CALL-IN MEDIA EXPOSURE AND STRENGTH OF CANDIDATE PREFERENCE:
AN ANALYSIS OF THE 1992 NES PRE-/POST-ELECTION SURVEY
Erik P. Bucy
Doctoral Student
College of Journalism
University of Maryland, College Park
2806 Clear Shot Drive #12
Silver Spring, MD 20906-6207
e-mail: [log in to unmask]
tel/fax: (301) 598-4351
Paper submitted to the Communication Theory and Methodology Division of
the
Association for Education in Journalism and Mass Communication
for presentation at the 1995 Annual Convention
in Washington, D.C.
Running Head: Call-In Media
August 9, 1995
CALL-IN MEDIA EXPOSURE AND STRENGTH OF CANDIDATE PREFERENCE:
AN ANALYSIS OF THE 1992 NES PRE-/POST-ELECTION SURVEY
ABSTRACT
Using the 1992 NES Pre-/Post-Election survey, this paper explores
the relationship between exposure to the "new media" (e.g. talk
radio and call-in television) and strength of candidate
preference.
Of voters reporting exposure to call-in media, a majority state
a
strong preference for a candidate over a weak preference.
Refuting
the conventional wisdom that talk shows are mainly the province
of
White, working-class conservatives, the analysis shows
middle-class
minorities and liberals to report the highest levels of talk
show
use. Exposure to call-in media, as measured by the crude NES
survey
question, does not directly predict strength of candidate
preference, however. To better get at the effects of the "new
media," more sophisticated measures are called for. The paper also
provides a demographic profile of call-in media users.
Introduction
Students of elections note that it is the innovations in political
campaigns, whether related to the adaptation of a new technology or an
innovative run by a third party candidate, that make elections worth
our
studying (Chaffee & Jamieson, 1994). "Predictable as the general
election-year routine might be, each campaign brings surprises"
(1994:261). Faced with a mainstream media environment that has become
increasingly negative, antagonistic, self-important, and cynical in
recent years (Dickson, Chapman, & Gershon, 1994; Patterson, 1993),
candidates in 1992 sought alternative channels for communicating with
voters (Lemert, 1994). The launching of Ross Perot's third-party
candidacy, for instance, was built largely around televised call-in
media appearances. In a study of the rise and fall of Perot's
candidacy,
Zaller and Hunt (1994) estimate that Perot persuaded roughly 20
percent
of those watching him on morning interview programs to support his
candidacy. Ridout (1993) attributes Bill Clinton's comeback in the weeks
leading up to the Democratic National Convention to a flurry of
appearances on nationally televised interview programs and talk
shows--ten in one month's time.
Though initially associated with small audience cable systems, "new
media" formats (including electronic town hall forums, political
call-in television programs, presidential debates that feature audience
questions, and morning variety and afternoon talk shows that invite
candidates to address questions from a live studio audience) were
quickly embraced by the networks. During the '92 election all of the
major television networks adopted some variant of the political talk
show or town hall format (Ridout, 1993). In addition to these
institutional adaptations, the new media may have been influential in
altering audience members perceptions of their political
participation
and knowledge (Bucy, 1995; Hollander, 1994). In a study using the
1992
pre-election North Carolina Poll, Hollander (1994) found that the
greater the use of talk shows, even when controlling for demographic and
political factors, the greater the perception of campaign knowledge.
Theoretically, call-in shows facilitated the introduction of a two-way
communication model into presidential politics, enabling at least
some
voters to speak directly to candidates in a dynamic described by
Newhagen (1994) as the "spontaneous interaction between political
figures and citizens." This direct interaction between voters and elites
through communication technology, a sort of political interactivity,
underscores the prominence and evolving nature of media in electoral
politics. Interactivity, or "the extent to which users can participate
in modifying the form and content of a mediated environment in real
time" (Steuer, 1995:46), is a focal concept of the new media. The
two-way flow that interactivity implies suggests a new avenue of
participation for voters and presents a new opportunity for candidates
during elections.
For political actors, call-in formats hold appeal because they make
relatively unfiltered communication possible with the electorate. For
voters, call-in formats facilitate direct questioning of politicians.
From both points of view, there is a relative lack of mediation and
diminished role for the journalist. This lack of mediation is an
important factor in distinguishing call-in programs from traditional
campaign news or advertising (Newhagen, 1994; Zaller & Hunt, 1994).
Although a relatively small number of voters directly interacted with
candidates in these new media programs, an argument can be made that
the
few ordinary citizens who did get through helped to redefine the
nature
of political messages broadcast through the mass media. Callers who
speak directly to candidates may serve as surrogates to members of the
viewing audience; home television viewers may personally identify
with
the callers who get through to Larry King or see reflections of
themselves in the typical electronic town hall forum participant.
Given the perception that the political system may now be more
accessible and responsive through the mass media, a change in the
reception of political messages may therefore be afoot; the challenge is
to capture this change and measure what new impact, if any,
interactive
formats have on voters. The present study investigates an emerging
area
of candidate evaluation, namely, how voters' preferences toward
presidential candidates are affected by exposure to the new media
formats.
The Decline of Mediating Institutions
Patterson (1993) argues that the McGovern-Fraser reforms of 1970
favoring primaries over caucuses in nominating elections left a
mediation vacuum between voters and the political system that gave the
press a more prominent role in selecting candidates. The rise of
direct-vote primaries thus diminished the influence of the political
parties as mediating institutions between the political system and the
electorate. From a more technological determinist perspective,
Ranney
(1983) maintains that the shift to television as the dominant medium
of
political communication may itself be a primary explanation for the
decline in the salience of parties. Through professional norms and
practices that value coverage of individuals rather than institutions,
broadcast journalism is structured to inform voters more about
prominent
and charismatic political personalities than political parties
(Chaffee,
Zhao, & Leshner, 1994).
In the absence of strong party associations, communication variables
assume a more prominent role in voter decisions (Chaffee & Choe,
1980).
The large number of voters who eventually voted for Independent
candidate Perot during the 1992 general election (19 million) provides
at least anecdotal evidence that traditional party ties were
relatively
weak and, consequently, voters were open to other sources of
direction
and information about their campaign choices, namely, from the
press.
Patterson (1980) has dubbed this campaign process the "mass media
election." In 1992, the rise of the new media began to erode even
journalists' roles as mediators (Lemert, 1994).
From their analysis of the Perot campaign, Zaller and Hunt (1994) found
the Texas billionaire was effective when he could appeal to voters
in an
essentially unmediated fashion, that is, with little or no
journalistic
interpretation between the message and the intended audience.
Unmediated
communication is defined by Zaller and Hunt as communication over
which
the source of a message retains control of (a) the timing of its
release, (b) the amount released, and (c) the informational context in
which it appears (1994:360). Besides call-in shows that allow for
viewer
questions and comment, other forms of unmediated political
communication
include paid political advertising that goes unchallenged either by
the
opposition or the news media, and direct mail (unless, of course, a
direct mail piece becomes the object of a controversy and subsequent
news coverage, is challenged and opened to reinterpretation).
Contesting the popular view that talk shows and call-in programs
altered the course of electoral politics in 1992, Zaller and Hunt (1994)
found little evidence that the new media became an important
long-term
factor in even Perot's campaign. Rather, they argue, "the larger and
longer-lasting effects of the interview and call-in programs...was to
signal the conventional press that Perot was a candidate who merited
serious attention" (1994:363). Just as the tabloid press drove the
Genifer Flowers philandering story onto the front pages of establishment
newspapers, so call-in programs could be said to have surveillance
and
agenda-setting effects on the mainstream media, serving as an early
warning system for up-and-coming news items and alerting the national p
ress what not to ignore. The effect of call-in shows on the
campaign,
then, seems to have been indirect rather than direct.
Despite the modest direct effects of the new media on public opinion,
Perot, as a third party candidate with no party apparatus in place
before he started, rose to national prominence largely because of his
radio and television interview and call-in appearances, including the
30-minute infomercials he aired on national TV to promote his
economic
proposals and campaign effort (Chaffee, Zhao, & Leshner, 1994).
After
sliding in the polls at the same time Perot was enjoying his ascent,
Bill Clinton embraced the new media in earnest in June 1992, appearing
on Larry King Live (twice), Arsenio Hall, MTV, morning variety
shows,
and two nationally televised town hall meetings. Whether due to these
relatively unmediated programs or not, Clinton dramatically improved
his
poll ratings, moving from third place at the beginning of June to
first
place by the beginning of July (Ridout, 1993). A few weeks after he
won
the presidency, Clinton said of the traditional press filter:
"Anyone
who lets himself be interpreted to the American people through these
intermediaries alone is nuts" (Golson & Range, 1992).
Data and Methods
Social science has long had difficulty proving whether mass media has
observable effects on the audience outside of controlled laboratory
settings (Iyengar & Kinder, 1987). As Bartels notes, the scholarly
literature on this subject has been much better at "refuting,
qualifying, and circumscribing the thesis of media impact than at
supporting it" (1993:267). Research more often than not concludes that
mass media in general, and television news in particular, simply
strengthen or reinforce existing beliefs and opinions (Kinder & Sears,
1985). Given the pervasiveness of the mass media and their virtual
monopoly over election information, the inability to prove a causal
connection between media messages and voter behavior is, according to
Bartels, "one of the most notable embarrassments of modern social
science" (1993:267). Indeed, with regard to political communication,
Iyengar and Kinder argue that the lack of a (universal) theory of
media
effects significantly impedes our understanding of how democracy
works
(1987).
This paper does not promise to prove what decades of political
communication research has not been able to accomplish; however, it does
endeavor to lay some groundwork for future investigations into the
newly
emergent political call-in programs popularized during the last
presidential election. As mentioned, the interactive nature of the new
media is central to our understanding of it and should,
consequently,
figure prominently into our research approach. If media effects are
difficult to discern from a simple one-way model of communication (with
a unidirectional information flow), then perhaps the added element
of
two-way interactive communication intrinsic to call-in formats may
produce different results.
Although still few in number, studies that look at the relationship
between new media programs and voters have tended to focus on call-in
media use and political knowledge, a traditional dependent variable.
Looking at data from two pre-election statewide surveys, one from
California, the other from North Carolina, Chaffee, Zhao, and Leshner
(1994) found that talk shows did, in fact, add to the public's
knowledge
about candidates and their issue positions. Hollander (1994)
analyzed
the same North Carolina survey data and compared it to the combined
1992
National Election Study (NES) Pre-/Post-Election Survey. In the NES
data, call-in show exposure is weakly but positively associated with
knowledge about the candidates and their parties. Call-in exposure
surf
aces as a slightly stronger predictor of knowledge in the North
Carolina
data (Hollander, 1994).
From the national data, Hollander also finds that call-in exposure
leads to greater campaign knowledge among the less educated, but
provides no additional knowledge for more educated respondents. Talk
show exposure does not influence such perceptions among more educated
respondents, perhaps because their exposure to televised political
information is offset by the in-depth analyses available in the print
media, which more literate voters would be apt to use.
Missing from most studies of call-in media written thus far is an
examination of the impact of call-in media exposure on candidate
preference. This paper uses the above 1992 NES Pre-/Post-Election Survey
conducted by the Inter-university Consortium for Political and
Social
Research at the University of Michigan to explore the relationship
between call-in exposure and strength of candidate preference. Both
call-in exposure and strength of candidate preference are simple
dichotomous, or nominal, variables (yes-no; strong-not strong). This
weakness in survey design necessarily limits the analysis by providing
a
narrow range of potential answers and constraining the explanatory
power
of the variables. Moreover, the call-in question measures
respondents'
exposure to call-in media rather than attention, which is seen as a
stronger, more reliable predictor of media effects (Chaffee &
Schleuder,
1986; Geiger & Newhagen, 1993; McLeod & McDonald, 1985). The
call-in
question asks: Do you listen to or watch talk shows on radio or TV
where
people call in to voice their opinions? (yes, no) The candidate
preference question queries:
How strong is your preference for the presidential candidate you will
vote for? (strong, not strong)
While the call-in question is not specific to a particular medium
(either television or radio) or in the precise program, these problems
are also found in questions measuring general television and
newspaper
exposure and attention (Hollander, 1994). And, despite the
limitations
of questions that translate into nominal variables, their analysis
can
still serve as a starting point for investigating potential talk
show
effects. Based on the above, it can be predicted that:
H1 Of those exposed to call-in media, more will report a strong
preference for their presidential candidate of choice than a weak
preference; call-in users will be more likely to vote for their
candidate than non-users.
Race has been found to predict certain political behaviors and
orientations. For instance, Shingles (1981) found that working-class
Blacks attend campaign rallies and display lawn signs and bumper
stickers more than working-class Whites. A 1992 election study of the
African-American enclave of Prince George's County, Maryland by
Newhagen
(1994) found that Blacks reported using call-in shows more than
Whites.
Newhagen (1994) also found, as did Zerbinos (1993), talk show media
users to be more likely to vote, to be issue-oriented, and to undertake
a number of other political behaviors. Newhagen's Prince George's
County
study found self-efficacy and race to be particularly strong
indicators
of call-in use. Social class, a traditionally strong indicator of
news
media use (Robinson, 1975), did not do well as a predictor of
call-in
use in the Prince George's County study, nor did ideology, political
(system) efficacy, or leader trust.
Taking race into consideration with call-in media exposure, it can be
predicted that:
H2 Minorities will report greater exposure to call-in media programs
than Whites, regardless of class.
Talk shows, and talk radio in particular, are generally considered to
be conservative in orientation (Times Mirror Center, 1993). As such,
they may attract an ideologically conservative audience. Political
commentator Fred Barnes (1994) recently advanced a connection between
talk radio and working-class conservatives (white males):
One of the things that I think is not to be underestimated in this
past [1994] campaign and as an ongoing phenomenon in reaching
this
particular group of people, is talk radio, and particularly
people
like Rush Limbaugh and others, who I think have a tremendous
impact on that group of people. These are people who drive
trucks,
they're out in their cars and trucks, they're working. They
listen
to these shows on television, these talk radio [sic] shows
that
tend to be more conservative than not, and they're affected by
them. And some people like Rush Limbaugh see their role as an
educating role, one where he will read some article that is on
the
editorial page of the Wall Street Journal to people who would
otherwise not be reading that page.
Given Barnes' anecdotal observations about talk shows and periodic
pieces in the popular press suggesting talk shows are a conservative
phenomenon (e.g. Corliss, 1995; Fineman, 1993; Kolbert, 1993; Lind,
1995; Tueque, 1993;), it can be predicted that:
H3 Conservative respondents will report higher levels of call-in media
exposure than liberal respondents; working-class conservatives will
report the highest levels of call-in exposure.
Perot's association with the call-in format gives rise to a final
hypothesis. Given the Independent candidate's reliance on unmediated
communication channels and programs to reach his constituency, it can
be
predicted that:
H4 Perot supporters will report higher levels of call-in media exposure
and they will report higher levels of strong preferences than
Clinton or
Bush supporters.
Results
A frequency run of the bivariate call-in question showed more
respondents reporting exposure to call-in media (yes, n=1,264 or 50.8
percent) than reported not seeing or listening to such programs (no,
n=1,044 or 42.0 percent). One hundred seventy seven respondents (7.1
percent) did not answer the question and were removed from further
analysis. A smaller number responded to the strength of candidate
preference question than did the call-in item (1,813 compared to 2,308).
Of these, more reported a strong preference (1,249 or 50.3 percent)
than
a preference that was not strong (564 or 22.7 percent). That 672
respondents, or 27 percent of the total survey population, did not
answer the preference question perhaps reveals the limitation of a
simple strong/not strong question design more than their own lack of an
opinion.
A crosstab analysis of media habits, demographic traits, and political
characteristics from the combined 1992 NES Pre-/Post-Election Survey
data shown in Table 1 shows that those exposed to call-in media are
heavy television viewers (70.6 percent, p<.001) and newspaper readers
(51.1 percent, p<.05) but only light-to-moderate radio listeners
(p<.01). The call-in audience perhaps associates the new media more with
political call-in television than radio talk shows. Half of all
call-in
viewers report having at least some college education while the
other
half report a high school education--or less. Eighty-two percent of
the
call-in audience was white (p<.001); 18 percent minority (a combined
category of Blacks, American Indians, and Asians, p<.001).
Career-wise, more call-in viewers are employed (61.4 percent, p<.05)
than are out of work or going to school (38.6 percent, p<.05). While
they tend to have jobs, members of the call-in audience are neither
educationally nor financially "rich." It is the lower-to-middle rungs
of
the socioeconomic ladder, those whose real wages have been in slow
decline in recent years (Phillips, 1990), who are the most avid call-in
media users. Only 10.5 percent of the call-in audience reported
having
done some graduate study and 16.2 percent reported a combined family
income of $60,000 or over (although income was not statistically
significant).
Politically, a roughly equal number of call-in viewers described
themselves as Democrats, 38 percent, as Republicans, 37 percent, while a
fewer number, 25 percent, identify themselves as Independent
(p<.05).
A Statistical Profile of Call-In Media Users[1]
(respondents reporting "yes" to the exposure question; n=1,264)
Table 1
Media Habits Light Users Moderate Users Heavy Users
Radio 44.1%** 33.3%** 22.6%**
Newspapers 32.4%* 16.6%* 51.1%*
Television 12.2%*** 17.2%*** 70.6%***
Demographic Traits Male Female
Gender 46.8% 53.2%
Total Male Female
Average Age 46.2 44.4 47.8
White Minority
Race 82.0%*** 18.0%***
White M Minority M White F Minority F
Race by Gender 84.0%* 16.0%* 80.4%*** 19.6%***
K-12 College Graduate Study
Education 50.0%* 39.5%* 10.5%*
Employed Unemployed
Job Status 61.4%* 38.6%*
$0-$25k $25k-$60k $60k and over
Income Level 44.1% 39.7% 16.2%
Political Variables Clinton Bush Perot
Candidate 53.9%_ 36.8%_ 9.3%_
Liberal Moderate Conservative
Ideology 26.9% 32.2% 41.0%
Democrat Independent Republican
Party ID 38.0%* 24.9%* 37.0%*
Strong Not Strong
Preference 70.3%* 29.7%*
Yes No
Perot Ever 1st Choice? 22.8%_ 77.2%_
_p<.10 *p<.05 **p<.01 ***p<.001
Though not significant for call-in exposure, the income variable was
significant in a crosstab of strength of candidate preference by
income.
In the crosstab, a greater percentage of lower and middle-income
respondents reported a strong preference for a candidate than upper
income voters (p<.001). With more money and leisure time to spend
thinking about politics (and potentially with more need, as they may
have financial interests at stake), upper income voters have more
opportunity and motivation to follow the election and develop a
political schema for sorting out the barrage of conflicting information
that inundates the public sphere during a campaign. This political
interest may then result in the development of more sophisticated
cognitions that enable them to distinguish between the different shades
of gray in political messages. Hence, their preferences are less
pronounced because they are able to see the good and bad in each
candidate.
Strength of preference demonstrated an inverse relationship to
education in another crosstab, however: the less education the stronger
the candidate preference, and the more education the weaker the
candidate preference (p<.001). Frequently, education is used as a proxy
for political sophistication (Sniderman, Brody, & Tetlock, 1991).
From
this, we can expect lower education voters neither to have the
requisite
interest in politics (made possible by leisure time and motivation)
nor
the cognitive tools necessary to develop a highly elaborated
political
schema, or framework for organizing their political beliefs. As
research
in political cognition has shown, college educated people place more
emphasis on performance criteria, while the less well-educated tend to
base their inferences on the observable attributes of candidates
(Miller, Wattenberg, & Malanchuk, 1986). They consequently will tend to
view candidates in superficial terms. In short, their political
cognitions will resemble those of low income voters.
A crosstab using strength of preference by call-in exposure yielded a
significant association and provided support for Hypothesis 1. Of
those
who reported exposure to call-in media, a higher percentage stated a
strong preference for a candidate than a weak preference, 70.3 percent
compared to 29.7 percent (p<.05). However, the results for those who
did
not report using call-in media were similar, though slightly less
pronounced; 65.5 percent held strong preferences compared to 34.5
percent who did not (p<.05). While this finding supports the first
hypothesis, it is important to test whether it still holds true after
other forms of media use, as well as important demographic and
political
factors, are taken into consideration.
Next, a crosstab using call-in media exposure by race was run,
revealing significant results that showed minorities to have higher
levels of reported call-in media exposure than Whites. Overall, 65.1
percent of minorities reported call-in exposure compared to 52.9
percent
for Whites (p<.001). This finding supports the first part of
Hypothesis
2 and corroborates Newhagen's (1994) finding that race can be a
predictor of call-in use. The same crosstab of call-in exposure by race,
this time controlling for income, showed minorities reporting higher
levels of call-in exposure than Whites across all income levels
(p<.05,
except for upper income), supporting the second part of the
hypothesis,
which said that minorities will report greater exposure to call-ins
than
Whites regardless of class.
Call-In Exposure by Race
(controlling for income level)
Table 2
Income
Hypothesis 2
Lower Middle Upper
Race (n=520) (n=468) (n=192)
White 55.3%* 52.8%*** 53.2%
Minority 65.4 %* 71.7%*** 58.1%
*p<.05 **p<.01 ***p<.001
Hypothesis 3 is addressed to the question of whether talk shows are
primarily a conservative, working-class phenomenon. A crosstab of
call-in exposure by ideology shows that slightly more self-described
liberals than conservatives (a combined category that includes
moderates
and conservatives) report watching or listening to call-in media,
although not at statistically significant levels. The same crosstab,
this time controlling for income, reveals a similar pattern and runs
contrary to conventional wisdom. At the lower- to middle-income
levels,
liberals report slightly more call-in exposure than conservatives
(the
results approach significance at the p<.10 level). Call-in use
becomes
the domain of conservatives only in the upper-income range
($60,000+),
which is rarely mentioned as the natural audience of talk shows.
These f
indings fail to support Hypothesis 3, which said that conservatives
will
report higher levels of call-in media exposure than liberals, and
that
working-class conservatives will report the highest levels of
call-in
exposure.
Call-In Exposure by Ideology
(controlling for income level)
Table 3
Income
Hypothesis 3
Lower Middle Upper
Ideology (n=523) (n=470) (n=192)
Liberal 63.4%_ 56.3% 43.4%*
Conservative 56.1 %_ 54.3% 56.4%*
_p<.10 *p<.05 **p<.01 ***p<.001
The first part of Hypothesis 4, that Perot supporters will report
higher levels of call-in exposure than Clinton or Bush supporters, was
born out by two crosstabs. The first question asked if Perot was
ever
your first choice for president and the second asked who voters would
be
voting for in the presidential election. The results of both
crosstabs,
using call-in as the dependent variable, approached significance
(p<.10). In the first instance, 58.2 percent of Perot-as-first-choice
voters reported call-in exposure, while 41.8 percent did not. In the
second crosstab, Perot voters reported the highest amounts of call-in
viewing of any other candidate following: 66.2 percent for Perot
compared to 56.3 percent for Clinton and 56.4 percent for Bush
supporters respectively. Only a third of Perot voters said they had not
watched or listened to any call-ins. The second part of the research
question, that Perot supporters would report stronger preferences than
Clinton or Bush supporters, was also born out by a second set of
crosstabs using strength of preference as the dependent variable. A
higher percentage of Perot supporters, 75.9 percent, reported a strong
preference than Clinton, 71.8 percent, or Bush, 62.9 percent,
supporters
(p<.001).
Table 4 presents the bivariate relationships (correlation coefficients)
for strength of preference and call-in exposure and a series of
demographic, political, and media variables. For strength of candidate
preference, education level emerges as the strongest correlate (.12,
p<.001) in the data. Radio use (.11, p<.01) and income (.09, p<.001)
are
the next most associated with a strong candidate preference, while
call-in exposure emerges with a very slight association (.05, p<.05).
The variable, "Was Perot ever your first choice for President," shows
the strongest negative correlation to strength of candidate
preference
(-.15, p<.001), followed by party identification (-.13, p<.001) and
television use (-.10, p<.001). From this we can infer that voters with
strong preferences were less inclined to be Perot supporters,
television
viewers, or diehard party faithful.
For call-in exposure, radio use shows the strongest association (.12,
p<.001), suggesting that some media use leads to more media use and
supporting the intuitive supposition that radio listeners are bound to
regularly encounter at least some talk radio programs.
Interestingly,
television use has the strongest negative correlation to call-in
exposure (-.22, p<.001). This counter-intuitive finding suggests that
the more people watch television, the less they will be exposed to
call-in programs. Perhaps this latter
contradictory finding is an artifact of a self-reporting measure.
Television viewers may equate call-in programs with talk radio and not
televised political call-ins such as Larry
King Live or the electronic town hall forums popularized during the 1992
campaign, so
Correlation of Strength of Preference and Call-In Exposure
with Demographic, Political, and Media Variables
Table 4
Variable Strength of Preference Call-In Exposure
Age -.09*** -.01
Education .12*** .01
Employment -.07** -.05*
Income .09*** .02
Race -.01 -.06**
Gender .00 .00
Candidate -.08*** -.04
Ideology .02 -.01
Party -.13*** -.04
Perot -.15*** .04
Newspaper .01 -.05*
Radio .11** .12***
Television -.10*** -.22***
Call-In .05* (Preference) .05*
*p<.05 **p<.01 ***p<.001
they may not report being exposed to them when asked in a survey.
Newspaper use and race were also both negatively correlated to call-in
exposure, but only slightly.
Since strength of preference is a dichotomous, nominal variable,
logistic regression was used to further assess the relationship between
the variables and the respondent's report of strength of preference.
Table 5 shows the variables Perot as 1st Choice and Republican party
membership to have the largest regression coefficients relative to
standard error. The negative sign on Perot as 1st Choice (-.68, p<.001)
indicates that respondents who are not Perot supporters are more
likely
to have strong preferences than those for whom Perot was ever their
first choice for president. While the crosstabs supporting the first
part of Hypothesis 4 showed substantial call-in use by Perot
supporters,
this new media exposure does not necessarily translate into a strong
preference. The call-in variable shows a small regression coefficient
relative to standard error and it is not statistically significant.
Compared to Democrats (the regression reference category for the
political party variable), Republicans are more likely to have a strong
preference for a candidate (.47, p<.001). For Independents, the
opposite
holds true (although at a level only approaching significance); the
odds
are they will have a decreased chance of a strong preference
compared to
Democrats (-.21, p<,10). As is increasingly the case, political
belief
systems may be organized around factors other than party affiliation
and
ideology and may be reconstituted with each election based on the
unique
mix of candidates, issues, and context specific to the campaign
(Kennamer, 1987). People for whom party identification does not operate
as a primary filter and cognitive organizer may be more receptive to
alternate information sources, including news reports, debates, and
political advertising (Chaffee, 1981). Following this logic, one might
expect Independents to be more media reliant for their candidate im
pression formations.
The main findings from this logistic regression showing Perot
supporters and Independents to not have strong candidate preferences
possibly derives from the negative tenor of much media coverage,
especially toward Perot (Patterson, 1993; Zaller & Hunt, 1995). What
Zaller and Hunt term the "bust" phase of Perot's popularity cycle was
more pronounced than for that of the other candidates (1995). Both
the
sheer amount of Perot's negative coverage, and the amount of it
relative
to positive coverage, were greater than for comparable former
candidates
Jimmy Carter and Gary Hart (Zaller & Hunt, 1995). Exposed to
coverage
that was largely antagonistic toward Perot, Independents were
understandably lukewarm partisans.[2]
Logistic Regression of Demographic, Political, and Media
Variables Against Strength of Preference[3]
Table 5
Variable Regression Coefficient S. E. Coefficient/S.E.
(n=1,670)
Age -.09* .04 -2.25
Employment -.15 .14 -1.07
Income
$0-$24,999 -.19* .08 -2.38
$25,000-$59,999 -.04 .07 -0.57
Race -.18 .16 -1.13
Gender -.10 .13 -0.77
(n=1,563)
Candidate .34* .15 2.27
Ideology -.33* .15 -2.20
Perot 1st Choice -.68*** .13 5.23
Party
Independent -.21_ .11 -1.91
Republican .47*** .08 5.88
Education
Elementary -.56** .20 -2.80
High School -.11 .10 -1.10
Some College .15 .11 1.36
College Degree .11 .13 0.85
(n=603)
Call-In -.10 .20 -0.50
Radio -.50** .18 -2.78
Newspaper
None -.06 .19 -0.32
1 Day -.48_ .25 -1.92
2 or More Days .51** .18 2.83
Television
1-2 Days .11 ,17 0.65
3-4 Days -.04 .17 -0.24
_p<.10 *p<.05 **p<.01 ***p<.001
Discussion and Conclusion
From this analysis we see that call-in media exposure is nominally
correlated with a strong preference for a candidate but recedes from
view when other variables are included in a regression model.
Politically, the talk show or call-in audience is made up of a roughly
equal number of Democrats and Republicans and a fewer number of
Independents. Call-in users are not uneducated; half of all respondents
who reported yes to the call-in question had at least some college
education. Interestingly, minorities report higher levels of call-in
exposure than Whites, with middle income minorities reporting the
highest levels. Consistent with other findings (e.g. Newhagen, 1994), a
conservative ideological orientation does not explain call-in use;
slightly more liberals report talk show exposure than conservatives. In
general, talk shows do not appear to be a conservative,
working-class,
white male phenomenon, despite the conventional wisdom (and
occasional
scholarly assertions, e.g. Pan & Kosicki, 1995).
Additionally, it is increasingly evident that call-in media, which
occupied a unique place in the presidential campaign of 1992, are
becoming a regular feature of the national political landscape. Call-in
exposure (especially as measured by the crude NES question item) did
not
directly predict strength of candidate preference; however, as a
salient
media genre during the 1992 election, talk shows may have
contributed at
different times to both Bill Clinton's and Ross Perot's rise in the
polls. In the case of Perot, call-ins appear to have contributed
substantially to his initial rush of support--his burst onto the
national scene (Zaller & Hunt, 1994). This radar blip then prompted the
conventional media to give Perot's campaign more attention and
closer
scrutiny.
Perot won the first small increment of his support through largely
unmediated communication. The press, impressed by this
evidence of
his mass appeal, judged Perot to be a credible candidate for
president and hence worthy of sustained coverage, which it
promptly
began to provide. In so doing, the press effectively
interposed
itself between Perot and the mass audience. This coverage
greatly
extended the reach of Perot's message, but it also had its
downside. Thereafter, the flow of positive and negative news about
Perot in the conventional press, as determined by reporters'
rules
of coverage, would determine the fate of Perot's candidacy
(Zaller
& Hunt, 1994:386).
Just as conventional media emerged as an important long-term factor in
Perot's continued support, so may another intervening variable (or
host
of intervening variables) determine the relationship between call-in
exposure and strength of candidate preference. As this study and a
great
body of other research has found, political party plays an important
role in candidate preferences. Indeed, several classic studies have
found political party (and ideology) to be so predictive of voter
preference in general elections that a definitive test of other
influences is rendered difficult, if not untenable (Berelson,
Lazarsfeld, & McPhee, 1954; Burnham, 1970; Campbell, Converse, Miller, &
Stokes, 1966). Issues such as the economy, taxes, and defense, which
this paper did not consider, may also determine candidate choices.
Nevertheless, call-in media do generate some observable direct effects.
Perhaps the most obvious of these is their ability to elicit direct
feedback from the audience in such forms as the billions of dollars in
annual purchases (and thousands of yearly on-air phone calls) to the
Home Shopping Channel or the millions of dollars in pledges to public
radio and television stations during their fundraising telethons.
During
the campaign of 1992 these "effects" were most evident in the volume
of
phone calls and offers of volunteer and financial support to the
innovative campaigns of both Ross Perot and Democratic primary candidate
Jerry Brown, who pioneered the use of the direct political appeal on
both mainstream (network) and cable broadcast media.[4] If the effects
of
unmediated political communication on the audience are ever to be
better isolated, more
sophisticated measures of call-in activity, exposure, and attention will need
to be
developed.[5]
References
Barnes, F. (1994). Remarks on "The Diane Rehm Show," WAMU-FM,
Washington, DC, December 5.
Bartels, L. M. (1993). Messages received: the political impact of media
exposure. American Political Science Review, 87(2), 267-284.
Berelson, B., Lazarsfeld, P. L., & McPhee, W. N. (1954). Voting: A study
of opinion formation in a presidential campaign. Chicago: University
of
Chicago Press.
Bucy, E. P. (1995, May). Where interactivity resides: An agenda for 'new
media' research in political campaigns. Paper presented at the
annual
meeting of the International Communication Association,
Albuquerque, NM.
Burnham, W. D. (1970). Critical elections and the mainsprings of
American politics. New York: Norton.
Campbell, A., Converse, P. Miller, W., & Stokes, D. (1966). Elections
and the political order. New York: Wiley.
Chaffee, S. (1981). Mass media in political campaigns: An expanding
role. In R. E. Rice and W. J. Paisley (Eds.) Public Communication
Campaigns. Beverly Hills: Sage.
Chaffee, S., & Choe, S. (1980). Time of decision and media use during
the Ford-Carter campaign. Public Opinion Quarterly, 44, 53-69.
Chaffee, S., & Schleuder, J. (1986). Measurement and effects of
attention to news media. Human Communication Research, 13, 76-107.
Chaffee, S, & Jamieson, K. H. (1994). Studies of the 1992 U.S. election
campaign: An overview. Communication Research, 21(3), 261-263.
Chaffee, S., Zhao, X., & Leshner, G. (1994). Political knowledge and the
campaign media of 1992. Communication Research, 21(3), 305-324.
Corliss, R. (1995). Look who's talking. Time, January 23.
Dickson, S., Chapman, B., & Gershon, P. (1994, August). A case in
cynicism: Network coverage of the 1992 presidential campaign. Paper
presented at the annual meeting of the Association for Education
in
Journalism and Mass Communication, Atlanta, GA.
Fineman, H. (1993). The power of talk. Newsweek, February 8.
Geiger, S., & Newhagen, J. (1993). Revealing the black box: Information
processing and media effects. Journal of Communication, 43(4),
42-50.
Golson, B., & Range, P. Interview with president Clinton. TV Guide.
November 21, 1992.
Hollander, B. A. (1994, August). Talk shows and political knowledge in
the 1992 presidential campaign. Paper presented at the annual
meeting of
the Association for Education in Journalism and Mass
Communication,
Atlanta, GA.
Iyengar, S., & Kinder, D. (1987). News that matters. Chicago: University
of Chicago Press.
Kennamer, J. D. (1987, August). Political party, candidate perceptions,
and media use as predictors of candidate preference. Paper presented
at
the annual meeting of the Association for Education in
Journalism and
Mass Communication, San Antonio, TX.
Kinder, D. R., & Sears, D. O. (1985). Public opinion and political
behavior. Vol. 2, in G. Lindzey and E. Aronson (Eds.) Handbook of social
psychology, 3rd ed. New York: Random House.
Kolbert, E. (1993). The people are heard, at least those who call talk
radio. The New York Times, January 29, A-12.
Lemert, J. B. (1994). Adapting to Clinton and the new media reality.
Media Studies Journal, 8(2), 53-58.
Lind, M. (1995). The south, the GOP, and America: The southern coup. The
New Republic, June 19.
Lodge, M., & Hamill, R. (1986). A partisan schema for political
information processing. American Political Science Review, 80(2),
505-519.
McLeod, J. M., & McDonald, D. G. (1985). Beyond simple exposure: Media
orientations and their impact on political processes. Communication
Research, 12(1), 3-33.
Miller, A. H., Wattenberg, M. P., & Malanchuk, O. (1986). Schematic
assessments of presidential candidates. American Political Science
Review, 80(2), 521-540.
Newhagen, J. E. (1994). Self-efficacy and call-in political television
show use. Communication Research, 21(3), 366-379.
Pan, Z., & Kosicki, G. M. (1995, May). Talk show exposure as an opinion
activity. Paper presented at the annual meeting of the International
Communication Association, Albuquerque, NM.
Patterson, T. E. (1980). The mass media election. New York: Praeger.
Patterson, T. E. (1993). Out of Order. New York: Knopf.
Phillips, K. (1990). The politics of rich and poor: Wealth and the
American electorate in the Reagan aftermath. New York: Random House.
Ranney, A. (1983). Channels of power: The impact of television on
American politics. New York: Basic Books.
Ridout, C. F. (1993). News coverage and talk shows in the 1992
presidential campaign. PS: Political Science & Politics, 26(4), 712-716.
Robinson, M. (1975). American political legitimacy in an era of
electronic journalism: Reflections on the evening news. In D. Carter and
R. Alder (Eds.), Television as a social force: New approaches
to TV
criticism. New York: Praeger.
Shingles, R. (1981). Black consciousness and political participation:
The missing link. American Political Science Review, 75, 76-91.
Sniderman, P. M., Brody, R. A., & Tetlock, P. E. (1991). Reasoning and
choice: Explorations in political psychology. New York: Cambridge
University Press.
Steuer, J. (1995). Defining virtual reality: Dimensions determining
telepresence. In F. Biocca and M. R. Levy (Eds.), Communication in the
age of virtual reality (pp. 33-56). Hillsdale, NJ: Lawrence
Erlbaum
Associates.
Times Mirror Center for the People and the Press (1993). News Release --
The vocal minority in American politics. Washington, DC.
Tueque, B. (1993). Press "1" for the Christian right. Newsweek, February
8.
Zaller, J., & Hunt, M. (1994). The rise and fall of candidate Perot:
Unmediated versus mediated politics--part I. Political Communication,
11
(4), 357-390.
Zaller, J., & Hunt, M. (1995). The rise and fall of candidate Perot: The
outsider versus the political system--part II. Political
Communication,
12(1), 97-123.
Zerbinos, E. (1993, August). Talk radio: Motivation or titillation?
Paper presented at the annual meeting of the Association for Education
in Journalism and Mass Communication, Kansas City, MO.
Endnotes
[1] The total n for those answering "yes" to the call-in exposure
que
stion is 1,264, but not all call-in viewers answered every question
so
the n varies in crosstab comparisons. Precise variable definitions
are
available from the author.
[2] This finding conflicts with the earlier reported crosst
ab showing a
higher percentage of Perot supporters reporting a stronger
preference
than Clinton or Bush supporters. However, a separate crosst
ab of
strength of preference by political party showed more Democrats w
ith
strong preferences than either Independents or Republicans (p<.001)
.
Additionally, a crosstab of strength of preference by Perot as 1st
Choice indicated that respondents who did not consider Perot their first
choice were more likely to have strong preferences, 72 percent, than
those who did, 55 percent (p<.001).
[3] Variables were entered in three
separate regressions to preserve
maximum number of cases due to missing
data. Categorical variables were
converted into a group of n-1 variabl
es, using the SPSS default
deviation contrasts subcommand, which uses t
he last category (typically
the highest level of any given variable) as
the reference category. The
value of the coefficient for the reference
category can be calculated by
taking the negative of the sum of the ot
her category coefficients.
Logistic regression coefficients are interpr
eted as a ratio of the
coefficient to the standard error; the larger th
e coefficient relative
to standard effort, the more likely are the odds
of an event occurring.
[4] Brown ultimately raised $6.5 million in pledges of $100 or
less and
received 180,000 calls to his initially derided then widely i
mitated 800
number (Brown for President campaign press release, July 19
92). The
author served as Brown's national scheduler and deputy press s
ecretary
during the '92 primaries.
[5] Unlike the NES survey, a New Yo
rk Times-CBS poll cited in Zaller and
Hunt (1994) framed call-in use as
a continuous variable rather than in
yes-no nominal terms, yielding ri
cher data.
|
