Chompsky's 'Manufacturing Consent' explains everything that's ever pissed you off about the news

by Dorian Minors

July 10, 2020

Analects  |  Newsletter


Say what you will about Noam Chompsky, he had some bright ideas. The book he co-authored with Edward Herman is often described as ideological. It’s a shame, because when you strip out the strident denounciations of the U.S. government, it’s extremely interesting and helps explain many of the bizarre and annoying trends we see in media today. It’s almost like they don’t want you to know…

filed under:

Article Status: Complete (for now).

Say what you will about Noam Chompsky, he had some bright ideas. He was one of the founders of modern cognitive theory, and has been influential in more academic fields than most people have time to google.

One important contribution he made was his work on the media with co-author Edward S. Herman: The Propaganda Model: Manufacturing Consent.

It’s a book that gets lobbed around academic circles, and talked about in student dorms, but it’s not frequently seen elsewhere. This might have something to do with its academic style. It’s not particularly easy to read and without a familiarity with the language or any kind of peer pressure, it’s very easy to put down. There’s also the fact that Noam Chompsky is a divisive figure. If you don’t like the man, you might not be interested in the message. It’s also not entirely unlikely that there’s an effort to minimise the promotion of the book, or at a minimum, motivation to ignore it. Herman and Chompsky really make an effort to shit all over the media and the U.S. Government. I’m sure neither are enthusiastic about that.

It’s a shame. The model itself is extremely interesting. More to the point, despite its age, it feels increasingly applicable given the total integration of internet and media. If you’ve ever been dissatisfied with the content in your media streams, this model probably explains why.

Since I found it difficult to find good elaborations of the model and how it might be at work today, I thought I may as well visit it myself. We’ll do two articles. The first, this one, is a description that’s a little more detailed than those I’ve found online. The reason being that the nuance of the model which makes it so applicable is often overlooked. The second article, which you can find here, is a piece covering some of the bizarre current media trends the model makes sense of.

The model itself is an economic model more than anything else. It outlines economic and political pressures. The book is littered with denunciations of the media and the U.S., and at times, other governments. While fairly reasonable critiques of the horrific abuses the U.S. has visited on other states and the attendant media complicity, it distracts a bit from the model itself. Stripping those critiques out of the model makes it a little more digestible, particularly for non-U.S. audiences.

One fact that rises quickly to the surface is that the model describes almost accidental properties. The major point is that information is filtered in five ways before it becomes news. Each of these filters means that certain news is more newsworthy than other kinds of news. As the authors put it “the raw… news must pass through successive filters, leaving only the cleansed residue fit to print”. But these filters aren’t in place by intent. They come about incidentally as features of certain ways the world works.

The authors make it clearest when they say, the “media do not function in the manner of the propaganda system of a totalitarian state. Rather, they permit—indeed even encourage—spirited debate, criticism, and dissent”. What the filters do, however, is limit that debate quite drastically to line up with the interests of powerful groups. The filters, arising not from people, but from circumstance, do all this subtly and “largely without awareness.”

This means that many of criticisms of the model for being anti-the journalistic spirit don’t particularly hold water. It’s not so much that journalists are shady, but rather that the system encourages a certain view of the world. More on that later, but it does explain why the bias we see is not so much a bunch of bald-faced lies. Instead it’s an earnest conveying and debating of news that makes it through the filters, and the de-emphasis or ignoring of news that doesn’t.

The second fact that becomes apparent is that the model has only increased in relevance given the advent of internet. While the internet allows us to diversify from what the book calls the “mainstream media” to small organisations and individuals, this increased freedom actually increases the influence of some of the filters the authors describe.

So, in the order they’re presented in the book, the five filters of the ‘Propaganda Model’.

The cost of doing business

Media outlets are usually private enterprises. This is, in fact, the primary virtue of media in democratically-oriented states. In private hands, the media can, in theory, do its job: provide a legitimate flow of information to the public, and in doing so, check the power of the government and hold decision-makers accountable. These things are not so easy to do when the media is in the hands of that very same government.

Unfortunately, this same feature introduces the pressures that accompany a market-orientation. For the media, like Method Man, “cash rules everything around me, C.R.E.A.M”.

The capital challenge

The first pressure is the need for capital. To get your news into the hands of the people, you need the cash to put it there. The more people, the more money. This obstacle is the primary reason for the “concentration of media ownership”. Larger companies have more money, which means they can reach larger audiences. Larger companies also have a tendency to buy up the smaller ones, and if not, prevent them competing simply because they can reach more people. While the exact nature of the breakdown may change from time to time, it has been true to say for a long time that a very small number of large companies produce the vast majority of the media we consume. This particular feature becomes extremely important to keep in mind later: the internet might have created more media options, but when we consider the third filter, “sources of information”, it doesn’t mean more information.

Market integration

The second is the influence of the market. Large companies are typically profit-seeking organisations. This isn’t a must, but profit-seeking and growth are pretty comfortable bed-fellows. These concentrated media entities are typically no different, especially since they already face the capital challenge. If the organisations aren’t owned by wealthy individuals or groups with a historical interest in increasing their wealth, they are generally integrated into the market. This leads to the pressure of “stockholders, directors, and bankers to focus on the bottom line”.

In short, media corporations are businesses, and like businesses they’re interested in making money.

The cost of doing business

These two influences, outlined by Herman and Chompsky, aren’t particularly ground-breaking. But it is certainly worth exploring how these pressures filter information. The most obvious is the need to conform to lines of information production that don’t overly threaten their business interests. Media companies can’t be overly risky with the content they produce without facing their own collapse. This is particularly true of media outlets owned by companies who also sell products, which are as or more common than media organisations that focus solely on content creation.

As such, there’s a pressure to filter out information that is threatening to the market, and filter in information that is beneficial. This particular point, I think, is more aptly addressed in the next filter “sources of income”, but it’s important to note that the pressure to conform to market-based messages doesn’t simply come from the desires of the consumers, but also the owners themselves.

Another, less obvious influence comes as a side effect of their contact with the market. The more distributed a media organisation is among shareholders, the more contact the organisation has with external influences. Larger companies require larger governence. The directors that provide this governence are frequently selected from the corporate sphere—who better to assist the organisation in a profit-orientation than business executives? This introduces even more top-down pressure to conform to market-friendly messaging.

Finally, profit-oriented companies are capacity-limited by government regulations. You can only sell what you’re allowed to sell, and media companies are governed by a system of licences. This legal dependency obviously opens the gate for influence directly from the government, and media organisations are routinely pressured via regulation to conform to certain narratives. This encourages media companies to dabble in political dealings themselves. By building ties with the government, media companies can favourably influence policy. But by building ties with the government, the media opens itself to indirect influence. This is most evident when the organisation selects government or ex-government officials for positions on its board of directors.

In sum, to own and run a media organisation requires an intimate relationship with the market. This has led to an extreme narrowing of the sources of information to mostly large companies with the most capital, as well as a market-friendly bias on the content of that information.

Sources of income


Herman and Chompsky go on to comment about the nature of advertising in media. Advertising, particularly with the advent of the internet, has become a dominant feature of our media streams. The introduction of advertising creates a cycle of influence. Advertising revenue lowers the cost of the media for viewers. This improves outreach, since more people will be interested in consuming cheaper media. More consumption excites more advertisers which kicks off the cycle again. The media organisations most successful at the cycle generally out-compete the ones who aren’t successful or don’t engage at all.

The obvious outcome of this is the least interesting, although powerful: content can’t be overly counter to the advertiser’s interests, or the advertisers will put their money elsewhere. The more interesting dynamic, though, is subtler.

In this model, the viewers of the media are not the consumers, the advertisers are. The advertisers are the ones paying for the service of serving ads to the viewers. The viewers, then, become the product. The media organisations, in order to remain competitive in attracting advertisers, must also attract an audience which buys products. It’s no good spending money advertising to people who ignore the advertising. As the authors put it, “the mass media are interested in attracting audiences with buying power, not audiences per se”.

This has important implications for the content of the media. The content becomes a kind of “filler”. It has to be attractive to an audience with money, but it can’t interrupt too much the “buying mood” of the people. It can’t, therefore, spend much time conveying messages that are too disruptive, powerful, upsetting, or distracting.

This is in fairly stark contrast to a common idea that the pressure to reach more people means that media organisations have to be more democratic and represent ideas that have a broad appeal. Rather, as the authors put it, it’s like “a voting system weighted by income”. The implications are actually similarly concerning. Because the media are targeting the wealthy, they are targeting the kind of elite groups that generate many of the pressures that comprise the filters. Wealthy audience members are more likely to be decision-makers themselves. As such, the model reinforces itself. The wealthy advertise the interests of the wealthy to the wealthy. Democratic, it is not.


A common alternative to advertisers these days are subscribers. This does indeed represent a shift away from the influence of advertisers, but it doesn’t quite get rid of the filter. The main issue with the advertising model is that the content has to draw in a certain audience. It does that by excluding the material the audience doesn’t like. Subscribers produce a similar pressure. Content creators who are reliant on subscribers are encouraged to lean into the topics their audience prefers. This content has to be what the authors call “a tolerable portrayal of reality”, but what people find tolerable isn’t always true. We are left, then, with a similar pressure to conform.

In sum, the sources of income determine the message for media outlets big and small. Audiences need to be attracted, and the best audiences have money. This encourages some news to be emphasised and some news to be ignored.

Sources of information

The third filter relates to where the information comes from in the first place. There’s a limit to how many journalists can be employed, and how many places they can be deployed to. This limit is determined by the cash the organisation has to fund the journalists. Larger organisations, therefore, are the sources of the most information. This particular point is ever more important since the advent of the internet. Since smaller organisations and individual content creators are simply unable to source information directly, they rely on the information produced by the larger companies, a process made effortless with the internet. This information has already been filtered, and it doesn’t therefore really represent the interests of the smaller organisations, but a subset of the interests of the larger companies.

There is another source of information, unrelated to the large media companies: ‘official sources’. The pressure to produce content, and the expense of journalism, places pressure on media producers to use the material and expertise distributed by governments, as well as corporate and other powerful interest groups. As Herman and Chompsky put it, “the large bureaucracies of the powerful subsidize the mass media, and gain special access by their contribution to reducing… costs”. This essentially means that a large proportion of the information that makes its way into the media stream is produced by the powerful groups the media is supposed to be holding accountable.

These two features lead also to some interesting dynamics. The first is the consistent theme. By virtue of the fact that the large companies are the source of much of the information, the information is already filtered to reflect market-friendly messages. However, a large proportion of that information has already been pre-filtered to contain messaging that’s friendly to various governmental and powerful group interests by the governments and the interest groups themselves. Perhaps more importantly, this opens the way for government and various interest groups to “manage” the media: emphasising certain flows of information, and minimising others in order to create a particular framework of messaging that can lessen or eliminate negative press.

In sum, information is costly to obtain. Smaller organisations rely on larger ones to produce content, and all rely on government sources and sources backed by powerful interests. This means that not only is information filtered according to the interests of large companies, it’s also pre-filtered by governments and powerful interest groups.

“Flak” as discipline

Herman and Chompsky describe ‘flak’ as negative responses to media content. This comprises all ‘flak’, from the letter of a reader to the legal action of a large firm. The important relationship here is between the ‘flak’ and the power of the entity producing the flak. A single letter from a single reader has little power, but the legal action of a large firm is more difficult to ignore.

Once again, the concept of flak is not a novel one. The authors take care, however, to note the presence of “flak machines”—well-resourced groups, backed by powerful interests, which coordinate campaigns of flak to curtail media narratives. The criticisms of these machines are often given space in the media narrative with little or no acknowledgement of their backers and sponsors.

In short, ‘flak’ is any negative response to media, from letters to legal action. Flak influences the media narrative, particularly when from well-resourced ‘flak machines’ backed by powerful interests whose motives are rarely questioned.

Anti-ideologies, or fear of the “enemy”

The last filter needs a bit of modification to stay current, as the authors acknowledged in later editions and comments. It’s most easily understood as a focus on the fear of an “enemy”. This enemy takes the form of some potential threat, against which the media focuses its attention. Another way to think about this might be the idea of anti-ideologies: an ideology that is strictly against a particular thing. At the time of the book this was anti-communism. Later it became anti-terrorism, alongside the “War on Terror”. Today, some have suggested that Islam is the “enemy”, and the anti-ideology is islamophobia.

The origin of the anti-ideology is not the media organisations themselves, but rather “vested interests” who reach “through the system to exercise a profound influence on the mass media”. The link between the other filters and this is not explicitly made, but by now it should be fairly clear. The information that shakes its way though the filtering process is skewed in such a way as to concentrate and exaggerate a broad threat to powerful interest groups, real or imagined. The broadness of the threat allows it to be “used against anybody advocating policies that threaten… interests”, and it also narrows the platform for narratives that might be different.

A final comment on these anti-ideologies is also a modification, but seems increasingly important. A key theme of two of the filters, “sources of income” and “the cost of doing business”, concerns the orientation of media outlets toward cashflow. There is no doubt that these anti-ideologies, or fear campaigns, produce a great deal of attention from viewers. As such, the media is encouraged to lean into these anti-ideologies for its own financial benefit.

In sum, the media focuses on ‘fear the enemy’ narratives at the expense of less biased coverage. In part this is because the other filters permit this manipulation, but also because it generates income.

To conclude

There’s a great deal of nuance there, but the main points are pretty accessible. Five filters, each almost accidental in nature. The cost of distributing content is high which supports profit-seeking organisations to succeed, with profit-seeking motives. The primary source of revenue maintains focus on a narrow market, and thus narrow topics of information. Information is costly to source, so the media rely on powerful groups to provide it. Powerful groups also resource entities to fight certain media narratives on their behalf. And finally, the media narratives are often dominated by broad ‘fear the enemy’ threads that generate income, support the interests of powerful groups, and take the oxygen out of other stories.

In sum, we have a media that, by its very ‘free-marketeer’ nature is heavily biased towards some news and away from others.

This last point is perhaps the most important. The model does not tell us that the media is lying to us. It doesn’t tell us that journalists are corrupt people. It tells us that the system is fundamentally skewed. It will emphasise only the news that makes it through the filters, and de-emphasise or ignore the news that doesn’t.

In fact, Herman and Chompsky explicitly comment on these journalists. They say that while there “may be some who are simply corrupt”, the system selects for those who reflect the interests of the system. Those that remain are “free to express themselves with little managerial control, and they will be able to assert, accurately, that they perceive no pressure to conform.” The structure of these filters means that it’s extremely labour intensive to find news outside the filtered information. It must then be supported, against the overwhelming weight of the material that does make it through the filters. To quote, “[t]he technical structure of the media” means that this simply can’t “be expressed between two commercials, or in seven hundred words, without the appearance of absurdity that is difficult to avoid when one is challenging familiar doctrine with no opportunity to develop facts or argument.” This is in addition to the typical dynamics that exist in any group: any given group of people will be attractive to similarly-minded people, will select similarly-minded people, and will repel people who are dissimilar. It’s why humans are so cliquey..

Coming up next, we go into how the model explains some of the stranger and more annoying trends in the media currently. In the meantime, keep these five filters in mind. You might start to find out that the world isn’t quite so terrifying or mundane as the news has you believing.

Ideologies you choose at btrmt.

Join over 2000 of us. Get the newsletter.