Manufacturing Consent

There's a puzzle at the heart of media criticism: both the Left and the Right are absolutely certain the media is biased against them. The Right points to journalists who vote Democrat four-to-one and donate overwhelmingly to liberal causes. The Left — specifically, Noam Chomsky and Edward Herman in their 1988 book Manufacturing Consent — argues that the media systematically serves the interests of the powerful despite the liberal leanings of individual journalists. The strange thing is that both sides have a point, and understanding why requires seeing that they're talking about different kinds of bias operating at different scales.¹

The Propaganda Model

Chomsky and Herman propose five structural filters that shape media output regardless of individual journalists' politics. The first is ownership: mass media is controlled by large corporations, so it naturally favors the kinds of things large corporations favor. The second is advertising dependence — Texaco isn't going to buy ad space in a publication that attacks Big Oil. The third is sourcing: journalists rely on official, convenient sources (the Pentagon, the White House, local police) who enjoy a presumption of trustworthiness, and the most credentialed "experts" are drawn from institutions with pro-establishment views. The fourth is "flak" — conservative organizations fund relentless nitpickers who shriek about every deviation from the official line, grinding journalists into unconscious compliance. The fifth is anti-communism (or, updated for the post-Cold War era, anti-terrorism) as a master ideology that can be deployed to smear any dissenter.¹

What makes this framework compelling isn't any single filter but how they reinforce each other. A journalist who parrots official sources can be sloppy and nobody calls them on it. A journalist who departs from the establishment line faces both sourcing difficulties and a barrage of flak. The result is that the range of acceptable discourse narrows to what Chomsky calls the "bounds of thinkable thought" — not by conspiracy, but by structural incentive.

Death Squad Democracy

The most powerful evidence in Manufacturing Consent comes not from domestic media analysis but from a pattern Chomsky and Herman document across the Third World. The pattern is horrifyingly consistent: the United States topples a government (usually because leftists were doing well in elections), installs a weak dictator, who then faces guerrilla resistance, spawns death squads to suppress it, and receives ever-increasing American military aid to deal with the resulting chaos. The dictator holds an election from which all real opposition is banned. The US media reports this as a "step toward democracy." Meanwhile, tens of thousands die.¹

The media's role in this isn't direct lying. It's emphasis and framing. "Worthy victims" killed by America's enemies (like Polish priest Jerzy Popieluszko, murdered by Communists) receive saturating coverage, outrage, and demands for justice. "Unworthy victims" killed by America's allies (like the thousands murdered in El Salvador and Guatemala) get a shrug and "violence sometimes happens." The technique of "legitimizing versus meaningless elections" ensures that sketchy votes in US-allied countries are praised as democratic progress while cleaner elections in US-enemy states are dismissed as shams.¹

The most revealing case study is Vietnam, usually considered the moment the media broke with the establishment. Chomsky argues this happened only within an extremely narrow window. The two acceptable positions were "the US is right to fight for the freedom of South Vietnam" and "the US is right to fight for South Vietnam, but the costs are too high." The position that nearly everyone in South Vietnam supported Ho Chi Minh, and that the US was essentially waging war on the South Vietnamese people themselves, was never expressed in mainstream media.¹

The Deeper Pattern

What's unsettling about Chomsky's work isn't the specific claims — many of which are now conventional history — but the mechanism. In a 1977 interview, Chomsky observed that the United States exercises tighter ideological control than many explicitly authoritarian states, despite having no secret police or concentration camps to speak of. During the Franco period, Spanish intellectuals visiting America were surprised by how narrow the range of political discourse was. "If a rational Fascist dictatorship were to exist," Chomsky mused, "it would choose the American system. State censorship is not necessary, or even very efficient, in comparison to the ideological controls exercised by systems that are more complex and more decentralized."²

This connects to his analysis of Watergate as a supposed "triumph of democracy." Nixon wasn't condemned for using reprehensible methods — similar methods had been used against minorities and social movements for decades. He was condemned for using them against people with power: the chairman of IBM, the Washington Post, senior Democratic donors. "Men of power against men of power." The real abuses of state power — the FBI's assassination of Black Panther leaders, COINTELPRO's systematic destruction of civil rights organizations — received no comparable reckoning.²

The student movement of the late 1960s briefly opened the ideological space. Revisionist historians challenged the orthodox Cold War narrative, and the orthodox position "simply dissolved, vanished" the moment it was seriously examined. But once the pressure of the student movement eased, the establishment intelligentsia worked systematically to close these breaches, describing the modest opening as "totalitarianism of the left" and the "SS marching through the corridors."²

Where the Model Breaks

All that said, the propaganda model has real limitations. Chomsky and Herman had complete control over which incidents to include, and they used it to select genuinely troubling cases while ignoring ones that cut against their thesis. The comparison of media coverage of the Kurdish genocide (by US enemy Iraq) versus the East Timor genocide (by US ally Indonesia) supports their argument — but including Israel (US ally, massively covered) or Iran (US enemy, barely covered) would muddy it considerably.¹

More fundamentally, the model struggles with the toxoplasma-like dynamics of modern media, where controversy propagates not because powerful interests select for it but because the information ecology rewards outrage regardless of who benefits. The structural filters Chomsky identified in 1988 still operate, but they now compete with algorithmic amplification of whatever generates engagement — a force that is often anti-establishment in the short term, even as it serves establishment interests in the long term by making coherent opposition impossible.

The propaganda model also doesn't account well for the genuine agency of journalists, who are not merely vessels for structural forces. The model predicts that the media will always serve power, but the history of investigative journalism includes enough real counterexamples — the Pentagon Papers, Watergate reporting, coverage of Abu Ghraib — to suggest that the filters are powerful but not total. The question is whether these exceptions are the system working or the system failing, and that depends on what you think the system is for.

What Chomsky gets right, and what connects his work to Legibility And State Power, is the insight that the most effective control is invisible even to the controllers. The editors and journalists who internalize the bounds of acceptable discourse don't experience themselves as constrained. They experience themselves as exercising professional judgment. The propaganda model is most powerful not as a description of conspiracy but as a description of how rational individuals, each acting in their own perceived interest, can collectively produce outcomes that serve power without any individual intending that result.