on outrage as its raw material, the consequences show up as physical events with body counts and insurance claims, not just as bad vibes online.

Right-coded lethal incidents in recent years include Tree of Life (11 dead, 2018), El Paso (23 dead, 2019), Buffalo (10 dead, 2022), and the January 6 Capitol assault (five deaths in connection with the events of that day; four subsequent officer suicides). Left-coded incidents include the June 14, 2017 shooting at the GOP congressional baseball practice in Alexandria — perpetrator James Hodgkinson, a 66-year-old Bernie Sanders volunteer from Illinois, nearly killed Majority Whip Steve Scalise, who spent months in recovery (Faulders & Karl, 2017) — and the June 8, 2022 attempt by Nicholas Roske on Justice Brett Kavanaugh.

The Roske case deserves more weight than it usually gets. Three weeks before the Dobbs v. Jackson Women's Health Organization decision was issued, Roske flew from Los Angeles to Washington with a Glock 17, ammunition, a tactical knife, zip ties, pepper spray, duct tape, and tools for breaking and entering. He called 911 from outside Justice Kavanaugh's home and told the dispatcher of his intent to kill the justice and then commit suicide; the motive he described to investigators was the upcoming abortion decision and the Uvalde, Texas school shooting. Roske pleaded guilty on April 8, 2025, and was sentenced on October 3, 2025, to 97 months in federal prison (U.S. Department of Justice, 2025). Prosecutors had asked for 30 years to life. The case stands as one of the most serious individual acts of attempted political violence against a federal official in modern American history.

The October 28, 2022 attack on Paul Pelosi, the husband of then-Speaker Nancy Pelosi, by David DePape in the couple's San Francisco home left Paul Pelosi with a fractured skull from a hammer blow. DePape was convicted in federal court of attempted kidnapping of a federal official and assault on a federal officer's immediate family member; he received a 30-year federal sentence in May 2024 and a separate 7-years-to-life state sentence (Eligon, 2024). Investigators found that DePape held a chaotic worldview that included QAnon-adjacent conspiracy theories, anti-LGBTQ rhetoric, and 2020 election denial — placing this incident on the right-coded side of the violence ledger, though his overall mental-health picture complicates clean categorization.

Threats to federal officials and election workers cover both sides. The U.S. Capitol Police reported 9,625 threat investigations in 2021, the highest figure on record, with threats reported against members of both parties (USCP, 2022). The Department of Justice's Election Threats Task Force reviewed more than 2,000 reports of hostile contact with election workers; about 11 percent met the criminal threshold (U.S. DOJ, 2022). The pattern of threats to election workers was overwhelmingly originating from Trump supporters acting on his stolen-election claims; the pattern of threats to Supreme Court justices and Republican members produced incidents on the other side. Both halves of the picture are real.

Treating Each Organization Honestly

Lazy partisan writing lumps every organization on the other side into a single undifferentiated mass. That habit is its own form of dishonesty, and I want to refuse it here. The following organizations are different from each other in important ways — in age, mission, internal discipline, legal record, and the evidence against and for them. They deserve separate treatments.

The NAACP, founded in 1909, is one of the oldest civil rights organizations in the United States, with a documented role in every major civil rights advance from anti-lynching legislation through Brown v. Board of Education and the Voting Rights Act of 1965. To treat the NAACP as interchangeable with newer or more radical groups is bad history. A conservative can disagree with specific NAACP positions on policing or voter ID without erasing the institution's century of disciplined, lawful, and largely successful civil rights advocacy. I disagree with the NAACP on plenty. I do not pretend it is the same thing as an anonymous activist network whose members wear masks on the street.

The Southern Poverty Law Center, founded in 1971, is a hate-tracking organization with real institutional credibility on extremist watch — and real, sustained criticism from a wide range of observers (including some on the left) about its hate-group designations, which critics argue have at times been applied too broadly to include mainstream conservative organizations. The SPLC should be cited where its methodology is sound and its evidence is specific, and criticized where it has elided the difference between organized hate and conservative policy disagreement. Both can be true at once, and a serious paper says so.

Antifa is not a single organization. It is a decentralized tactic and identity — an antifascist orientation without formal membership, dues, or a national board. That makes it harder to criticize coherently, because there is no "Antifa headquarters" to indict, and harder to defend, because there are no internal accountability mechanisms when self-described antifa actors engage in property destruction or assaults. FBI Director Christopher Wray testified to Congress in 2020 that antifa is "more of an ideology or movement" but that the bureau was actively investigating "violent anarchist extremists" within it. The Portland federal courthouse attacks of 2020 are real. The Atlanta Stop Cop City confrontations of 2023, in which activist Manuel Esteban Paez Terán ("Tortuguita") was killed by Georgia state troopers and a trooper was wounded, are real. Sixty-one activists were later charged under Georgia's RICO statute — a sweeping use of organized-crime law against protest activity that civil liberties groups including the ACLU criticized as overreach. Antifa-coded violence is a real category. It is also a different category from an organized, hierarchical network with a leader, a chain of command, and seditious conspiracy convictions.

Black Lives Matter is two different things that share a name. The BLM movement — the broad social mobilization following the deaths of Trayvon Martin (2012), Michael Brown (2014), and George Floyd (2020) — is a decentralized movement with hundreds of millions of sympathetic participants and a broadly successful civil rights agenda on specific policy questions including chokehold bans and body cameras. The BLM Global Network Foundation is a specific 501(c)(3) organization that raised approximately $90 million in 2020 and subsequently became the subject of sustained financial controversy, including reporting on real-estate purchases and questions about executive compensation that drew criticism from local BLM chapters and outside auditors. The movement and the foundation are not the same thing. Both criticism of the foundation's finances and acknowledgment of the movement's legitimate civil rights claims belong in the record.

The Proud Boys, founded by Gavin McInnes in 2016, were characterized by federal prosecutors at trial as a hierarchical organization with a chairman, a chain of command, and an operational plan for January 6, 2021. The seditious conspiracy convictions of Enrique Tarrio (22 years), Ethan Nordean (18 years), Joseph Biggs (17 years), Zachary Rehl (15 years), and Dominic Pezzola (10 years) are not in dispute and were the result of jury trials in federal court (Hsu & Weiner, 2023). Treating the Proud Boys as a loose drinking club is not honest. They were convicted, at the highest level of their organization, of conspiring to use force against the United States government.

The Oath Keepers, founded in 2009 by Stewart Rhodes, were the subject of the most consequential domestic seditious conspiracy prosecution since the 1995 Oklahoma City bombing investigation. Rhodes was sentenced to 18 years on May 25, 2023, with Judge Amit Mehta applying a terrorism enhancement to the guidelines — the first time a January 6 sentencing court had done so (Hsu, 2023). Florida Oath Keepers leader Kelly Meggs received 12 years the same day. The Oath Keepers' recruitment of current and former military and law enforcement personnel, and the explicit role of their "quick reaction force" stationed in Virginia on January 6, are established facts on the record of multiple federal trials.

Patriot Front, a successor formation to the white nationalist Vanguard America after the 2017 Charlottesville rally, has staged choreographed flash marches in Boston, Washington, Nashville, and elsewhere, with masked formations carrying flags and shields. On June 11, 2022, 31 Patriot Front members were arrested near a Pride event in Coeur d'Alene, Idaho, with riot shields, smoke grenades, and an operational plan drawn up by their leader Thomas Rousseau (Sottile, 2022). Seven were ultimately convicted of conspiracy to riot. Treating Patriot Front as merely "young men with strong opinions" is not honest; it is an organized white-nationalist formation with a documented operational pattern.

These treatments are not equivalent because the organizations are not equivalent. They are, however, parallel in approach: each has been examined as itself, with its actual record, rather than lumped together with anyone else who happens to share a partisan valence. That is what I mean by organization-level honesty.