were found to support the Democrat FDR (Rosten 1937). Further studies since then
have confirmed the pro-Democrat tendency in political preferences among journalists
(Wilhoit et al. 1985, Beyle et al. 1996). However, many of those studies have focused
on Washington, or more generally, the East coast media elite, which may have resulted
in biased measures of social and educational background and overestimation of liberal
tendencies (Shoemaker and Reese 1996). Conservative media watchdogs like the Media
Research Center (
) place much emphasis on collecting
survey evidence about the political views of the media. The definition of ‘liberal’ being
contentious and in itself highly politicised in the U.S., such collections of journalists’
views appear selective and often seem to focus deliberately on issues like abortion and
affirmative action. In similar vein, political pundits like Ann Coulter (2003) and Bill
O’Reilly, and self-confessed ‘whistle blowers’ like Bernard Goldberg (2002, 2003)
build their case against the ‘liberal media’ to a large extent around an almost caricature-
like image of journalists and editors:
Goldberg’s definition of liberalism is a conservative activist’s wet dream of hot button and
wedge issues: abortion, affirmative action, feminism, homosexuality, and taxes. Liberals
may think of themselves as striving for educational opportunities, good wages and benefits,
progressive taxation, environmental conservation, civil liberties, and corporate
accountability. Those nuances, and in fact many of those concerns, are lost in Goldberg’s
narrative. (Harmon 2005: 111)
Conservative media watchdogs, right-wing pundits, and Republican campaigners are
among the most outspoken proponents of the liberal media bias claim. Since the 1980s,
they have seized upon these repeated survey findings about the predominance of liberal
political opinions within the journalistic profession. As Domke et al. (1999: 45ff.) report
for three recent presidential elections (1988, 1992 and 1996), complaints from political
elites and campaigners about slant in the media have become increasingly common
(rising by about 120% over those three campaigns). Also, they found that claims about
liberal bias completely dominated, constituting between 92% and 96% of all bias claims
during the campaigns. In similar vein, the news media themselves, when reporting about
bias, tend to treat media bias synonymous with liberal bias. In a Lexis-Nexis search,
Niven (2002) found over 600 newspaper articles, published during twelve months in
2000 and 2001, that discussed the topic of media bias. 81% of them talked about liberal
bias, only 5% about a conservative bias (3ff.). Harmon (2005) corroborated that finding.
Also using Lexis-Nexis, he found that since 1990
‘At ratios of typically about four to one, the term “liberal media” popped up much more
often than the term “conservative media.”’ (111f.)
And it can be shown that public debates and complaints about liberal bias occur not as
responses to suddenly increasing negative treatment of conservative ideas or Republican
4
Convention