This page, created by Matt Garfield, is part of an undergraduate assignment at Davidson College.
Research
|
Works Cited |
II. Literature Review
There is a significant amount of past research into the effects
of campaign endorsements, much of it coming in the form of case studies examining
particular
election cycles. Most of this research, however, focuses on how much endorsements
affect election outcomes, rather than how much they affect media coverage
levels. In that regard, this paper can break new ground by looking at the
relationship between endorsements and levels of media coverage.
David Myers concluded in a 1991 study that most newspapers are suffering
declining circulation numbers, and therefore the influence they exert on
politics has lessened. He found that a relatively small proportion of readers
pay close attention to the editorial pages, but those who do read them are
more likely to be high-profile community leaders who can carry the message
to the broader community (Pulliam 7, 2004).
Robert Erikson published a landmark study of the 1964
presidential election, finding that in 223 counties in the northern United
States, an
endorsement
from a local newspaper resulted in a gain of five percentage points on average
for Lyndon Johnson. However, if two newspapers in the same county endorsed
different candidates, Johnson lost votes to his opponent, Barry Goldwater
(Erikson 215, 1976). Erikson used multiple regression analysis to illustrate
a positive relationship between endorsements and voting decisions (Erikson
212, 1976). He concluded that newspaper endorsements exert influence to a “small
but significant” extent. “Newspaper readers are rarely allowed
to be selective because circulation monopolies generally provide newspapers
with captive audiences in their local market areas,” he wrote. Erikson
found that even people who did not read newspapers “may receive voting
cues from others who do (Erikson 223, 1976).”
Political scientist Tim Counts studied the role of
endorsements on presidential votes, taking the following variables into
account: 1) party
registration,
2) percentage of residents who own the home in which they reside, 3) population
per square mile, 4) population change and 5) editorial endorsements. He looked
specifically at the 1948 and 1960 elections in which most newspapers endorsed
the Republican candidates, who were each ultimately defeated by Democrats.
He concluded that home ownership and population changes play a significant
role in influencing voters’ decisions, while newspapers are not likely
to exert any influence (Counts 647, 1985).
Byron St. Dizier conducted a 1985 experiment that is
believed to be “the
first setting under which endorsements and party identifications were tested
solely against one another (St. Dizier 594, 1985).” He concluded that “newspaper
endorsements have more impact than political party affiliation when information
about a candidate is minimal (King 3, 1997).”
Robert Hurd and Michael Singletary focused on the 1980
presidential vote, finding that voters were not significantly influenced
by newspaper
endorsements
in that election. Instead, there was a stronger correlation to variables
such as “strength of partisanship, education and the number of articles
read about the campaign (Hurd and Singletary 337, 1984).”
Cecilie Gaziano conducted research on the effects of
newspaper ownership on endorsement choices. She argued in a 1989 study
that chain-owned
newspapers
tended to be “homogenous in endorsement patterns (Gaziano 844, 1989).” Contrary
to previous popular opinion, she found that owners of larger newspapers exerted
less editorial control than those at smaller newspapers. In examining the
1976 and 1980 presidential elections, Gaziano concluded that newspapers endorsing
Republican candidates tended to be in markets with “two newspapers
under the same ownership.” In 1984 and 1988, newspapers with no competition
were most likely to back the Republican nominee. Finally, in every year except
1980, newspapers endorsing the Democratic nominee were usually in “two-paper
or multi-paper markets (Gaziano 844, 1989).”
Kenneth Rystrom examined 51 California newspapers over a 10-year period and
found that chain-owned newspapers tended to be slightly more liberal in ideology
than their independently owned counterparts, but that endorsements in state-wide
races from both types did not significantly influence voting results (Rystrom
451, 1987).
Daniel Pulliam of Butler University examined Congressional
endorsements given by nine newspapers in Indiana, also seeking to determine
the role that newspaper
ownership plays in endorsement decisions. Pulliam found that locally owned
newspapers are more likely to endorse “based upon the ownership’s
perspective on the issue,” while corporately owned newspapers are more
likely to choose based on “tradition, rather than ideology.” He
found that some newspapers choose not to make endorsements at all, citing
a desire to maintain credibility and foster a healthy relationship in the
community (Pulliam 48, 2004).
Journalism professor Peter Clarke conducted a national study of press coverage
and editorials in Congressional races, and found that newspapers tend to
show much more affection toward incumbent candidates than challengers, on
both their editorial and news pages (King 2, 1997). Clarke, however, did
not include presidential races in the scope of his research.
A 1996 survey conducted by Editor & Publisher Magazine found that an
increasing number of newspapers are choosing not to make endorsements. The
study found that 69.9% of responding newspapers, or 415, adhered to a “no-endorsement” policy,
up from 67% in 1992. By comparison, an Editor & Publisher poll taken
in 1940 found that only 13.4% of newspapers remained neutral in presidential
races (Jones 1, 1996).
Studies also indicate that newspapers with larger circulations are more likely
to make endorsements than those with smaller circulations. A 2000 survey conducted
by Editor & Publisher found that 50% of newspapers with circulations less
than 20,000 said they would endorse candidates. The percentage rose to 80% for
newspapers with circulations greater than 20,000 (Pulliam 14, 2004).
Why are newspaper endorsements on the decline? Editor & Publisher put this
question to leading newspaper editors across the nation, and received a variety
of responses. Bill Kovach, curator of the Nieman Foundation for Journalism, said
that because of an increase in chain ownership nationwide, there are fewer U.S.
cities with more than one major daily newspaper. Because of this lack of competition,
newspapers see less sense in making political endorsements, said Kovach. “If
we are the only voice in town, maybe we shouldn’t be on the editorial pages,” he
said (Jones 1, 1996). Other editors, including Andy Stone, editor of the Aspen
(Colo.) Times, believe that taking sides in national elections does not serve
the interests of their readers. “As a local paper, we feel our area of
expertise, and where we have any credibility at all, is in the local arena, so
it’s important to limit our endorsements to that area,” said Stone.
Other editors say their newspapers remain neutral in order to avoid being viewed
as biased, thus alienating many readers. “We don’t want an outright
endorsement to affect how people read our paper,” said Charles Broadwell,
editor of the Fayetteville (N.C.) Observer (Jones 2, 1996).
Some editors disagree with this line of reasoning,
arguing that newspapers show a lack of courage when they avoid taking public
stands
on elections. “You
have to be a bold voice and let people know where you stand,” said Sig
Gissler, former editor of the Milwaukee Journal. Newspapers “give their
opinion all the time. What makes (the presidential election) so special?
Remaining vigorous in the community is to speak out on things that are important
(Jones
2,3, 1996).”
Contrary to popular opinion, Editor & Publisher
found that historically, newspapers that choose to make endorsements overwhelmingly
support Republican
candidates. However, the four terms served by President Franklin D. Roosevelt
illustrate that this trend has not always had a positive effect on voters.
Roosevelt never won the support of a majority of newspapers, and in 1944,
60% of newspapers
backed his opponent, Thomas Dewey. Similarly, John F. Kennedy won the presidency
in 1960 with endorsements only from 16.4% of newspapers, while opponent
Richard Nixon had 57.7% (Jones 1, 1996).
© Davidson College, 2004, Department of Political
Science, Davidson College, Davidson, NC 28035
Send comments, questions, and suggestions to Dr.
Patrick Sellers, Professor of Political Science.
Created: 4/29/2004. Last updated: 5/03/2004.