Report: Media
Economics of Media Bias
The Challenge to the Presumption of Bias
Sources of a Liberal News Cartel
Liberal Journalists and Media Bias
Journalists versus Owners in the Production
of News
Maintenance of Bias: The Problem of Entrants
==============================
Cato Institute
Can The Media Be
So Liberal? The Economics Of Media Bias
Daniel
Sutter
Cato
Journal, Vol. 20, No. 3 (Winter 2001). Copyright © Cato Institute. All rights
reserved.
Daniel Sutter is
Assistant Professor of Economics at the University of Oklahoma. He thanks Tyler
Cowen, Joe McGarrity, Marc Poitras, Bob Tollison, and a referee for helpful
comments on earlier drafts.
Many
conservatives charge that the national news media exhibit a liberal bias,
despite surface appearances of impartiality. Charges of a liberal bias
essentially require the existence of a news cartel. Is the structure of the
news industry capable of sustaining a cartel? A liberal news cartel requires
collusion among news organizations and constraint of maverick outlets. Cartels
are always vulnerable to defection.
Indeed, many
critics who accuse the media of a liberal bias are likely skeptical of the
ability of businesses to maintain stable cartels without government assistance.
Competition usually forces firms to cater to their customer’s preferences. Yet
critics allege that all major national news organizations present the same
biased coverage, which is more liberal than the median voter. A liberal media
represents a failure in the news market.1
The documentation
of media bias has become something of a cottage industry since Edith Efron’s
(1971) pioneering study. Critics accusing the media of either a liberal or
conservative bias make use of surveys of working journalists, content analysis
of stories covered, and anecdotes about stories killed or not pursued to make
their case.2 But a conclusive measure of political bias in the news has been elusive.
That the Media Research Center and Fairness and Accuracy in Reporting
respectively point to the same news as demonstrating liberal and conservative
biases indicates that we lack such a measure. Unfortunately, we cannot simply
“test” the news and determine once and for all if a liberal bias exists.3 I adopt a
different approach here. I do not attempt to document the existence of news
bias. Rather, I ask what might generate and sustain a liberal news media. If we
cannot measure bias directly, we should consider the conditions for its
survival and whether the news industry meets these conditions. The parallel is
the economist’s approach to market power. Monopoly power can be difficult to
measure directly. A firm with market power must face little current or
potential competition. A successful cartel requires entry barriers and a
mechanism to prevent defection by member firms. Economists will likely reject
charges of monopoly power in a contestable market, one with small sunk costs.4 Scholars attempting
to document liberal bias have not asked questions of this type.
Specifically I
consider three questions concerning the liberal media charge. First, is the
source of bias on the demand (news consumer) or supply side of the news market?
Two potential supply-side sources can be distinguished, owners and journalists.
Although owners have used their news organizations to further their favored
political causes, corporate ownership of the media makes this less likely.
Proponents of the liberal bias charge place great weight on the numerous
surveys concerning journalists’ personal political views and voting patterns, which
suggests employees as the likely source of bias.
Biased news will
alienate many potential customers with centrist or right-of-center views. A
smaller audience reduces advertising revenues and profit. The second question
then is, why do profitmaximizing owners allow their reporters to indulge their
liberal views at the organization’s expense? Professionalism gives journalists
considerable leeway to set standards for the quality of their product. But if
journalistic independence and professionalism only hurt the news organization’s
bottom line, owners could eliminate such independence.
The third
question concerns entry of conservative news organizations into the market to
undercut the liberal cartel. Entry could involve a change in programming at an
existing organization, or creation of a new organization. If all current
organizations present liberal news, a single right-of-center organization would
have half of the political spectrum to themselves. The only conservative firm
in a liberal-dominated market could likely draw larger audiences than possible
as a member of the cartel. A liberal news cartel creates a profit opportunity
for a news organization willing to listen to the conservative critics.
The probability
of sustaining a news cartel diminishes as the number of firms in the market increases.
Technology, and to a lesser extent regulation, have combined to keep the number
of news organizations of any one type in the national news market relatively
small.
Until about 20
years ago there were basically only the three television networks, two weekly
magazines, and perhaps four newspapers in the national media market. All but a
handful of cities had only one daily newspaper and the three network affiliates
with news operations; even multi-paper cities have only two or three dailies.
As cable television, satellite printing, and the Internet increase the number
of news organizations in a truly national news market, the profit incentive for
product diversification will become overwhelming. I conclude that the
conditions for sustaining a news cartel, while tenuous in the past, will vanish
in the near future. The news industry will almost certainly feature
organizations catering to a range of political perspectives.
We require some
means of defining the existence of bias in the news. Bias cannot merely be in
the eyes of the beholder, because each of us would like news stories to confirm
the validity of our views. Consequently, I apply the spatial model of politics
to the news media’s product.5 The most reasonable way to define bias is relative
to the views of the median voter. A liberal news organization would be located
to the left of the median voter. And the deviation from the median voter’s
position must be nontrivial for bias to be a policy issue of significance.
I do not consider
here the details of classifying a story or locating an organization on the
spectrum based on their many stories.6 A question of relevance, though,
concerns charges of bias in the news media as a whole. Must all organizations
share the same partisan orientation for the media to be biased? Or is the mean
or median of organization positions the appropriate measure? Operation of the marketplace
of ideas requires only that all views get a hearing before the court of public
opinion, which suggests bias adversely affects this market only if it extends
across all organizations. A situation where most organizations exhibit a
left-of-center bias but at least one is located to the right of the median
voter can result from product differentiation and a majority liberal audience.
The rejection by citizens of partisan views of either the left or right does
not represent inefficiency in the transmission of political views. As a
practical matter, conservative critics charge that all the major news organizations
share the liberal bias. Bozell and Baker (1990), for instance, do not attempt
to make fine distinctions between the bias of the television networks.
Consequently, I will assume that liberal news bias involves all the major
national news organizations offering a product which deviates significantly
from the median voter.7
If all news
organizations must offer essentially the same left-ofcenter news to validate
the critics’ charges, then the liberal media charge requires a news cartel.
Typically economists describe a cartel as restricting output to raise price and
generate supra-competitive profits. Advertising is the dominant source of
revenue for news organizations, so the media are not trying to raise the price
they charge to customers (readers and viewers). Yet if all organizations supply
left-of-center news, customer preferences are not being fully satisfied. Furthermore,
a cartel is subject to defection by member firms; a firm can increase its sales
and profits at the others’ expense by lowering its price.
The following
example illustrates that a liberal news media can be reasonably described as a
cartel. Suppose for simplicity we distinguish three positions on the political
spectrum: liberal, moderate, and conservative. Each consumer has a preferred
position on this spectrum; for concreteness, let there be 300 moderates, 150
liberals, and 150 conservatives in the audience. Each news organization must
adopt a position on the political spectrum for their news coverage. Consumers patronize
the organization whose news coverage comes closest to their personal views;
conservatives will choose the conservative outlet if one is available. A group
of consumers divides equally among the competitors if two or more news
organizations offer the closest match. Assume that all consumers consume news,
even if their favored news is not provided.
One news
organization, under these circumstances, could locate anywhere along the
political spectrum and attain an audience of 600. A liberal news monopoly is
not implausible. Suppose two or more organizations exist. If all organizations
provided liberal news they will split the market, which is the same result as
if all organizations provided moderate news. Yet a liberal news cartel is
unstable. When two firms participate in a liberal cartel, each organization has
an audience of 300. But one firm increases its audience to 450 if it shifts its
news location to moderate, since it captures all the moderate and conservative news
consumers. Firms have an incentive to defect from a cartel as the number of
organizations increases. With two or three firms we would expect convergence to
the middle; all organizations would provide moderate news and split the market.
Once four firms were in the market, product differentiation could result: two
firms with moderate news, one liberal and one conservative organization, each
with an audience of 150. Although a firm might then present liberal news, a
uniform liberal news media does not result. Given the nature of the strategic
problem the affected firms face, we can reasonably characterize the liberal
media charge as requiring a sustainable cartel.8
What is the
source of bias? Political bias in the news media has three potential sources.
On the demand side, a disproportionate number of consumers of the news might
have liberal views with the news media merely providing the product their
customers demand. Indeed, a demand side explanation does not imply market
failure. Either owners or journalists could also be a supply-side source of
biased news. A supply-side source of liberal bias would constitute a market failure,
and raises the question of intentionality. Is the liberal media part of a
conspiracy? A supply-side explanation of bias must address the stability of the
liberal news cartel.
The set of news
consumers may differ from the set of voters. If liberals constitute a
disproportionate percentage of news consumers, the median voter may regard the
news media’s programming as liberal. Yet the media are simply responding to
their customers’ preferences. News organizations serving their median customer
will appear liberal from the perspective of the entire electorate. A demandside
explanation does not imply that the news market fails.
Clearly, this is
an empirical claim. Some evidence suggests that news consumers may be more
liberal than voters. Robert Entman (1989: 141–43) provides evidence (drawn from
the Michigan Current Population Survey) that nonvoters are more liberal than
voters, which implies voters are not a representative sample of the population.
Newspaper and newsmagazine subscription rates increase with education, and
surveys typically show college graduates are more liberal than the population
as a whole. Brian Goff and Robert Tollison (1990) show that newspaper
circulation per capita across states increases with a measure of the
liberalness of state voters. The available evidence, however, is far from
conclusive in establishing news consumers as a source of bias.
Two important
caveats apply even if the median news consumer were demonstrably more liberal
than the median voter. First, not all viewers are to the left of the median
voter. News organizations will engage in product differentiation when
profitable. If the majority of news consumers are located to the left of the
median voter, this might explain why most news organizations provide liberal
news. But all organizations must locate left of the median voter to explain the
liberal media. The audience must be quite skewed for profitmaximizing news
organizations to all provide liberal news. Suppose we increase the proportion
of consumers with liberal views in the previous model of program choice. If
three organizations exist, 400 of the 600 consumers must be liberal so that no
organization wishes to provide moderate news. The degree of divergence from the
voting population required increases with the number of media organizations. With
four news organizations at least 450 of the 600 consumers must prefer liberal
news to sustain a liberal news cartel.
Second, even if
we establish that the media and its viewership are left-of-center, this
correlation does not determine the direction of causality. A demand-side
explanation assumes that causality runs from audience to programming, that
liberals disproportionately consume news and the media provide the product they
demand. But causality could easily run in the other direction: the media’s
liberal bias could alienate conservatives, producing over time a liberal
audience. For causation to run from programming to audience, a potential
viewer’s news consumption decision (versus other forms of entertainment) must
depend upon the product’s partisan bias. In the context of the model of
programming choice, conservatives must choose not to consume liberal news if
conservative (or moderate) news is not offered.
Certainly most
people would prefer news which conforms with and reinforces their political
beliefs, and should turn away from news which challenges these beliefs. I am
considering a liberal national news market. Given the other sources of news,
like local newspapers or television, disaffected conservatives might stop
watching. A liberal news media could produce a liberal audience.9 Because of the
difficulties in explaining a news cartel based on a liberal audience, I turn now
to supply-side explanations.
Economists
typically assume firms maximize profit, but owners as consumers wish to
maximize utility. Utility maximization does not necessarily mean profit
maximization since some amenities can be more easily acquired in production
than through the market. Owners will generally be willing to trade some profit
for other goals. Media owners can use their news organizations to advance their
favored political causes. Owners’ ideology is a potential supply-side source of
bias in the media.10
Clearly many
media owners over the years have been strongly identified with political
parties or causes and used their organizations to achieve these goals.11 Henry Luce,
founder of Time, remarked of his magazine’s coverage
of the 1952 presidential election: “Eisenhower was right for the country for a
large number of reasons, therefore, it was Time’s duty to
explain why the country needed Ike. Any other form of objectivity would have
been unfair and uninvolved” (quoted in Halberstam 1979: 59). The Chandler
family for many years ran the Los Angeles Times to benefit their
favored (mostly Republican) candidates. Today media moguls Ted Turner and
Rupert Murdoch are known for their strong and contrasting political views.
The political
views of owners provide weak grounds for a liberal news cartel. The problem is
not that owners will never sacrifice profits for political goals; many
instances of this abound. Rather, all media owners must be willing to sacrifice
profits for liberal political goals to sustain a news cartel. This assumption
is open to two objections. First, although examples of liberal media owners can
be found (Philip and Katherine Graham, Ted Turner), more often prominent media
organizations have been used by owners to advance conservative causes.
Many owners
notable for their political views have right-of-center views: Luce of Time, Chandler of the
Los Angeles Times, William Paley of CBS, and Robert
McCormick of the Chicago Tribune. Proponents of the liberal media
proposition have not established convincingly that all media owners are
liberal.12
Second,
contending that all media owners trade profits for liberal political goals
clashes with evidence that profit maximization is an increasingly important
goal for the media and the changing structure of media ownership. While we lack
an objective measure of concern with the bottom line across organizations or
over time, increased emphasis on profits following the acquisition of a
newspaper or television station by a media conglomerate is a familiar refrain
(Bagdikian 1997, Underwood 1993). The tendency of some media scholars to equate
cost-cutting measures with profit maximization confuses matters; a news
organization may pursue profits by offering a high-quality product.
Cost-cutting does not demonstrate increased emphasis on profit
maximization, only perhaps a change in product offered. Nonetheless, most close
observers of the media report a growing emphasis on the bottom line. And
surveys of journalists report a perceived increase in emphasis on profit
margins by management (Underwood 1993: 117–26, Weaver and Wilhoit 1996: 60–67).
Unless these observations are entirely off the mark, owners must be less
willing to trade profits for political goals now than in the past.
The growth of
corporate ownership of the media strengthens the case for increased emphasis on
profit as a goal. Newspapers in the United States typically began as
family-owned businesses, but the inheritance laws forced the sale of or
issuance of stock by most of these companies (Lacy and Simon 1993, Bagdikian
1997). The growth of newspaper groups has been substantial. In 1920, 31 groups
owned 7.5 percent of the nation’s daily newspapers. By 1986, 127 groups owned
69.9 percent of daily papers. In addition, the size of groups increased, from
4.9 papers per group in 1920 to 9.1 papers per group in 1986. Gannett, the
largest newspaper group, owned 90 daily papers in 1986 (Lacy and Simon 1993:
132–33). The growth of corporate ownership has occurred in other media as well,
highlighted by takeovers of all three major U.S. television networks in the
1980s.
Media outlets are
also increasingly owned by widely held public corporations in contrast to
family-owned or narrowly held corporations. The breadth or concentration of
ownership affects the likelihood of pursuit of goals other than profit
maximization. Consensus on goals besides profit is more
likely with a narrowly owned family business than with thousands of
stockholders. Profit is a goal which everyone who invests in a business can
agree on. Other goals are far less general. Even if all stockholders agree news
organizations should champion causes, they will disagree about which
causes
to champion: environmental protection, the labor movement, protection of property
rights, school choice, and so on. Some stockholders might desire a liberal bias
and some a conservative bias, and others not interested in politics will wish
to pursue only profit maximization. Harold Demsetz and Kenneth Lehn (1985) show
that media companies have more concentrated ownership and more family ownership
than other types of firms. Demsetz and Lehn argue that the amenity potential of
media firms (that they can be used to influence public opinion) explains their
concentrated ownership. Nonetheless, their evidence suggests that an exogenous
decrease in the concentration of media ownership should produce less
consumption of amenities and more emphasis on profit.
Journalists
themselves are a second supply-side source of bias. Charges of a liberal bias
in the news place great weight on surveys revealing the liberal views of a
majority of journalists. The survey evidence is consistent and strong. In
Robert Lichter, Stanley Rothman, and Linda Lichter’s (1986: 20–53) 1979–80
survey of journalists at elite media organizations, 54 percent of respondents
identified themselves as left-of-center, versus 17 percent right-of-center. The
journalists who voted for a major party candidate in presidential elections
between 1964 and 1976 overwhelmingly went for Democrats: Lyndon Johnson 94
percent, Hubert Humphrey 87 percent, and George McGovern and Jimmy Carter 81
percent each. David Weaver and Cleveland Wilhoit (1996: 15–19) in 1992 found
that 47.3 percent of journalists identified themselves as left or left-leaning
versus 21.7 percent right-of-center. In terms of party affiliation, 44 percent
identified with the Democrats versus 16 percent Republicans and 34 percent
Independents. Of the Washington journalists surveyed by Stephen Hess (1981:
87), 42 percent identified themselves as liberal versus 19 percent
conservative.13
How did the
journalism profession come to be dominated by liberals, and how is this
domination maintained? An established partisan bias can be sustained through
self-selection. The costs of membership in the organization for individuals
with dissenting viewpoints can be extremely high. Colleagues will make
statements about politics at the water cooler, on breaks, and at lunch which
dissenters find objectionable. Saying nothing in response often involves a
psychic cost, while responding brings potential social sanctions (Kuran 1995).
A bias in the topics of media coverage provides dissenting individuals with inherently
disconcerting assignments. Dissenters bear a higher cost of putting aside their
feelings to do the story and do it well, and failure to do so results in lower
quality job performance. Dissenters will tend to quit at a higher rate and have
lower job performance, as measured by supervisors sharing the organization’s
dominant values. Lower performance leads to a higher probability of dismissal
and lower probability of promotion or desirable future assignments. And
dissenting individuals might come to adopt the organization’s dominant
viewpoint, not out of strategic considerations but because of belief plasticity
(Klein 1994). Individuals with the dominant view will find a career in the
media more rewarding, and are more likely to invest in the training necessary
to enter the field.
This leaves unresolved
the establishment of bias. One possible answer could be the joint distribution
of journalistic talent and political views across the population. People with
the talent, temperament, and personality to be journalists might also be
inclined toward liberal political causes:
Now the
thing that God puts in a man that makes him a creative person makes him very
sensitive to social nuances and that sort of thing. And overwhelmingly—not by a
simple majority, but overwhelmingly—people with those tendencies tend to be on
the liberal side of the spectrum. People on the conservative side of the
political spectrum end up as vice presidents at General Motors. . . . If you go
out and put together a television news department, and you hire people purely
on the basis of their ability with no regard whatever for their ideology . . .
and you put together 200 men you are quite likely to end up with 175 men who
would be quite proud to call themselves liberal [quoted in Keely 1971: 51].
This argument
rests entirely on the empirical nature of the distribution of talent and
ideology in society. The predominance of liberal political views in Hollywood
and the arts does lend some credibility to the premise. Relative opportunity
costs can also explain career choices. Individuals who strongly object to the
current socioeconomic order will find it difficult to justify to themselves the
pursuit of wealth and personal goals through the system. A job which allows one
to work toward changing the system, in however modest a fashion, provides
higher nonpecuniary benefits for such a person. Journalists are intellectuals,
a class long dominated by individuals hostile to business.
A third source of
biased journalists lies in the screening function played by journalism schools
which train an increasing fraction of reporters, particularly at leading news
organizations.14 As long as news organizations require the acquisition of these
credentials, journalism programs choose the set of candidates available to
employers.
Surveys reveal
that journalism students have even more pronounced liberal views than working
journalists. Indeed, in 1982, 85 percent of Columbia Graduate School of
Journalism students identified themselves as liberal, versus 11 percent
conservative (Lichter, Rothman, and Lichter 1986: 48). Faculty could
conceivably gear the curriculum to attract students with liberal leanings and
repel young conservatives. Selection could extend to grades and
recommendations.15
Conservative critics
typically see the survey evidence, supported by some content analysis, as
settling the issue. Objectivity in reporting is impossible; reporters cannot
abandon their worldview in attempting to report on daily events. Examples of
intentional bias, such as slanting stories to advance their views given the
constraint of surface neutrality, are rare (see Efron 1971 for some examples,
though). More generally, advocates of the liberal media hypothesis argue that bias
is unintentional and unavoidable. Liberal reporters may simply know more
liberal sources and find their explanations of events more persuasive. As
Bozell and Baker (1990: 3) note:
But
though bias in the media exists, it is rarely a conscious attempt to distort
the news. It stems from the fact that most members of the media elite have
little contact with conservatives and make little effort to understand the
conservative viewpoint. Their friends are liberals, their sources are liberals,
what they read and hear is written by liberals.
The surveys
establish though only an inherent tendency in journalists’ reporting, if
unchecked. Media owners and their agents (top editors and producers) could take
steps to prevent journalists’ personal views from biasing their reporting to
prevent the content of news from being liberal. Critics of the liberal media
proposition point to the incentive for the major news organizations to push
their owners’ conservative, pro-free–market politics.16 The critics on
the left properly point out a weak link in the liberal media argument. I
believe, however, that these critics misidentify the media owners’ motive for
controlling journalists. Corporate media owners will be more interested in
maximizing profits than pushing a political agenda. Being a member of a liberal
news cartel costs individual media organizations (and perhaps the industry as a
whole) potential profit. Corporations will try to control liberal journalists
if liberal news reduces profit. Proponents of the liberal media thesis must
explain why media owners fail to control their employees. Owners apparently are
content to let journalists indulge their liberal views at the expense of the organization’s
profit.
A conflict of
interest between employers and employees does not imply that workers will get
their way. All owners want their employees to exert effort to achieve the
organization’s goal (profit). Effort is costly for employees to exert, and
owners are residual claimants for (at least most of) the company’s profits.
Employers, though, have a number of methods of controlling shirking by
employees. Firms can design employment contracts to overcome the
principal-agent problem.17 The question for the liberal media thesis then becomes why control of
employees might be more of a problem for news organizations than other
businesses. The nature of journalism provides a possible answer. Journalism is
a profession; reviewing performance and detecting shirking requires a
significant level of professional expertise. Professionalism creates certain
standards that must be met in choosing sources and presenting a story. Only
someone familiar with these standards can criticize a journalist’s performance
(detect shirking). Objectivity provides reporters with a formula to use against
critics of a story (Tuchman 1978, Miraldi 1990, Bennett 1996).
Owners, though,
can hire editors or producers with knowledge of journalistic practices to
control liberal bias in reporters. As an empirical matter, owners do not seem
to be choosing top supervisory personnel to offset liberal reporters. Weaver
and Wilhoit (1996: 17) find that 57.2 percent of executives (who supervise
editorial personnel) at prominent news organizations describe themselves as
left or left-leaning, compared to 59.4 percent of staffers at these
organizations. But even an editor intent on controlling journalists would find the
task daunting. News stories are not a standardized product; rather, each is
unique. An editor very familiar with the techniques of reporting will typically
lack the detailed knowledge necessary to criticize the choice of sources on a
story or beat. Hess (1981) documents the virtual lack of editorial oversight
for Washington reporters, a particular specialist elite within the set of
journalists. Print reporters particularly enjoy a great deal of autonomy by
other measures as well: reporters initiated 69 percent of their stories in his
study and 51 percent of stories were not edited by the home office at all (Hess
1981: 6, 8). Another factor weakening editorial control is the use of anonymous
sources. Editors cannot evaluate the credibility or potential bias in the range
of sources for a story if the reporter will not divulge names (Isaacs 1986).
Professionalism
and peer review increase autonomy and independence in many fields. Academics
use peer review in tenure decisions, department rankings, and accreditation
decisions to promote goals which they value (lower teaching loads, graduate
programs, more support for research) over those of their institutions
(undergraduate teaching).18 If professionalism creates slack in the principal-agent relationship,
journalists with liberal personal political views could take up this slack in
the form of biased news coverage. In this view liberal bias in the news simply
reflects the potential for discretion, just as peer review contributes to a
liberal bias in academia.
The above
arguments suggest that the cost of controlling reporters will be high. News
organizations will closely control journalists only if the benefit of the
effort exceeds the cost. The gain from controlling bias, the reduction in
profit from liberal news, might be small for at least two reasons. First,
allowing partisan news coverage could reduce the cost of producing news
coverage. Individuals with strong political views will accept lower pay to do
the type of reporting they believe in.19 If the majority of journalists have
left-of-center views, liberal news might cost less to supply than unbiased
news. Traditionally salaries in the news industry have been modest. And the
cost of trying to produce a right-of-center product with predominantly liberal
reporters could be quite high. Profit-maximizing news organizations will trade
off reduced revenue from a smaller audience if offset by a larger reduction in
cost.
Second, the
reduction in audience due to liberal news may be small. Americans have little
interest in politics in general and foreign news in particular; clearly
reporters have more interest in the political game than the median audience
member. News coverage, particularly of politics, has traditionally been a
financial drain (Schudson 1995). CBS established its distinguished news
department to enhance the network’s reputation, not because of a direct return
on the investment (Halberstam 1979).20 If lost revenue from biased coverage is
$5 million a year instead of $500 million, news organizations have little
incentive to crack the whip and keep their reporters in line. In terms of the
programming model, some portion of consumers may not be able to discern the
partisan position of an organization’s coverage. These consumers might divide
evenly among the different organizations regardless of their position on the
spectrum. Also the number of organizations producing in a given news market
determines the audience gain at stake. A news monopoly can indulge partisan
news at a lower cost than an outlet which faces competition. Hence monopoly
newspapers should exhibit more bias than magazines.
Even if existing
news organizations did overcome their difficulties and maintain a liberal
cartel, other organizations might enter the market and challenge the prevailing
bias. An entrant could capture the moderate and conservative audience.
Successful cartels must have means of punishing defectors and preventing entry.
Yet proponents of the liberal media thesis offer little discussion of the
barriers to entry and punishment mechanism which sustain the news cartel.
Licensing by the
Federal Communications Commission constitutes a barrier to entry for radio and
television. Indeed, for many years the FCC helped maintain the oligopoly
position of the three over-the-air networks and delayed the development of
cable television. The advent of cable, however, has reduced restrictions on
entry, even though regulations still limit access to cable systems. Local
access and educational programs take up scarce channels, while the must-carry
provisions for local stations in the 1992 Cable Bill restrict access of new stations
to cable systems. The last two decades have nonetheless witnessed the
successfull entry of CNN and Fox to the television market.
The First
Amendment exempts the newspaper industry from direct regulation. Entry into the
market is possible, at least in principle, as the example of USA
Today demonstrates. Economies of scale do create declining average costs for
newspapers over a given range of production, and circulation interacts with
advertising rates. Consequently, only the largest cities can sustain two or
more daily papers and significant barriers to entry exist in the local
newspaper market (Lacy and Simon 1993). Only two new major metropolitan daily
papers were established in the United States during the decade of the 1980s:
the St. Louis Sun, which lasted less than a year, and
the Washington Times, which loses money but survives due
to the owners’ willingness to subsidize the paper’s partisan output (Lacy and Simon
1993: 134). And USA Today had a total pre-tax loss of $800 million
in its first nine years of operation (Underwood 1993: 99). Entry barriers in
the magazine industry are low and numerous policy magazines exhibit a partisan
orientation, including National Review, Mother
Jones, The American Spectator, Village
Voice, and The New Republic. Note that partisan policy magazines
span the political spectrum. The uniformity of bias which the liberal media thesis
requires does not exist. While partisan policy magazines do not attain a market
presence close to Time or Newsweek, failure to win
a large audience is not a barrier to entry.
The public
sector’s considerable public relations apparatus subsidizes mainstream
journalists and creates a barrier to entry for new firms. As Paul Weaver (1994:
212) explains:
Much
more consequential is the massive public information and public relations
apparatus maintained by government at all levels at a cost to taxpayers of
billions of dollars a year. . . . Usually a single call to a single
press-relations person is enough to produce an entire day’s or week’s worth of
interviews and inspection tours, supplemented by nearly any amount of official
information, often on extremely short notice. Doing it yourself would take days
if it could be done at all. This is a subsidy that isn’t available to all
comers. An ordinary citizen seeking interviews and observation opportunities out
of mere personal interest wouldn’t get the time of day. A representative of a
nonaccredited or out-of-field medium is also apt to consider press relations a
way of providing some journalists with protections from competition from other
journalists.
Journalists from prominent
news organizations are also most likely to have access to top administration
officials and members of Congress and receive leaks and scoops. The ability to
consistently get the story first provides a distinct advantage to incumbent
organizations. If barriers to entry prevent a new (conservative) organization
from entering the news market, ideologues could always purchase an existing organization.21 All three major
television networks were sold in the 1980s. Only the news division need be
modified in the organization; a television network’s prime time and daytime
lineups could be left intact. Given the alleged prevailing liberal bias,
product differentiation could easily increase audience and advertising
revenues. In addition to this profit incentive, the marginal benefit to
conservatives of one countervailing source of news should be high. Further,
wealthy conservatives willing to trade profits for political goals might invest
in the takeover effort to break the liberal news cartel.
What prevents a
change of orientation at an existing news organization, or how can cartel
members punish a cheating firm? Professionalism provides journalists some
ability to enforce their accepted practices on news organizations through peer
evaluation.22 If the new owners of a television network set about trying to impose a
conservative perspective on the news product, other journalists could criticize
the compromise of journalistic independence and deterioration of the quality of
coverage. A reduction in the professional evaluation of the organization’s
product may adversely affect audience size and advertising revenue. Liberal
journalists could themselves raise costs for a new conservative news
orientation through work slowdowns and resignations. Indeed, more than 60 staff
members of the Chicago Sun-Times took generous
severance pay and resigned when Rupert Murdoch purchased the paper (Shawcross
1997: 175). In time new staff could be hired or trained, but a short-run
increase in cost and deterioration of quality would be nearly inevitable.
Ultimately the
reader must judge the strength of the liberal media thesis. I have taken a
different approach to the topic and raised some economic objections. Liberal
bias must have a source and maintenance of bias effectively requires a news
cartel. The source of bias remains unclear, although journalists themselves are
the most plausible source. Even if so, the establishment of ideological bias
within the profession is a mystery. And two weak points still remain in the cartel
argument: why media owners allow journalists to indulge their liberal views at
the expense of potential profits, and what protects the cartel against
defectors and entrants.
Journalistic
professionalism and independence play a role in answering both objections.
Professionalism and the uniqueness of each story make the reporter’s judgments
(regarding sources, a lead, and a frame) difficult to second-guess, creating an
environment in which the journalist’s political views could matter. A news
organization which attempts to impose a conservative slant in its coverage will
meet resistance from professional journalists who see independence from
ownership as critical for good reporting. The larger question for the liberal
media thesis then is why news organizations accept journalistic professionalism.
Perhaps media owners do not have the will to battle their employees over money
and working conditions, preferring to accommodate employees’ demands to improve
their public image. Yet newspapers (including the Washington
Post) have battled their printers’ unions to control costs. So a general
unwillingness to oppose employee demands will not provide an answer.
Professionalism must generate some benefit for news organizations. A higher
quality product as judged by journalists could possibly translate into larger
audiences and advertising revenues. Yet evidence casts doubt on this.
Coverage of
Watergate won Bob Woodward and Carl Bernstein and the Washington
Post many accolades but did not increase profitability as measured by stock
price (McChesney 1987). Many journalists decry the changes implemented by
Gannett first in the USA Today and later in their chain papers to
create a more reader-friendly product (Underwood 1993: 95–105). Budget cuts by
CBS News in the 1980s lowered the quality of their product in the view of
journalists. Are these organizations worse off financially as a result?
Ownership is most likely to indulge professionalism if skilled journalists
value independence and will trade salary in return. Competition among news organizations
for the best journalists determines the value of their compensation package.
The form of the package depends on journalists’ preferences and the cost to the
organization of supplying types of compensation. Profit rises as long as the
news journalists want to produce does not reduce the audience too much.
Two sources of
potential barriers to entry or cost disadvantage to new organizations lie in
the public sector: FCC limits on the number of television stations and government’s
public relations apparatus. Public-sector actions seem an inadequate basis on
which to explain a liberal news cartel. Republicans held the presidency for 28
of the 40 years between 1952 and 1992. The White House could have directed scoops
and interviews to conservative reporters and news organizations during these
years. If FCC regulations were the source of a liberal news cartel, the Nixon
administration could have knocked out a liberal news cartel via direct action
instead of having Vice President Agnew speak out about bias.23 Government
regulation supported cartels in trucking, railroads, and airlines until the
late 1970s; was news merely another example of government-sponsored
cartelization? I suspect not. Support for trucking and airline cartels was less
risky for politicians than a news cartel because politicians had no direct
stake in the output of these industries.
Technology will
likely eliminate any prospects of a liberal news cartel. A successful cartel
typically requires a small number of firms. The cooperation necessary to
maintain a cartel is more likely and the ability to identify and punish cartel
violators greater with a modest number of firms. The past two decades have
witnessed an increase in the number of national news organizations; CNN and USA
Today are just two examples. Many regional newspapers take on more of a national
character due to web sites and satellite printing. Fox News is slowly becoming
available to more cable systems.
As the number of
news organizations increases, the incentive for partisan differentiation by
some becomes quite strong. Partisan media bias, as defined here as a deviation
of the product of all news organizations from the position of the median voter,
may be a thing of the past. The minimum efficient scale of production limited
and then eliminated competition between papers in all but the largest
metropolitan areas. And the rise of the three television networks combined with
FCC regulation limited the number of television news organizations. But as the
number of television networks and national newspapers expands, the potential
for a successful news cartel diminishes. A proliferation of reporting
techniques and partisan views should occur in the decades to come as the number
of national news organizations increase. Fox News already attempts to
differentiate its product from the perceived liberal media bias. At least some
of the new organizations should find it profitable to cater to conservative viewers.
I have addressed
the question of partisan political bias in the news, a bias in favor of the
Democratic party or liberal issue positions at the expense of Republican
candidates and conservative or libertarian causes. News gathering and reporting
can have other effects on the news, biases which may be as significant as party
bias. Coverage of government activities might be excessively positive because
reporters fear retaliation (lack of access, complaints to superiors) from
government officials over negative stories. Fairness in reporting always allows
government officials to be quoted and thus to provide their spin on events.
Academic critiques of journalism in fact focus more on these other biases.
Analysis of a possible link between journalistic practices and the expansion of
government, though, remains a subject for future research.
1Charges of
liberal bias typically concern the national news media, including the
television networks, the major weekly newsmagazines, and leading papers like
the New York Times and Washington
Post. These organizations comprise only a portion of the news industry. Other
portions of the industry, notably talk radio and local newspapers, have been
accused of a conservative bias. Liberal bias in the national news market is of
concern since the national media have a greater impact on the political agenda
than these other outlets do.
2See Bozell and
Baker (1990) and Baker (1994) for a sample of the types of evidence of liberal
media bias.
3Documentation of
bias requires considering the impact of normal incentives and procedures on
reporting. Washington reporters seek access to inside sources due to
competition for stories. Reporters and sources engage in a repeated game, so
mutual cooperation is likely to evolve; hostile reporters could see their
administration sources dry up. If we observe reporters not being critical of a
Democratic administration, the behavior may not be a result of bias.
4Although claims
of both a liberal and conservative bias in the media must face these types of
questions, this paper focuses on potential pitfalls of a liberal news cartel.
Sutter (2000) considers the charge that corporate ownership and advertising
creates a conservative, probusiness bias in the media.
5On the spatial
model see Downs (1957). Goff and Tollison (1990) and Endersby and Ognianova
(1997) have previously applied the spatial model to the political content of
the news media.
6Every
organization runs many stories which will likely exhibit different points of
view, and many stories will lack any partisan content. An organization may be
located based on a measure of central tendency (the median or mean story) or a
measure of dispersion (variance or extreme stories). Different measures may
produce different evaluations of an organization. I require only that some
means of locating organizations exist.
7Bias across a
majority of organizations can affect policy outcomes. Suppose the median voter’s
preference helps determine policy outcomes. If two out of three television
networks exhibit a liberal bias and voters likely to be influenced by bias are
evenly distributed across the three networks, the median voter will end up more
liberal than otherwise. Political parties trying to win elections would
understandably be concerned about bias that changes even 1 or 2 percent of
votes.
8The exact outcome
in models of this type depends on the assumptions made (Owen and Wildman 1992:
64–100). Allowing audience members not to consume news if no organization provides
their most preferred news product would strengthen the incentive for a single
firm to provide moderate news. Also a liberal cartel might have a smaller total
audience than a moderate cartel. The difficulty of maintaining a liberal cartel
would still result.
9Of course, if we
had a measure of media bias, we could examine the causality issue directly.
10Media owners
might also desire a large and activist government as a means of increasing demand
for their product (Crain and Tollison 1997). The gains from such a strategy
will likely be remote and modest, since rational ignorance limits citizens’
demand for political information. I focus on ideological consumption by owners.
11Prior to the rise
of commercial media (the penny press) many newspapers were supported by
political parties. I am concerned with the use of commercial media for
political purposes.
12Bozell and Baker
(1990: 86–98) present evidence concerning the giving patterns of media organizations’
charitable foundations.
13Of journalists in
the Brookings survey, 51 percent agreed that the Washington press corps had a
political bias. Of those who believed a bias existed, 96 percent perceived a
liberal bias versus 1 percent perceiving a conservative bias (Hess 1981: 87).
14In 1992, of the
working journalists surveyed by Weaver and Wilhoit (1996: 35), 82 percent had
at least a four-year college degree, and 11.4 percent had a graduate degree.
Furthermore, 40 percent of the journalists with college degrees had majored in
journalism.
15This impact need
not involve conscious discrimination on the part of professors. Students who find
a course interesting tend to work harder and do better. And professors tend to provide
a higher evaluation of the abilities of protege´ s.
16As Hertsgaard
(1988: 85–6) puts it: The deeper flaw in the liberal-press thesis, however, was
that it completely ignored those whom journalists worked for. Reporters could
be as liberal as they wished and it would not change what news they were
allowed to report or how they could report it. America’s major news
organizations were owned and controlled by some of the largest and richest
corporations in the United States. These firms were in turn owned and
controlled by individuals whose politics were, in general, anything but
liberal. Why would they employ journalists who consistently covered news in
ways they did not like? For other examples of this argument see Bagdikian
(1997), Parenti (1986), and Kellner (1990).
17Indeed, media
scholars (Schiller 1981, Endersby and Ognianova 1997) argue that objectivity or
fairness in reporting was introduced by news organizations to target news
toward a mass audience.
18On the
difficulties which monitoring professionals creates for firms, see Alchian and Demsetz
(1972). Other professions secure occupational licensing, which allows members
to reduce entry into the field and create rents for practitioners.
19Frank (1996)
presents survey evidence concerning students’ willingness to accept lower pay
for jobs offering greater intangible moral rewards.
20This raises the
question of why the TV networks provide news service at a loss. A combination
of habit and the public service requirement imposed by government regulators probably
explains this.
21A group of
conservatives did in fact attempt to buy CBS in 1985.
22Efron (1971)
suggests that members of the journalistic community who criticize the prevailing
liberal bias face considerable ostracism from other professionals.
23The Nixon
administration did threaten not to renew the licenses of television stations owned
by the Washington Post company.
Alchian, A.A.,
and Demsetz, H. (1972) “Production, Information Costs, and Economic
Organization.” American Economic Review 62: 777–95.
Bagdikian, B.
(1997) The Media Monopoly. 5th ed. Boston:
Beacon Press.
Baker, B.H.
(1994) How to Identify, Expose and Correct Liberal Media Bias. Alexandria,
Va.: Media Research Center.
Bozell, L.B., and
Baker, B.H. (eds.) (1990) That’s the Way It Isn’t: A Reference Guide to Media
Bias. Alexandria, Va.: Media Research Center.
Crain, W.M., and
Tollison, R.D. (1997) “Expansive Government and the Media.” Unpublished paper,
George Mason University.
Demsetz, H., and
Lehn, K. (1985) “The Structure of Corporate Ownership: Causes and
Consequences.” Journal of Political Economy 93 (December): 1155–77.
Downs, A. (1957) An
Economic Theory of Democracy. New York: Harper and Row.
Efron, E. (1971) The
News Twisters. Los Angeles: Nash Publishing. Endersby, J.W., and Ognianova, E. (1997)
“A Spatial Model of Ideology and Political Communication.” Harvard
International Journal of Press/Politics 2: 23–39.
Entman, R.M.
(1989) Democracy Without Citizens. New York:
Oxford University Press.
Frank, R.H.
(1996) “What Price the Moral High Ground?” Southern
Economic Journal 63(1):1–17.
Goff, B., and
Tollison, R.D. (1990) “Why Is the Media So Liberal?” Journal
of Public Finance and Public Choice 1:13–21.
Halberstam, D.
(1979) The Powers That Be. New York:
Alfred A. Knopf.
Hertsgaard, M.
(1988) On Bended Knee: The Press and the Reagan Presidency. New York:
Farrar, Straus Giroux. Hess, S. (1981) The Washington Reporters. Washington,
D.C.: Brookings Institution.
Isaacs, N.E.
(1986) Untended Gates: The Mismanaged Press. New York: Columbia
University Press.
Keely, J. (1971) The
Left-Leaning Antenna: Political Bias in Television. New Rochelle,
N.Y.: Arlington House.
Kellner, D.
(1990) Television and the Crisis of Democracy. Boulder, Colo.:
Westview Press.
Klein, D.B.
(1994) “If Government Is So Villainous, How Come Government Officials Don’t
Seem Like Villains?” Economics and Philosophy 10: 91– 106.
Kuran, T. (1995) Private
Truths, Public Lies. Cambridge Mass.: Harvard University Press.
Lacy, S., and
Simon, T.F. (1993) The Economics and Regulation of United States
Newspapers. Norwood N.J.: Ablex Publishing.
Lichter, S.R.;
Rothman, S.; and Lichter, L.S. (1986) The Media Elite. Bethesda, Md.:
Adler and Adler.
McChesney, F.S.
(1987) “Sensationalism, Newspaper Profits and the Marginal Value of Watergate.”
Economic Inquiry 25:135–44.
Miraldi, R.
(1990) Muckraking and Objectivity. New York:
Greenwood Press.
Owen, B.M., and
Wildman, S.S. (1992) Video Economics. Cambridge, Mass.: Harvard University
Press.
Parenti, M.
(1986) Inventing Reality. New York: St. Martin’s Press.
Schiller, D.
(1981) Objectivity and the News. Philadelphia:
University of Pennsylvania Press.
Schudson, M.
(1995) The Power of News. Cambridge, Mass.: Harvard University
Press.
Shawcross, W.
(1997) Murdoch: The Making of a Media Empire. New York: Touchstone
Books.
Sutter, D. (2000)
Political Bias in the News: A Skeptical Examination. Unpublished book
manuscript, University of Oklahoma.
Tuchman, G.
(1978) Making News. New York: Free Press.
Underwood, D.
(1993) When MBAs Rule the Newsroom. New York:
Columbia University Press.
Weaver, D.H., and
Wilhoit, G.C. (1996) The American Journalist in the 1990s. Mahwah, N.J.:
Lawrence Erlbaum.
Weaver, P.H.
(1994) News and the Culture of Lying. New York: Free
Press. CAN THE MEDIA BE SO LIBERAL?
==============================