Journalism Studies, Volume 1, Number 1, 2000, pp.

101–112

The Propaganda Model: a retrospective

EDWARD S. HERMAN
University of Pennsylvania, USA

ABSTRACT This article describes the “propaganda model” of media behavior and performance,
initially set down in the book Manufacturing Consent, and addresses some of the scholarly
criticisms leveled against the model since its inception a decade ago. Drawing on contemporary
examples of media reporting, the article argues for the continuing, if not increasing, relevance
of the propaganda model in explaining current media behavior.

KEY WORDS Conspiracy Theory, Mass Media, Political Economy, Propaganda

In Manufacturing Consent: the political and contend that the media behave
economy of the mass media, Noam fairly and responsibly.1 The exclusion
Chomsky and I put forward a of the propaganda model perspective is
“propaganda model” as a framework noteworthy for one reason, because
for analysing and understanding how that perspective is consistent with long-
the mainstream US media work and standing and widely held elite views
why they perform as they do (Herman that “the masses are notoriously short-
and Chomsky, 1988). We had long sighted” (Bailey, 1948, p. 13) and are
been impressed by the regularity with “often poor judges of their own inter-
which the media operate on the basis ests” (Lasswell, 1933, p. 527), so that
of a set of ideological premises, de- “our statesmen must deceive them”
pend heavily and uncritically on elite (Bailey, 1948, p. 13); and they “can be
information sources and participate in managed only by a specialized class
propaganda campaigns helpful to elite whose personal interests reach beyond
interests. In trying to explain why media the locality” (Walter Lippmann, 1921, p.
perform in this way we looked to struc- 310). In Lippmann’s view, the “manu-
tural factors as the only possible root of facture of consent” by an elite class
the systematic patterns of media be- had already become “a self-conscious
havior and performance. art and a regular organ of popular
Because the propaganda model government” by the 1920s (Lippman,
challenges basic premises and sug- 1921, p. 248).
gests that the media serve antidemo- Clearly, the manufacture of consent
cratic ends, it is commonly excluded by a “specialized class” that can over-
from mainstream debates on media ride the short-sighted perspectives of
bias. Such debates typically include the masses must entail media control
conservatives, who criticize the media by that class. Political scientist Thomas
for excessive liberalism and an adver- Ferguson contends that the major me-
sarial stance toward government and dia, “controlled by large proŽ t-maximiz-
business, and centrists and liberals, ing investors do not encourage the
who deny the charge of adversarialism dissemination of news and analyses
ISSN 1461-670X Print; 1469-9699 Online/00/010101-12 Ó 2000 Taylor & Francis Ltd
102 EDWARD S. HERMAN

that are likely to lead to popular indig- and they are funded largely by adver-
nation and, perhaps, government ac- tisers who are also proŽ t-seeking enti-
tion hostile to the interests of all large ties, and who want their
investors, themselves included” (Fer- advertisements to appear in a support-
guson, 1995, p. 400). Political scientist ive selling environment. The media
Ben Page provides evidence that there also lean heavily on government and
are common “elite-mass gaps”, with major business Ž rms as information
“ordinary citizens … considerably less sources and both efŽ ciency and politi-
enthusiastic than foreign policy elites cal considerations and, frequently,
about the use of force abroad, about overlapping interests, cause a certain
economic or (especially) military aid or degree of solidarity to prevail among
arms sales, and about free-trade the government, major media and other
agreements. The average American is corporate businesses. Government
much more concerned than foreign pol- and large non-media business Ž rms
icy elites about jobs and income at are also best positioned (and
home” (Page, 1996, p. 118). Page sufŽ ciently wealthy) to be able to press-
notes that “the problem for public delib- ure the media with threats of with-
eration is most severe when ofŽ cials of drawal of advertising or TV licenses,
both parties and most mainstream me- libel suits and other direct and indirect
dia take positions that are similar to modes of attack. The media are also
each other and opposed to the public” constrained by the dominant ideology,
(Page, 1996, p. 119). The propaganda which heavily featured anti-commu-
model explains the “elite-mass gaps”, nism before and during the Cold War
as well as elite and mainstream media era, and was often mobilized to induce
hostility to this mode of analysis: re- the media to support (or refrain from
fusal to allow it entry into the debate is criticizing) US attacks on small states
understandable given that the gaps are that were labeled communist.
embarrassing and suggest that the me- These factors are linked together,
dia do serve narrow elite interests. re ecting the multi-leveled capability of
This article brie y describes the government and powerful business
propaganda model, addresses some of entities and collectives (e.g. the Busi-
the criticisms that have been leveled ness Roundtable; the US Chamber of
against it and discusses how the model Commerce; the vast number of well-
holds up a decade or so after its publi- heeled industry lobbies and front
cation.2 Examples are provided to illus- groups) to exert power over the  ow of
trate the ways in which the propaganda information. We noted that the Ž ve fac-
model helps explain the nature of me- tors involved—ownership, advertising,
dia coverage of important political top- sourcing,  ak and anti-communist
ics at the turn of the century. ideology—work as “Ž lters” through
which information must pass, and that
individually and often in cumulative
The Propaganda Model fashion they greatly in uence media
choices. We stressed that the Ž lters
What is the propaganda model and work mainly by the independent action
how does it work? Its crucial structural of many individuals and organizations;
factors derive from the fact that the and these frequently, but not always,
dominant media are Ž rmly embedded have a common view of issues as well
in the market system. They are proŽ t- as similar interests. In short, the propa-
seeking businesses, owned by very ganda model describes a decentralized
wealthy people (or other companies); and non-conspiratorial market system
THE PROPAGANDA MODEL 103

of control and processing, although at ganda model explained everything or

times the government or one or more that it illustrated media omnipotence
private actors may take initiatives and and complete effectiveness in manu-
mobilize co-ordinated elite handling of facturing consent. It is a model of me-
an issue. dia behavior and performance, not of
Propaganda campaigns can occur media effects. We explicitly pointed to
only when they are consistent with the the existence of alternative media,
interests of those controlling and man- grassroots information sources and
aging the Ž lters. For example, these public scepticism about media truthful-
managers all accepted the view that ness as important limits on media ef-
the Polish government’s crackdown on fectiveness in propaganda service, and
the Solidarity Union in 1980 and 1981 we urged the support and more vigor-
was extremely newsworthy and de- ous use of the existing alternatives.
served severe condemnation; whereas Both Chomsky and I have often pointed
the same interests did not Ž nd the to the general public’s persistent re-
Turkish military government’s equally fusal to fall into line with the media and
brutal crackdown on trade unions in elite over the morality of the Vietnam
Turkey at about the same time to be War, the desirability of the assault on
newsworthy or reprehensible. In the lat- Nicaragua in the 1980s and the merits
ter case the US government and busi- of the North American Free Trade
ness community liked the military Agreement in the 1990s, among other
government’s anticommunist stance matters. The power of the US propa-
and open-door economic policy; the ganda system lies in its ability to mobi-
crackdown on Turkish unions had the lize an elite consensus, to give the
merit of weakening the left and keeping appearance of democratic consent,
wages down. In the Polish case, propa- and to create enough confusion, mis-
ganda points could be scored against a understanding and apathy in the gen-
Soviet-supported government, and eral population to allow elite programs
concern could be expressed for work- to go forward. We also emphasized the
ers whose wages were not paid by fact that there are often differences
Free World employers. The Ž t of this within the elite that open up space for
dichotomization to corporate interests some debate and even occasional (but
and anticommunist ideology is obvious. very rare) attacks on the intent as well
We used the concepts of “worthy” as the tactical means of achieving elite
and “unworthy” victims to describe this ends.
dichotomization, with a trace of irony as Although the propaganda model was
the varying treatment was clearly re- generally well received on the left,
lated to political and economic advan- some complained of an allegedly pessi-
tage rather than anything like actual mistic thrust and implication of hope-
worth. In fact, the Polish trade unionists less odds to be overcome. A closely
quickly ceased to be worthy when com- related objection concerned its appli-
munism was overthrown and the work- cability to local con icts where the pos-
ers were struggling against a sibility of effective resistance was often
western-oriented neoliberal regime. greater than in the case of national
The travails of today’s Polish workers, issues; but the propaganda model does
like those of Turkish workers, do not not suggest that local and even larger
pass through the propaganda model victories are impossible, especially
Ž lters. Both groups are unworthy vic- where the elites are divided or have
tims at this point. limited interest in an issue. For exam-
We never claimed that the propa- ple, coverage of issues such as gun
104 EDWARD S. HERMAN

control, school prayer and abortion Many of the critiques displayed a

rights may well receive more varied barely concealed anger, and in most of
treatment than, for instance, global them the propaganda model was dis-
trade, taxation and economic policy. missed with a few superŽ cial clichés
Moreover, well-organized campaigns (conspiratorial, simplistic, etc.), without
by labor, human rights or environmen- fair presentation or subjecting it to the
tal organizations that are Ž ghting test of evidence. Let me discuss brie y
against abusive local businesses can some of the main criticisms.
sometimes elicit positive media cover-
age. In fact, we would like to think that
the propaganda model can help ac- Conspiracy Theory
tivists understand where they might
best deploy their efforts to in uence We explained in Manufacturing Con-
mainstream media coverage of issues. sent that critical analyses such as ours
The model does suggest that the would inevitably elicit cries of con-
mainstream media, as elite institutions, spiracy theory, and in a futile effort to
commonly frame news and allow de- prevent this we devoted several pages
bate only within the parameters of elite of the preface to an explicit rejection of
perspectives; and that when the elite is conspiracy and an attempt to show that
really concerned and uniŽ ed and/or the propaganda model is best de-
when ordinary citizens are not aware of scribed as a “guided market system”.
their own stake in an issue or are im- Mainstream critics still made the
mobilized by effective propaganda, the charge, partly because they are too
media will serve elite interests uncom- lazy to read a complex work, partly
promisingly. because they know that falsely ac-
cusing a radical critique of conspiracy
theory will not cost them anything, and
Mainstream Liberal and partly because of their superŽ cial as-
Academic “Left” Critiques sumption that, since the media com-
prise thousands of “independent”
Many liberals and some academic me- journalists and companies, any Ž nding
dia analysts of the left did not like the that they follow a “party line” that
propaganda model. Some asked serves the state must rest on an as-
rhetorically where we got the infor- sumed conspiracy. (In fact, it can result
mation used to condemn the main- from a widespread gullible acceptance
stream media if not from the media of ofŽ cial handouts, common internal-
themselves (a tired apologetic point ized beliefs, common policies estab-
that we answered at length in our pref- lished from above within the
ace). Many of these critics found re- organizations based on ideology and/or
pugnant a wholesale condemnation of interests, and fear of reprisal for critical
a system they believed to be basically analyses from within the organization or
sound, its inequalities of access regret- from the outside.) The apologists can-
table but tolerable, its pluralism and not abide the notion that institutional
competition effectively responding to factors can cause a “free” media to act
consumer demands. In the postmod- like lemmings in jointly disseminating
ernist mode, global analyses and glo- false and even silly propaganda; such a
bal solutions are rejected and derided charge must assume a conspiracy.
and individual struggles and small vic- Sometimes the critics latched on to a
tories are stressed, even by nominally word or phrase that suggests a collec-
left thinkers. tive purpose or function, occasionally
THE PROPAGANDA MODEL 105

ironically, to make their case. Commu- ready used it in a 1981 book, Corpo-
nications professor Robert Entman, for rate Control, Corporate Power, and that
example, stated that we damaged our I was mainly responsible for the chap-
case by alleging that media coverage ter in Manufacturing Consent that pre-
of the 1973 Paris accord on Vietnam sented the propaganda model. Of
“was consciously ‘designed by the loyal course, whether traceable to Chom-
media to serve the needs of state skian linguistics or industrial organiza-
power’ … which comes close to en- tion theory, the substantive issues are:
dorsing a conspiracy theory, which the are the assumptions plausible? does
authors explicitly disavow early on” the model work? But showing a poss-
(Entman, 1990, p. 126). The word ible esoteric origin is a form of putdown
“consciously” was Entman’s, and he that suggests remoteness from and
neglected numerous statements about lack of touch with real media people.
the media’s treatment of the Paris ac-
cord that did not Ž t his effort to bring us Failure to Touch Base with
“close to” a conspiracy theory. To say
that we “disavow” a conspiracy theory
Reporters
is also misleading: we went to great
Romano did in fact follow up with the
pains to show that our view is closer to
admonition that we had failed to ask
a free market model; we argued that
reporters why they did what they did.
the media comprise numerous inde-
He implied, without offering any evi-
pendent entities that operate on the
dence, that the journalistic bias we crit-
basis of common outlooks, incentives
icized might have been revealed as for
and pressures from the market,
good cause, if we had only asked for
government and internal organizational
an explanation. However, apart from
forces.
the fact that we did speak with quite a
The propaganda model explains me-
few reporters, the criticism is inane. Are
dia behavior and performance in struc-
reporters even aware of the deeper
tural terms, and intent is an
sources of bias they may internalize?
unmeasurable red herring. All we know
will they not tend to rationalize their
is that the media and journalists often
behavior? More important, if we Ž nd,
mislead in tandem—some no doubt in-
for example, that in reporting on the
ternalize a propaganda line as true,
Nicaraguan and Salvadoran elections
some may know it is false, but the point
of 1984, they asked different questions
is unknowable and unimportant.
in the two elections in exact accord
with the propaganda line of the US
government, would asking journalists
Chomskian Linguistics what went on in their minds serve any
useful purpose? This line of criticism,
Some of the criticisms of the propa-
like the insistence on inquiry into re-
ganda model have been laughable.
porter–proprietor intentions, is a diver-
Carlin Romano, in his review in Tikkun,
sionary ploy that essentially denies the
located the weakness of the model in
legitimacy of a quantitative (or sci-
Chomskian linguistic theories that al-
entiŽ c) analysis of media performance.
legedly view everything as rooted in
deep structures (Romano, 1989). He
was unaware that the rooting of corpo- Failure to Take Account of Media
rate behavior and performance in struc- Professionalism and Objectivity
ture is the core of modern industrial
organization analysis, that I had al- A more sophisticated version of the last
106 EDWARD S. HERMAN

argument, put forward by communica- whereas Hallin’s focus on “profession-

tions professor Daniel Hallin, is that we alism” fares badly. Hallin acknow-
failed to take account of the maturing of ledges that “the administration was
journalist professionalism, which he able more often than not to prevail in
claims to be “central to understanding the battle to determine the dominant
how the media operate” (Hallin, 1994, frame of television coverage”, “the
p. 13). Hallin also states that in protect- broad patterns in framing the story can
ing and rehabilitating the public sphere be accounted for almost entirely by the
“professionalism is surely part of the evolution of policy and elite debate in
answer” (Hallin, 1994, p. 4). Washington”, and “coherent state-
However, professionalism and objec- ments of alternative visions of the world
tivity rules are fuzzy and  exible con- order and US policy rarely appeared in
cepts and are not likely to override the the news” (Hallin, 1994, pp. 64, 74, 77).
claims and demands of deeper power This is exactly what the propaganda
and control relationships. Professional- model would forecast. If, as Hallin con-
ism arose in journalism in the years tends, a majority of the public opposed
when the newspaper business was the elite view, what kind of “profes-
becoming less competitive and more sionalism” allows a virtually complete
dependent on advertising. Profession- suppression of the issues as the
alism was not an antagonistic move- majority perceives them?
ment by the workers against the press Hallin mentions a “nascent alterna-
owners, but was actively encouraged
tive perspective” in reporting on El
by many of the latter. It gave a badge
Salvador—a “human rights” frame-
of legitimacy to journalism, ostensibly
work—that “never caught hold”. The
assuring readers that the news would
propaganda model can explain why it
not be in uenced by the biases of own-
never took hold;3 Hallin does not. Even
ers, advertisers or the journalists them-
selves. In certain circumstances it has though 700 journalists were present at
provided a degree of autonomy, but the Salvadoran election of 1982, alleg-
professionalism has also internalized edly “often skeptical” of election in-
some of the commercial values that tegrity (Hallin, 1994, p. 72), why did it
media owners hold most dear, like rely- yield a “public relations victory” for the
ing on inexpensive ofŽ cial sources as administration and a major falsiŽ cation
the credible news source. As Ben of reality (as described in Manufactur-
Bagdikian has noted, professionalism ing Consent)? Hallin does not explain
has made journalists oblivious to the this. He never mentions the OfŽ ce of
compromises with authority they are Public Diplomacy, the Ž ring of New
constantly making (Bagdikian, 1987, York Times reporter Raymond Bonner,
p. 180). Hallin himself acknowledges, or the work of the  ak machines.4 He
moreover, that professional journalism does not explain the failure of the
can allow something close to complete media to report even a tiny fraction of
government control through domination the crimes of the contras in Nicaragua
of sources (Hallin, 1994, pp. 64, 70). and the death machines in El Salvador
Although Hallin claims that the and Guatemala, in contrast with media
propaganda model cannot explain the in ation of Sandinista misdeeds and
case of media coverage of the Central the double standard in reporting on
American wars of the 1980s, when the Nicaraguan election of 1984. Given
there was considerable domestic hos- the elite divisions and public hostility
tility to the Reagan policies, in fact the to the Reagan policy, media subser-
model works extremely well there, vience was phenomenal and arguably
THE PROPAGANDA MODEL 107

exceeded that which the propaganda the negative reactions of the serious
model might have anticipated.5 left-of-center media analysts such as
Philip Schlesinger, James Curran, Pe-
ter Golding, Graham Murdock and
Failure to Explain Continued John Eldridge, as well as that of Daniel
Opposition and Resistance Hallin. Of these critics, only
Both Hallin and historian Walter LaFe- Schlesinger both summarizes the ele-
ber in a review in the New York Times ments of our model and discusses our
(LaFeber, 1988) pointed to the contin- evidence. He acknowledges that the
ued opposition to Reagan’s Central case studies make telling points, but in
America policy as somehow incompat- the end he Ž nds ours “a highly deter-
ible with the model. These critics failed ministic vision of how the media oper-
to comprehend that the propaganda ate coupled with a straightforward
model is about how the media work, functionalist conception of ideology”
not how effective they are. Even the (Schlesinger, 1989, p. 297). SpeciŽ -
sophisticated and sympathetic Philip cally, he claims that we failed to explain
Schlesinger calls ours an “effects” the weights to be given our Ž ve Ž lters;
model, that “assumes that dominant we did not allow for external in uences,
agendas are reproduced in public opin- nor did we offer a “thoroughgoing
ion”, but he immediately quotes our analysis of the ways in which economic
statement that the “system is not all dynamics operate to structure both the
powerful… Government and the elite range and form of press presentations”
domination of the media have not suc- (quoting Graham Murdock); and al-
ceeded in overcoming the Vietnam though we put forward “a powerful ef-
syndrome” (Schlesinger, 1989, p. 301). fects model” we admit that the system
Nowhere does he cite us saying any- is not all powerful, which calls into
thing like his summary of our alleged question our determinism.
views on effects. We also stated ex- The criticism of the propaganda
plicitly with regard to Central America model for being deterministic ignores
that the elite was sufŽ ciently divided several important considerations. Any
over tactics to allow space and con- model involves deterministic elements,
siderable debate. We did stress, how- so that this criticism is a straw person
ever, that the parameters of debate did unless the critics also show that the
not extend to fundamental challenges system is not logically consistent, oper-
to the US intervention (Herman and ates on false premises, or that the pre-
Chomsky, 1988, pp. xii–xiii). dictive power of the determining
By the logic of this form of criticism of variables is poor. The critics often ac-
the propaganda model, the fact that knowledge that the case studies we
many Soviet citizens did not swallow present are powerful, but they do not
the lines put forward by Pravda demon- show where the alleged determinism
strates that Pravda was not serving a leads to error nor do they offer or point
state propaganda function. to alternative models that would do a
better job.6
The propaganda model deals with
The Propaganda Model is too extraordinarily complex sets of events,
Mechanical and Functionalist and and only claims to offer a broad frame-
Ignores the Existence of Space, work of analysis, a Ž rst approximation,
Contestation and Interaction that requires modiŽ cation depending
on local and special factors, and that
This set of criticisms is at the heart of may be entirely inapplicable in some
108 EDWARD S. HERMAN

cases; but if it offers insight in numer- and Murdock, 1991, p. 19). The Gold-
ous important cases that have large ing–Murdock claim that the propa-
effects and cumulative ideological ganda model focuses on “strategic
force, it is arguably serviceable unless interventions” is a surprising misread-
a better model is provided. Usually the ing, as the model’s Ž lters are built-in
critics stick to generalities and offer no and operate mainly through the inter-
critical detail or alternative model; when nalized recognition and enforcement of
they do provide alternatives, the results constraints and choices based on the
are not impressive.7 structure of power. Strategic interven-
The criticism of the propaganda tions certainly occur, but are of dis-
model for functionalism is also dubious, tinctly secondary importance.
and the critics sometimes seem to call It is also untrue that the propaganda
for more functionalism. The model model implies no constraints on media
does describe a system in which the owners and managers; we recognized
media serve the elite, but by complex and spelled out the circumstances un-
processes incorporated into the model der which the media will be relatively
that involve mechanisms and policies open—mainly, when there are elite dis-
whereby the powerful protect their in- agreements and when other groups in
terests naturally and without overt con- society are interested in, informed
spiracy. This would seem one of the about and organized to Ž ght about is-
model’s merits; it shows a dynamic and sues. The propaganda model does
self-protecting system in operation. start from the premise that a critical
The same corporate community that political economy will put front and cen-
in uences the media through its power ter the analysis of the locus of media
as owner, dominant funder (advertis- control and the mechanisms by which
ing) and major news source also un- the powerful are able to dominate
derwrites the efforts of Accuracy in the  ow of messages and limit the
Media and the American Enterprise In- space of contesting parties. The limits
stitute to in uence the media through on their power are certainly important,
harassment and the provision of right- but why should these get Ž rst place,
thinking experts. Critics of propaganda except as a means of minimizing the
model functionalism such as Eldridge power of the dominant interests,
and Schlesinger contradictorily point to in ating the elements of contestation,
the merit of analyses that focus on and pretending that the marginalized
“how sources organize media strate- have more strength than they really
gies” to achieve their ends.8 Apparently possess?
it is admirable to analyze corporate mi-
cro strategies to in uence the media,
but to focus on global corporate efforts Enhanced Relevance of the
to in uence the media is illegitimate Propaganda Model
functionalism.
Golding and Murdock criticize the The dramatic changes in the economy,
model for its focus on “strategic inter- the communications industries and
ventions”, that allegedly cause us to politics over the past dozen years have
“overlook the contradictions in the sys- tended on balance to enhance the ap-
tem. Owners, advertisers and key pol- plicability of the propaganda model.
itical personnel cannot always do as The Ž rst two Ž lters—ownership and ad-
they wish.” Analyzing “the nature and vertising—have become ever more im-
sources of these limits” is a “key task” portant. The decline of public
of critical political economy (Golding broadcasting, the increase in corporate
THE PROPAGANDA MODEL 109

power and global reach and the merg- on behalf of corporate America (Carey,
ers and centralization of the media 1995; Stauber and Rampton, 1995).
have made bottom-line considerations This industry understands how to utilize
more in uential in the US, in Europe journalistic conventions to serve its
and many other countries. The compe- own ends. Studies of news sources
tition for advertisers has become more reveal that a signiŽ cant proportion of
intense and the boundaries between news originates in public relations re-
editorial and advertising departments leases. There are, by one count,
have weakened further. Newsrooms 20,000 more public relations agents
have been more thoroughly incorpor- working to doctor the news today than
ated into transnational corporate em- there are journalists writing it (Dowie,
pires, with budget cuts and even less 1995, pp. 3–4).
management enthusiasm for investi- The Ž fth Ž lter—anti-communist ideol-
gative journalism that would challenge ogy—is possibly weakened by the col-
the structure of power (Herman and lapse of the Soviet Union and global
McChesney, 1997). In short, the pro- socialism, but this is easily offset by the
fessional autonomy of journalists has greater ideological force of the belief in
been reduced. the “miracle of the market” (Reagan).
Some argue that the Internet and the There is now an almost religious faith
new communication technologies are in the market, at least among the elite,
breaking the corporate stranglehold on so that regardless of evidence markets
journalism and opening an unpre- are assumed to be benevolent and
cedented era of interactive democratic non-market mechanisms are suspect.
media. There is no evidence to support When the Soviet economy stagnated in
this view with regard to journalism and the 1980s, it was attributed to the ab-
mass communication. In fact, one sence of markets; the disintegration of
could argue that the new technologies capitalist Russia in the 1990s is blamed
are exacerbating the problem. They on politicians and workers failing to let
permit media Ž rms to shrink staff even markets work their magic. Journalism
as they achieve greater outputs, and has internalized this ideology. Adding it
they make possible global distribution to the Ž fth Ž lter in a world where the
systems that reduce the number of me- global power of market institutions
dia entities. Although the new technolo- makes non-market options seem
gies have great potential for democratic utopian gives us an ideological pack-
communication, there is little reason to age of immense strength.
expect the Internet to serve democratic
ends if it is left to the market (Herman
and McChesney, 1997, pp. 117–35). Further Applications
The third and fourth Ž lters—sourcing
and  ak—have also strengthened as The propaganda model Ž ts exceedingly
mechanisms of elite in uence. A re- well the media’s treatment of the pass-
duction in the resources devoted to age of the North American Free Trade
journalism means that those who sub- Agreement (NAFTA) and the sub-
sidize the media by providing sources sequent Mexican crisis and meltdown
for copy gain greater leverage. More- of 1994–5. Once again there was a
over, work by people such as Alex sharp split between the preferences of
Carey, John Stauber and Sheldon ordinary citizens and the elite and busi-
Rampton has helped us see how the ness community; polls consistently
public relations industry has been able showed substantial majorities opposed
to manipulate press coverage of issues to NAFTA—and to the bailout of in-
110 EDWARD S. HERMAN

vestors in Mexican securities—but the bilitating strikes such as that involving

elite in favor. Media news coverage, Caterpillar were treated in a very low
selection of “experts”, and opinion key, and in a notable illustration of the
columns were skewed accordingly; applicability of the propaganda model
their judgement was that the beneŽ ts of the long Pittston miners’ strike was ac-
NAFTA were obvious, were agreed to corded much less attention than the
by all qualiŽ ed authorities, and that only strike of miners in the Soviet Union
demagogues and “special interests” (Tasini, 1994, pp. 7–9). For years the
were opposed.9 The effort of labor to media found the evidence that the ma-
in uence the outcome of the NAFTA jority of ordinary citizens were doing
debates was harshly criticized in both badly in the New World Economic Or-
the New York Times and the Washing- der of marginal interest; they discov-
ton Post, with no comparable criticism ered this issue only under the impetus
of corporate or governmental (US and of Pat Buchanan’s right-wing populist
Mexican) lobbying and propaganda. outcries during the 1996 presidential
With the Mexican meltdown beginning election campaign.
in December 1994, the media were Another striking application of the
clear that NAFTA was not to blame, propaganda model can be seen in the
and in virtual lock-step they supported media’s treatment of the chemical in-
the Mexican (investor) bailout, despite dustry and its regulation (Herman,
poll reports of massive general public 1999, pp. 231–52). Because of the in-
opposition. Experts and media repeat- dustry’s power, as well as the media’s
edly explained that the merit of NAFTA receptivity to the demands of the
was that it had “locked Mexico in” so business community, the media have
that it could not resort to controls to normalized a system described by
protect itself from severe de ation and Rachel Carson in Silent Spring as
unemployment. They were oblivious to “deliberately poisoning us, then polic-
the profoundly undemocratic nature of ing the results”. Industry is permitted to
this lock-in (Herman, 1999, pp. 183–6). produce and sell chemicals (and now
As is suggested by the treatment of also bioengineered foods) without inde-
NAFTA and of labor’s right to partici- pendent and prior proof of safety, and
pate in its debates, the propaganda the “policing” by the Environmental
model applies to domestic as well as Protection Agency has been badly
foreign policy issues. Labor has been compromised by underfunding and
under siege in the US for the past political limits on enforcement as well
several decades, but you would hardly as testing. Although industry denials
know this from the mainstream media. of harm from its products—from lead
A 1994 Business Week article noted in gasoline to asbestos and Agent
that “over the past dozen years … US Orange—and fraudulent testing have
industry has conducted one of the most been notorious for many years, the me-
successful union wars ever”, helped by dia still use the phrase “junk science” to
“illegally Ž ring thousands of workers for refer to the science employed by envi-
exercising their right to organize”, with ronmentalists and lawyers suing the in-
unlawful Ž rings occurring in “one-third dustry on behalf of its victims, not that
of all representation elections in the sponsored by industry. They have in-
late ’80s” (Bernstein, 1994, pp. 70). But ternalized industry usage, just as they
this successful war was carried out have normalized a status quo of caveat
quietly, with media co-operation. The emptor (buyer beware) rather than of
decertiŽ cation of unions, use of safety Ž rst.
replacement workers and long and de- In the health insurance controversy
THE PROPAGANDA MODEL 111

of 1992–3, the media’s refusal to take spelled out in more detail the contest-
the single-payer option seriously, de- ing forces both within and outside the
spite apparent widespread public sup- media and the conditions under which
port and the effectiveness of the these are likely to be in uential. How-
system in Canada, served well the in- ever, we made these points, and it is
terests of the insurance and medical quite possible that nothing we could
service complex (Canham-Clyne, have done would have prevented our
1994). The uncritical media reporting being labeled conspiracy theorists,
and commentary on the alleged ur- rigid determinists and deniers of the
gency of Ž scal restraint and a balanced possibility that people can resist (even
budget in the years 1992–6 Ž t well the as we called for resistance).
business community’s desire to reduce The propaganda model remains a
the social budget and weaken regu- very workable framework for analyzing
lation. The applicability of the propa- and understanding the mainstream me-
ganda model in these and other cases, dia–perhaps even more so than in
including the “drug wars”, seems clear 1988. As noted earlier in reference to
(Chomsky, 1991, pp. 114–21). Central America, the media’s perform-
ance often surpassed expectations of
media subservience to government
Final Note propaganda. It did so, also, in their
reporting on the Persian Gulf and
In retrospect, perhaps we should have Yugoslav wars of 1990 and 1999, re-
made it clearer that the propaganda spectively (Kellner, 1992; Mowlana et
model was about media behavior and al., 1992; Herman, 1999, pp. 161–6;
performance, with uncertain and vari- Chomsky, 1999). We are still waiting
able effects. Perhaps we should have for our critics to provide a better model.

Notes
1
For a discussion and illustration of the narrow focus of mainstream debates on media bias see Chomsky,
1989, pp. 160–80.
2
Noam Chomsky analyses some of these criticisms in Chomsky, 1989: Appendix 1.
3
The government, a primary and powerful source, was pressing its case hard, and trying to downgrade human
rights issues: it mobilized its resources of “ ak” and anti-communist ideology to keep the media in line; and
the owners and quite a few journalists of the dominant media were not unsympathetic to the government
attempt to overthrow the Sandinista government.
4
On the Ž ring of Raymond Bonner from the New York Times, its signiŽ cance in curbing media criticism of US
policy and the role of the Wall Street Journal as a media  ak machine, see “The Wall Street Journal As
Propaganda Agency”, chapter 9 in Herman, 1999.
5
For compelling documentation on this extraordinary subservience, see Chomsky, 1989, pp. 197–261.
6
I should note that the case studies in Manufacturing Consent are only a small proportion of those that
Chomsky and I have done that support the analysis of the propaganda model. Special mention should be
made of those covering the Middle East, Central America and terrorism. See especially Chomsky, 1983, 1986,
1989; Herman, 1982; Herman and O’Sullivan, 1990.
7
In fact, the only attempt to offer an alternative model was in Nicholas Lemann’s review of Manufacturing
Consent (Lemann 1989). For an analysis of this effort see Chomsky 1989, pp. 145–8.
8
The quotation is from Schlesinger (1989: 284), cited approvingly by Eldridge (1993: 29).
9
For a fuller account, “NAFTA, Mexican Meltdown and the Propaganda System”, chapter 14 in Herman, 1999.
See also Lee, 1995.

References
Bagdikian, Ben (1987) The Media Monopoly, Boston: Beacon Press.
Bailey, Thomas (1948) The Man in the Street: the impact of American public opinion on foreign policy, New York:
Macmillan.
112 EDWARD S. HERMAN

Bernstein, Aaron (1994) “The Workplace”, Business Week 23 May.

Canham-Clyne, John (1994) “When ‘Both Sides’ Aren’t Enough: the restricted debate over health care reform”,
EXTRA!, January–February.
Carey, Alex (1995) Taking the Risk Out of Democracy, Sydney NSW, Australia: University of New South Wales
Press.
Chomsky, Noam (1983) Fateful Triangle, Boston: South End Press.
Chomsky, Noam (1986) Pirates & Emperors, New York: Claremont Research and Publications.
Chomsky, Noam (1989) Necessary Illusions, Boston: South End Press.
Chomsky, Noam (1991) Deterring Democracy, London: Verso.
Chomsky, Noam (1999) The New Military Humanism: lessons from Kosovo, Monroe, ME: Common Courage
Press.
Dowie, Mark (1995) “Torches of Liberty”, introduction to John Stauber and Sheldon Rampton, Toxic Sludge Is
Good For You, Monroe, ME: Common Courage Press, 1996.
Eldridge, John (Ed.) (1993) Getting The Message, New York: Routledge.
Entman, Robert (1990) “News as Propaganda”, Journal of Communication 40, pp. 124–7.
Ferguson, Thomas (1995) Golden Rule, Chicago: University of Chicago Press.
Golding, Peter and Murdock, Graham (1991) “Culture, Communications, and Political Economy”, in James Curran
and Michael Gurevitch (Eds) Mass Media and Society, London: Edward Arnold, pp. 15–32.
Hallin, Daniel (1994) We Keep America on Top of the World, New York: Routledge.
Herman, Edward (1981) Corporate Control, Corporate Power, Cambridge UK: Cambridge University Press.
Herman, Edward (1982) The Real Terror Network, Boston: South End Press.
Herman, Edward (1999) The Myth of the Liberal Media, New York: Peter Lang Publishing.
Herman, Edward and Chomsky, Noam (1988) Manufacturing Consent, New York: Pantheon Books.
Herman, Edward and O’Sullivan, Gerry (1990) The Terrorism Industry, New York: Pantheon Books.
Herman, Edward and McChesney, Robert (1997) The Global Media, London: Cassell.
Kellner, Douglas (1992) The Persian Gulf TV War, Boulder, CO: Westview Press.
LaFeber, Walter (1988) “Whose News?” New York Times, 6 November.
Lasswell, Harold (1933) “Propaganda”, Encyclopaedia of the Social Sciences, New York: Macmillan, pp. 521–8.
Lee, Thea (1995) “False Prophets: the selling of NAFTA”, Economic Policy Institute.
Lemann, Nicholas (1989) “White House Watch”, The New Republic 2–16 January.
Bernstein, Aaron (1994) “The Workplace”, Business Week 23 May.
Canham-Clyne, John (1994) “When ‘Both Sides’ Aren’t Enough: the restricted debate over health care reform”,
EXTRA!, January–February.
Carey, Alex (1995) Taking the Risk Out of Democracy, Sydney NSW, Australia: University of New South Wales
Press.
Chomsky, Noam (1983) Fateful Triangle, Boston: South End Press.
Chomsky, Noam (1986) Pirates & Emperors, New York: Claremont Research and Publications.
Chomsky, Noam (1989) Necessary Illusions, Boston: South End Press.
Chomsky, Noam (1991) Deterring Democracy, London: Verso.
Chomsky, Noam (1999) The New Military Humanism: lessons from Kosovo, Monroe, ME: Common Courage
Press.
Dowie, Mark (1995) “Torches of Liberty”, introduction to John Stauber and Sheldon Rampton, Toxic Sludge Is
Good For You, Monroe, ME: Common Courate Press, 1996.
Eldridge, John (Ed.) (1993) Getting The Message, New York: Routledge.
Entman, Robert (1990) “News as Propaganda”, Journal of Communication 40, pp. 124–7.
Ferguson, Thomas (1995) Golden Rule, Chicago: University of Chicago Press.
Golding, Peter and Murdock, Graham (1991) “Culture, Communications, and Political Economy”, in James Curran
and Michael Gurevitch (Eds) Mass Media and Society, London: Edward Arnold, pp. 15–32.
Hallin, Daniel (1994) We Keep America on Top of the World, New York: Routledge.
Herman, Edward (1981) Corporate Control, Corporate Power, Cambridge UK: Cambridge University Press.
Herman, Edward (1982) The Real Terror Network, Boston: South End Press.
Herman, Edward (1999) The Myth of the Liberal Media, New York: Peter Lang Publishing.
Herman, Edward and Chomsky, Noam (1988) Manufacturing Consent, New York: Pantheon Books.
Herman, Edward and O’Sullivan, Gerry (1990) The Terrorism Industry, New York: Pantheon Books.
Herman, Edward and McChesney, Robert (1997) The Global Media, London: Cassell.
Kellner, Douglas (1992) The Persian Gulf TV War, Boulder, CO: Westview Press.
LaFeber, Walter (1988) “Whose News?” New York Times, 6 November.
Lasswell, Harold (1933) “Propaganda”, Encyclopaedia of the Social Sciences, New York: Macmillan, pp. 521–8.
Lee, Thea (1995) “False Prophets: the selling of NAFTA”, Economic Policy Institute.
Lemann, Nicholas (1989) “White House Watch”, The New Republic 2–16 January.
Lippmann, Walter (1921) Public Opinion, London: Allen and Unwin.
Mowlana, Hamid, Gerbner, George and Schiller, Herbert (Eds) (1992) Triumph of the Image, Boulder, CO:
Westview Press.
Page, Benjamin (1996) Who Deliberates? Chicago: University of Chicago Press.
Romano, Carlin (1989) “Slouching Toward Pressology”, Tikkun, p. 4.
Schlesinger, Philip (1989) “From Production to Propaganda”, Media, Culture and Society, p. 11.
Stauber, John and Rampton, Sheldon (1995) Toxic Sludge Is Good For You, Monroe, ME: Common Courage
Press.
Tasini, Jonathan (1990) “Lost in the Margins: Labor and the Media”, A Special FAIR Report.