Intercity Competition and Local Election Coverage
INTERCITY COMPETITION AND LOCAL ELECTION COVERAGE:
A CASE STUDY
by
James B. White
Doctoral Candidate
Northwestern University
Department of Communication Studies
1881 Sheridan Road
Evanston, IL 60208-2236
Phone: (708) 491-5839
e-mail: [log in to unmask]
Prepared for presentation at the Association for Education in Journalism
and Mass Communication annual meeting, Washington, DC, August, 9-12,
1995
ABSTRACT
That competition in media markets enhances the political information available
to the
public is a widespread assumption. Yet empirical findings on
the effect of competition
on content are ambiguous. This paper reports findings from a case
study of "umbrella"
or intercity competition, and its effect on state legislative and
congressional
election coverage. The paper concludes that umbrella competition may
cause an increase
in local political coverage initially, but this coverage may
quickly taper off; and th
at papers may respond differently to umbrella competition.
INTRODUCTION
A widespread assumption of both communications scholars and policy makers is
that
vigorous competition among news media firms enhances the
political information
available to the public. Indeed, telecommunications deregulatory
measures are now
making their way through Congress premised in part on this
assumption. However, few
empirical studies of the effect of media market structure on news
content have been
conducted. Those that have provide conflicting findings, seldom
consider political news
content in any detail, and fail to take into account contemporary media market
structure.
This paper examines the impact of "umbrella" or intercity newspaper competition
on
local election coverage in a major metropolitan area. The
umbrella model holds that
newspaper competition is more likely to occur across rather than
within circulation
"layers." The paper will first review the current literature on
newspaper competition
and content, and discuss the umbrella model of newspaper market
structure. It will then
present findings from a case study of umbrella competition. The paper concludes
that
newspapers engaged in umbrella competition may increase coverage
of local politics
initially, but this attention may taper off as time goes on; and
that different size
papers may respond differently to umbrella competition.
LITERATURE REVIEW
Newspaper monopoly and democratic politics
The declining number of newspaper markets in which there exist competing dailies
has
been a concern of policy makers and students of political
communication for decades. It
is assumed that market concentration in the newspaper industry reduces the
amount of
political views and opinions available to the public.
As Lacy and Simon (1993) explain, the "underlying assumption is that the utility
a
community receives is positively correlated with the number of
ideas and bits of
information available. . . . If one newspaper closes, the supply will
decline because
the amount of information and opinion material produced by a
surviving newspaper is
less than that provided by two competitive newspapers" (p.
224-225). This assumption is
not only evident among social scientists and economists studying political
communication, but among communications policy makers as well. The
Newspaper
Preservation Act, which suspends antitrust laws to sustain independent
editorial voices
in some markets, is based on the assumption that two competitive newspapers
provides
readers with more and better quality news and information than a
single, monopoly
paper.
Despite the vigor with which policy makers and the judiciary have attempted to
preserve
newspaper competition on democratic grounds, the empirical evidence supporting
the
assumed relationship between competition and newspaper content
is mixed. The research
to date has found that competition between newspapers results in
more space for local
news, and more pictures and opinion columns (Rarick & Hartman,
1966). Competition also
seems to spark increased investment in a newspaper's editorial
component. Newspapers
facing competition were found to spend more money on wire
services and staff reporters,
have a larger news hole, and devote more space to color and photographs (Litman
&
Bridges, 1986). However, another study attributed the increased use
of wire service
copy in part to a decline in the number of newspapers (Donohue &
Glasser, 1978). And
newspaper competition was found to exert an influence on the
competitiveness of
election campaigns (Vermeer, 1995).
Other marketplace or "system" characteristics have been found to exert an
influence on
newspaper content. Donohue, Olien and Tichenor (1985) found
that increasing pluralism
or diversity in society was associated with an increase in the
amount of conflict r
eporting in Minnesota newspapers.
But other studies show less support for faith in a competitive marketplace.
Using data
on 91 papers, Entman (1989) found that "competition fails to
exert much influence" on
the diversity of opinion, seriousness, fairness and
responsiveness of news coverage (p.
96). Similarly, McCombs (1988) examined two newspapers under both competitive
and
monopoly situations and found few "significant content differences"
(p. 136). And
Weaver and Mullins (1975) found little difference between papers in
competitive situati
ons, suggesting that competing papers duplicate rather than
diversify content.
McCombs (1988) argues that the reason competing mainstream newspapers do not
look much
different than monopoly papers is that all news organizations
share identical
"professional values, beliefs, and practices" (p. 136). Therefore,
regardless of the
competitiveness of the market, the news product will look pretty
much the same across
publications.
Aside from the inconclusiveness of findings in this area, research on market
structure
and political news suffers from two other limitations. First,
few studies focus on
political news content in any detail. Rather, most studies,
including those reviewed
above, examine broad content categories such as local news, human
interest news,
sports, business and advertising. Other studies examine indirect
measures of content
such as the size of the news hole, the amount of staff versus
wire copy, and the
financial investment in editorial resources. Therefore, the bulk of
empirical evidence
accumulated to date has little direct relevance to questions
posed by democratic
theory. These questions include: What is the impact of market
structure on the supply
of competing opinions about political races and issues? How are
market characteristics
related to the amount of coverage afforded to different types of
political news, like
policy positions, the political horse race, scandal or candidate
personality? And how
does the relationship between political districts and media
market boundaries influence
coverage? These questions and no doubt many more directly relevant to concerns
about
democracy and information are still very much in need of
empirical investigation.
Part of the reason for this is a reluctance among scholars to define and measure
content "quality." Indeed, judging whether political information
produced by the media
is "better" in some market situations than in others is a risky
business. Entman and
Wildman (1992) identify two schools of thought in the debate over
content quality. The
"market economics" school, informed by traditional economic
theory, argues that quality
is best measured by consumer demand; that is, those media products that do best
in the
market are assumed to be best for the "marketplace of ideas." The "social
value"
school, on the other hand, contends that content quality is that
which serves a general
"public interest." Public interest for this school is not an aggregation of
individual
consumer demands, but rather, a community goal derived from deliberation among
citizens. Failure to reconcile these points of view has led many
scholars studying
media to avoid using content measures that explicitly require
normative justifications.
The second limitation of the literature is that few studies operationalize
competition
in a way that accurately reflects current newspaper market
structure. Much of the
research that has been done on competition and content compares
differences between pap
ers that are engaged in direct newspaper competition and those
that are not. Direct
competition refers to "face-to-face competition of newspapers of
general appeal located
in and identified with a single city" (Rosse, 1980). But most media scholars
agree t
hat the bulk of competition between newspaper firms now occurs
between papers not
located in the same city nor occupying similar circulation "layers."
The decline in direct newspaper competition has been well-documented (Lacy &
Simon,
1993; Bagdikian, 1992; Rosse, 1980; Rosse & Dertouzos, 1979).
But coinciding with this
decline has been an increase in the number of suburban and
satellite city dailies which
are taking advantage of population shifts from center cities to outlying areas.
This
trend has received less scholarly attention. In 1980, Rosse
concluded that since these
satellite city and suburban markets tend to be small, these
papers seldom encounter
direct competition. However, the increasing ability of large
papers to target specific
groups of readers through zoning has enabled them to compete in
these smaller markets.
This "umbrella model" of newspaper competition will be discussed
below.
Newspaper market structure and umbrella competition
Newspaper market concentration is usually not the result of illegal mergers and
unfair
market practices. Rather, the decline of newspaper competition
has its roots in the
economics of the industry itself. Most media markets, newspaper
markets included, are
characterized by economies of scale which are marked by declining
average costs; that
is, the average cost of producing each unit of a product declines
as more units are
produced. So, the more pages and/or copies of a newspaper that are
produced, the che
aper the cost of producing each individual page and/or copy
becomes. Economies of scale
encourage newspaper firms to increase circulation.
The result, as Owen (1975) concludes: "Large newspapers will tend to drive out
smaller
ones; two smaller newspapers can both gain by merger" (p. 17).
Economies of scale exist
in markets where one or a few firms can service the market more efficiently
than many
firms. But what is the relevant market for newspapers? As Owen
points out, if economies
of scale were the only factor determining the shape of newspaper markets,
theoretically we would expect one national (or even global) newspaper. This
clearly is
not the case.
Several factors counteract the expansive influence of scale economies. First of
all,
the demands of advertisers and readers are localized. Readers
demand local news and
advertisers demand local audiences who live and shop in the areas
in which they sell th
eir products. Also, attracting a larger readership requires a
higher quality product,
which requires an increased investment (Owen, 1975). While the
average cost of
producing the first copy of a newspaper may decline as circulation
increases, a higher
investment in that first copy may be necessary to attract
additional readers.
Thus, in the United States, newspaper markets tend to be limited geographically
to
areas that are meaningful to subscribers and advertisers. The
competing influences of
economies of scale on the one hand, and a demand by both readers
and advertisers for ge
ographically specific products on the other, explain the current
multi-layered
structure of American newspapers. The umbrella model of newspaper
competition
illustrates this phenomenon.
The United States newspaper industry is made up of several layers of papers
which
include large national publications like USA Today, the Wall
Street Journal and the
national edition of the New York Times; large regional papers,
like the Los Angeles
Times, Chicago Tribune, and Philadelphia Inquirer; medium-sized,
small city and
suburban dailies; and local weekly papers. This market structure has
been referred to
as "umbrella competition" (Rosse & Dertouzos, 1979). The model
gets its name from its
most common illustration, which resembles layers of umbrellas of
decreasing size.
Rosse and Dertouzos (1979) argued that due to the existence of scale economies,
newspapers are more likely to differentiate their products to
appeal to different
audiences rather than compete directly. As they explain, "survival
of newspaper firms
depends on each being able to whittle out a profitable audience
segment" (p. 444). Even
large regional or national papers can be thought of as serving "local"
communities.
The only difference between the New York Times national edition
and a small, satellite
city daily newspaper "is the size of the local community in which
each happens to be
located" (p. 445).
Rosse's model suggests that, despite a decline in direct newspaper
competition--that
is, competition within layers--newspapers may compete with
those in other layers. One
way competition can occur between layers is through "zoning."
Most metropolitan dailies
with suburban or satellite city circulation print different editions of their
paper,
or special sections, with news specific to those areas. Zoning is
an attempt by a
single paper to satisfy a segmented and geographically dispersed
audience while still
taking advantage of economies of scale.
A few empirical studies have investigated the "umbrella model" theory. In a
study of
competition among Southern California dailies, Tillinghast
(1988) found that intercity
competition led to shifts in circulation and advertising. Lacy
and Sohn (1990) tested
the assumption of substitutability by comparing
circulation-content correlations of
metropolitan dailies and suburban non-dailies. They found weak
support for this
hypothesis. Finally, Lacy (1988) found that suburban newspapers
competing with larger
metropolitan dailies tend to respond by increasing their news
hole as well as local
news coverage.
The umbrella model suggests that those testing content effects by examining the
few
remaining cases of direct newspaper competition may be looking
in the wrong place. To
examine the effect of umbrella competition on content, one has to
examine the content
that reaches readers where the circulation zones of newspapers
occupying different
circulation layers overlap. This often occurs between zoned
editions of central-city
papers and their smaller, satellite-city competitors. Recent
developments in the suburb
an Chicago newspaper market provide a unique opportunity to
examine political news
content before and after the onset of umbrella competition between
a major metropolitan
and a satellite city daily newspaper.
DATA, METHODS AND HYPOTHESES
How would we expect the umbrella model of competition to affect the coverage of
political news? Umbrella competition occurs when larger newspapers
"invade" the
circulation zones of smaller newspapers. To attract readers from these
smaller markets,
larger papers must increase the amount of coverage of that local area. We would
expect
this increase to include coverage of local politics, since, as mentioned above,
newspaper markets for circulation are defined by "local community
interests" (Rosse and
Dertouzos, 1979, p. 444).
It is also important to point out that we would expect this increase at least to
some
degree by both papers involved in competition. Since both are
vying for the same local
audience we should expect a similar emphasis in their coverage of
local news. In
economic parlance, the products should be substitutable. It is not
clear what degree of
similarity among media products is necessary for them to be considered good
substitutes. Unlike other products, which are assumed to be
differentiated on the basis
of price, media products are differentiated on the basis of content. In cases
of
umbrella competition, larger papers moving into a smaller paper's
market must attract
readers "by being a substitute for the local paper, while
differentiating itself with
more news about areas outside the local market" (Lacy, 1988, p.
400). Just what mix of
local and non-local news is necessary most likely varies from
market to market.
Amount of coverage is only one way newspapers appeal to reader interest. The
prominence
with which news stories are played is also an indication of their importance.
We might
expect, then, for newspapers involved in umbrella competition to increase the
pro
minence of local news coverage through story placement (front or
inside page), and the
use of photographs and graphics to call attention to stories.
Three hypotheses concerning the relationship between umbrella competition and
the
coverage of local politics can be drawn from the preceding
discussion. These are:
H1: Umbrella competition will result in more coverage of local politics.
H2: This increase in coverage will be similar across all papers involved in the
competition.
And finally,
H3: Umbrella competition will result in more prominent coverage of local
politics.
In January 1991 the Chicago Tribune opened its Northwest bureau, a regional news
office
covering some of the city's most rapidly growing and affluent suburbs. News
produced
by the bureau appears primarily in the Tribune's "Metro
Northwest" section, one of the
seven zoned metro sections of the paper. The circulation zone of
the Northwest
Edition--which in 1994 had a daily circulation of 83,840--overlaps the
circulation
region of the Arlington Heights Daily Herald, a satellite-city
daily newspaper. The
Daily Herald's 1994 circulation was 134,495. However, in the
circulation area that
overlapped with the Tribune's Northwest Edition--suburban Cook
County, Illinois--the
Daily Herald had just over 70,000 subscribers.
This study focuses on coverage of state legislative and congressional elections.
Election coverage was chosen for both practical and conceptual
reasons. First, election
coverage occurs in a relatively condensed period of time, making data
collection
somewhat easier than considering political news during non-election
periods. Second,
the campaign season is one of the periods when the need for
political information is
highest. Thus, it is an ideal period to study the adequacy of
political news coverage.
Data was collected from four general election years--1988, 1990, 1992, and 1994.
This
provided two years of data preceding the introduction of the
Tribune's bureau, and two
years after. It also ensures that both the pre-competition and
competition periods
included a presidential election, which has been found to affect
the amount of coverage
of some smaller district elections (Kahn, 1991).
From one month prior to each November general election through the day after the
election, all stories on Illinois State Senate, State House, and
congressional
districts falling within the area of competition between the two
newspapers were coded.[1]
These different types of political districts were chosen to see how, if at all,
umbrella competition impacts differently on coverage of districts of
varying size.
Previous studies of news coverage of elections has found that
district size and fit
with media markets influences the amount of news coverage
districts receive (Goldenberg
and Traugott, 1987). Also, differences in coverage of various districts provide
a test
of the second hypothesis. A similar response to competition by both newspapers
would
mean similar attention to the different districts considered
here.
Two categories of data were recorded. "Coverage amount" was measured as the
number of
paragraphs on districts of interest for each election period.
"Coverage prominence" was
measured by recording the various types of design and graphic elements
associated with
stories.
FINDINGS
Figures 1 and 2 in the appendix show support for the first hypothesis, that
umbrella
competition will result in increased coverage of local
politics. Figure 1 shows the
number of paragraphs appearing in the Chicago Tribune about each
type of district in
the various electoral periods. Remarkably, only 1 paragraph of
text was written about a
Northwest suburban race in 1988. The coverage was afforded to a congressional
race. No
stories contained information about suburban legislative races that year. The
amount
of coverage increases slightly in 1990. After the introduction of
the Tribune's
Northwest bureau, coverage of all districts more than doubles.
Coverage is less in 1994
than in the preceding election period, but is still higher than in either
pre-competition period.
Figure 2 shows Daily Herald coverage. It shows a similar pattern to Figure 1,
but with
much larger frequencies in all categories of coverage. The
Daily Herald responded to
the presence of the Tribune bureau by dramatically increasing
coverage of local
elections, state legislative races in particular. Like Tribune
coverage, Daily Herald
coverage declines in 1994, but is still higher than coverage in
either pre-competition
period.
The second hypothesis--that both papers involved in umbrella competition will
show
similar increases in coverage--is less well-supported. As can be
seen from Figures 1
and 2, the papers did not provide similar amounts of coverage of
local news, and they
emphasized different types of districts.
Tables 1 and 2 show the results from a crosstabulation of political district
with
election year, for each newspaper. A chi-square statistic was
calculated to test for
the significance of the relationship between the two
nominal-level variables.[2] The
tables show the same frequencies depicted in bar chart form in
Figures 1 and 2. They
also show column percentages, which more clearly reveal which
political districts were
given more coverage by each newspaper.
Both the Tribune and Daily Herald afforded more than half of their coverage to
state
House of Representative races in 1990. Much of this coverage
concerned an unusually
competitive race for a single House seat. The Daily Herald
continued to emphasize state
legislative races after the introduction of the Tribune bureau. The Tribune,
however,
devoted most of its 1992 coverage to congressional races, and its
1994 coverage to
State Senate and congressional races. State House races became less
and less a part of
Tribune coverage.
Mixed support also was found for the third hypothesis, that umbrella competition
will
result in more prominent coverage of local politics. An
analysis of various graphic and
design techniques used to give more prominent play to stories suggests that,
although
running a smaller amount of coverage, the Tribune was more likely
after competition to
increase coverage prominence. The effect of competition on Daily
Herald election
coverage prominence is less clear.
Tables 3-10 show crosstabulations of several design characteristics with
election year.
Data were collected on the following design characteristics: the use of
photographs
with stories; the use of graphics, most often district maps or
eye-catching "logos"
that identified the story with the paper's continuing election
coverage; story
location, on the front page, inside the front section, on the front of
the second
section, or inside the second section[3]; and whether the story
"led" a section front. [4]
Clear changes in prominence are evident in Tribune coverage after
introduction of the
Northwest bureau along all dimensions. Tribune coverage after
competition began was
more likely to be accompanied by a photograph or graphic, and more
likely to be placed
on a section front or lead the page of a section front. Daily
Herald crosstabulations
do not show patterns that suggest content effects of competition.
While more coverage
appeared after introduction of the Tribune bureau, that coverage
was less likely to be
accompanied by a photograph or graphic than in 1988. Coverage
location and the
likelihood of coverage to lead a section front also show no clear
relationship to
competition.
The data on prominence also are relevant to the second hypothesis. They suggest
that
the Tribune responded to competition more by increasing the
prominence of coverage,
than by increasing the amount. The Daily Herald, on the other
hand, increased the
amount of coverage dramatically, but did not substantially increase
the prominence of
this coverage.
Determinants of election coverage
The nominal-level analysis above suggests that umbrella competition does indeed
effect
local election coverage. But this analysis does not control
for a number of electoral
factors which also could be responsible for an increase in
coverage. For instance,
previous studies have found that the competitiveness of election
races and the number
of open seats also impact news coverage (Goldenberg & Traugott,
1987). Nor does the
prior analysis identify the independent effect of newspaper
competition on the amount
of coverage afforded to political races.
To address these questions, the number of paragraphs written on each political
race was
regressed on competition and a number of control variables. The unit of
analysis was
the political race. Competition was operationalized as a dummy
variable, with 1 rep
resenting an election year occuring after the introduction of the
Tribune bureau, and 0
representing an election year before. Control variables included the
competitiveness
of the race, operationalized as the percentage of the vote going
to the winner; a dummy
variable indicating whether the race was for an open seat; and whether the race
was
for a state house, state senate or congressional seat.[5]
Tables 11 and 12 in the Appendix show the results of the regression analyses.
For the
Daily Herald, umbrella competition is shown to have a
statistically significant impact
on coverage after other factors are controlled for. As the
regression coefficient for
competition indicates, stories appearing after introduction of
the Tribune bureau
tended to receive about 33 more paragraphs of coverage than stories
in the
pre-competition era. And according to its beta, competition was the second
strongest
variable in the equation, behind electoral competitiveness. All
variables showed an
impact in the expected direction. Since competitiveness of race
was operationalized as
the winner's percentage of the total vote, it is inversely
related to the amount of
coverage. The regression equation explained 54 percent of the
variance in Daily Herald
coverage.
Umbrella competition and the competitiveness of races also exerted a significant
impact
on Tribune coverage. But this impact was less statistically significant and
less
powerful than it was for the Daily Herald. In fact, the
regression equation for the
Tribune provides a much less adequate explanation of coverage, with
an R-square of .20.
Clearly, other factors not accounted for in the equation were responsible for
much of
the Tribune coverage. Another surprise from the Tribune analysis
is the direction of
the effect of open elections. It was expected that open seats
would receive more
coverage than seats with incumbents, a tendency that has been
demonstrated in numerous
studies of election coverage. For the Tribune, however, races for
open seats tended to
receive somewhat less coverage than seats with incumbents. One
can only speculate as to
why this was the case. Perhaps, races between political newcomers only enhance
the
amount of news coverage in major regional papers like the Tribune
when the political
jurisdiction being contested is large. For smaller local
districts, the ignonimity of
novice candidates may further decrease, rather than enhance, the
race's attractiveness
to readers of large, regional papers.
DISCUSSION
This analysis suggests several tentative conclusions and areas for further
investigation. First, this study suggests that umbrella competition does
indeed affect
political news content. It appears that newspapers engaged in
umbrella competition will
respond, at least initially, by increasing the amount of coverage they afford
to local
political news. However, both papers studied here reduced their coverage of
local
politics after an initial high point which immediately followed the
onset of
competition.
Newspapers, then, may enhance local coverage as an immediate response to
competition,
but eventually return political coverage to pre-competitive
levels. An analysis that
considers more election periods than those studied here as well
as responses to compe
tition in different markets is needed to determine whether
content increases due to
competition are sustained, or quickly trail off.
Second, this study suggests that while newspapers involved in umbrella
competition may
increase the overall amount of local political coverage, they
may emphasize different
types of political districts. It has been demonstrated elsewhere
that news media or
ganizations serving large markets tend to pay more attention to
large political
districts that are congruent with their market boundaries (Campbell et
al, 1984;
Goldenberg and Traugott, 1987; Stewart and Reynolds, 1990). Also, it
has been suggested
that large newspapers "invading" a smaller paper's market must provide enough
local
coverage to be considered a good substitute. Tribune coverage
after introduction of its
bureau can be explained as an attempt to satisfy both of these demands.
By emphasizing congressional and Illinois State Senate districts, the largest of
the
three types of districts considered here, the Tribune may have
been able to better take
advantage of economies of scale while at the same time increase its appeal to a
local
audience. Despite their size, congressional and State Senate
districts still might be
considered matters of local interest to Northwest suburban
readers. However, all of the
congressional districts and some of the Senate districts fell only partially
within
the overlapping circulation zones of the two papers. It is
reasonable to assume that
the same congressional and State Senate stories appearing in the
Northwest Edition may
have been used in other Tribune editions which overlapped
portions of these districts.
A more extensive analysis, which examines content for all zoned
editions of major
metropolitan dailies like the Tribune, is needed to see exactly how
larger papers
balance the demand for local news in zoned regions with the
pressures to attract large
audiences.
Finally, this study suggests that design and graphic elements also are used as a
response to umbrella competition. Combining local election
coverage with attractive
visual elements as well as packaging coverage to emphasize its
importance are ways that
newspapers can appeal to local readers. Although, as was the case with the
amount of
coverage, it appears that papers may use prominence as a
strategic response to
competition to varying degrees. In this study, the large metropolitan
paper, while
showing less of an increase in the amount of local election
coverage, was more likely
than its smaller competitor to increase the prominence of
coverage after competition
began. Unfortunately, the reasons for this difference are not
apparent from this study.
Perhaps limitations on space available for zoned coverage would force a
center-city
daily to rely more on prominence than amount of coverage to
attract local readers.
In general, this case study provides evidence that readers who live in markets
where
umbrella competition takes place have available to them more,
and more prominent news
about local political elections, at least for a time. Assuming
that more is better, t
his bodes well for the relationship between the free market and
democratic politics.
However, any such conclusion would be premature before more
extensive measures of
political news content can be examined across numerous media
markets.
APPENDIX Table 1: Crosstabulation of amount of Tribune coverage on
congressional and State Legislative races by year. Cells show the number
of paragraphs with column percentages in parentheses.
1991
Tribune Bureau Introduced
District Type
1988
1990
1992
1994
Congressional
1 (100)
13 (22)
167 (53)
51 (34)
State Senate
0
13 (22)
72 (23)
81 (53)
State House
0
34 (56)
77 (24)
20 (13)
Total
1 (100)
60 (100)
316 (100)
152 (100)
Pearson Chi-square: 82.96 p < .01
Table 2: Crosstabulation of amount of Daily Herald coverage on
congressional and State Legislative races by year. Cells show the number
of paragraphs with column percentages in parentheses.
1991
Tribune Bureau Introduced
District Type
1988
1990
1992
1994
Congressional
83 (28)
34 (12)
326 (17)
155 (25)
State Senate
56 (19)
56 (20)
679 (34)
224 (36)
State House
157 (53)
186 (68)
964 (49)
240 (39)
Total
296 (100)
276 (100)
1969 (100)
619 (100)
Pearson Chi-square: 105.41 p < .01
Tables 3-10: Crosstabulation of Tribune and Daily Herald design
characteristics by year. Cells show the amount of coverage (in
paragraphs) accompanied by each design element. Column percentages are
in parentheses.
Story accompanied by photograph
3. Tribune
1991
Tribune Bureau Introduced
Photograph
1988
1990
1992
1994
No
1 (100)
60 (100)
240 (76)
55 (36)
Yes
0
0
76 (24)
97 (64)
Total
1 (100)
60 (100)
316 (100)
152 (100)
Pearson Chi-square: 107.25 p < .01
4. Daily Herald
1991
Tribune Bureau Introduced
Photograph
1988
1990
1992
1994
No
64 (22)
184 (67)
1252 (64)
438 (52)
Yes
232 (78)
92 (33)
717 (36)
398 (48)
Total
296 (100)
276 (100)
1969 (100)
836 (100)
Pearson Chi-square: 204.02 p < .01
Story accompanied by graphic
5. Tribune
1991
Tribune Bureau Introduced
Graphic
1988
1990
1992
1994
No
1 (100)
29 (48)
79 (25)
47 (31)
Yes
0
31 (52)
237 (75)
105 (69)
Total
1 (100)
60 (100)
316 (100)
152 (100)
Pearson Chi-square: 15.85 p < .01
6. Daily Herald
1991
Tribune Bureau Introduced
Graphic
1988
1990
1992
1994
No
64 (22)
179 (65)
342 (17)
281 (34)
Yes
232 (78)
97 (35)
1627 (83)
555 (66)
Total
296 (100)
276 (100)
1969 (100)
836 (100)
Pearson Chi-square: 323.60 p < .01
Location of story
7. Tribune
1991
Tribune Bureau Introduced
Location
1988
1990
1992
1994
Front Page
0
0
0
33 (22)
Inside Front Section
0
13 (22)
0
0
Front of 2nd Section
0
7 (12)
178 (56)
70 (46)
Inside 2nd Section
1 (100)
40 (67)
138 (44)
49 (32)
Total
1 (100)
60 (100)
316 (100)
152 (100)
Pearson Chi-square: 217.71 p < .01
8. Daily Herald
1991
Tribune Bureau Introduced
Location
1988
1990
1992
1994
Front Page
0
32 (12)
240 (12)
107 (13)
Inside Front Section
128 (43)
75 (27)
317 (16)
404 (48)
Front of 2nd Section
37 (13)
69 (25)
806 (41)
311 (37)
Inside 2nd Section
131 (44)
100 (36)
606 (31)
14 (2)
Total
296 (100)
276 (100)
1969 (100)
836 (100)
Pearson Chi-square: 621.36 p < .01
Lead story on section front
9. Tribune
1991
Tribune Bureau Introduced
Lead story
1988
1990
1992
1994
No
1 (100)
60 (100)
267 (84.5)
127 (84)
Yes
0
0
49 (15.5)
25 (16)
Total
1 (100)
60 (100)
316 (100)
152 (100)
Pearson Chi-square: 11.29 p < .05
10. Daily Herald
1991
Tribune Bureau Introduced
Lead story
1988
1990
1992
1994
No
284 (96)
227 (82)
1880 (95.5)
594 (71)
Yes
12 (4)
49 (18)
89 (4.5)
242 (29)
Total
296 (100)
276 (100)
1969 (100)
836 (100)
Pearson Chi-square: 368 p < .01
Table 11: Determinants of campaign coverage for the Daily Herald.
Independent Variables
b
S.E. of b
Beta
Umbrella Competition
33.29**
7.28
.38
Competitiveness of Race
- 136.87**
22.74
-.49
Open Seat
15.57
9.96
.13
Congressional Race
12.55
10.24
.10
State Senate Race
5.53
8.30
.05
R-square: .54
Adjusted R-square: .51
N = 82
*p < .05 **p < .001
Table 12: Determinants of campaign coverage for the Chicago Tribune.
Independent Variables
b
S.E. of b
Beta
Umbrella Competition
10.94*
4.44
.27
Competitiveness of Race
- 36.03*
13.73
-.28
Open Seat
- 8.98
6.07
-.17
Congressional Race
10.20
6.22
.18
State Senate Race
2.55
5.04
.05
R-square: .20
Adjusted R-square: .15
N = 82
*p < .05 **p < .001
REFERENCES
Entman, Robert M. (1989). Democracy without citizens: Media and the decay of
American
politics. New York: Oxford University Press.
Entman, Robert M. & Wildman, Steven S. (1992). Reconciling economic and
non-economic
perspectives on media policy: Transcending the "marketplace of
ideas." Journal of
Communication, 42(1).
Bagkikian, Ben H. (1992). The media monopoly (4th ed.). Boston: Beacon Press.
Campbell, James E., Alford, John R., & Henry, Keith (1984). Television markets
and
Congressional elections. Legislative Studies Quarterly, 4, 665-678.
Donohue, Thomas R. & Glasser, Theodore L. (1978). Homogeneity in coverage of
Connecticut newspapers. Journalism Quarterly, 55, 592-596.
Donohue, George A., Olien, Clarice N. & Tichenor, Phillip J. (1985). Reporting
conflict
by pluralism, newspaper type and ownership. Journalism Quarterly, 62, 489-499,
507.
Goldenberg, Edie N. & Traugott, Michael W. (1987). Mass media in U.S.
Congressional
elections. Legislative Studies Quarterly, 3, 317-339.
Kahn, Kim Fridkin (1991). Senate elections in the News: Examining campaign
coverage.
Legislative Studies Quarterly, 3, 349-374.
Lacy, Stephen and Simon, Todd F. (1993). The economics and regulation of United
States
newspapers. Norwood, NJ: Ablex Publishing.
Lacy, Stephen and Sohn, Ardyth B. (1990). Correlations of newspaper content with
circulation in the suburbs: A case study. Journalism Quarterly, 67,
785-793.
Lacy, Stephen (1988). The impact of intercity competition on daily newspaper
content.
Journalism Quarterly, 65, 399-406.
Litman, Barry R. and Bridges, Janet (1986). An economic analysis of daily
newspaper
performance. Newspaper Research Journal, Spring, 9-26.
McCombs, Maxwell E. (1988). Concentration, monopoly, and content. In Picard,
R.G.,
Winter, J.P., McCombs, M.E., & Lacy, S. (Eds.), Press concentration
and monopoly: New
perspectives on newspaper ownership and operation (pp.
129-137). Norwood, NJ: Ablex
Publishing.
Owen, Bruce M. (1975). Economics and freedom of expression. Cambridge, MA:
Ballinger.
Rarick, Galen & Hartman, Barrie (1966). The effects of competition on one daily
newspaper's content. Journalism Quarterly, 43, 459-463.
Rosse, James N. (1980). The decline of direct newspaper competition. Journal of
Communication, Spring.
Rosse, James N. & Dertouzos, James N. (1979). The evolution of one newspaper
cities. In
Symposium on media concentration. Washington, DC: Federal Trade Commission.
Stewart III, Charles & Reynolds, Mark (1990). Television markets and U.S. Senate
elections. Legislative Studies Quarterly, 4, 495-523.
Tillinghast, Diana Stover (1988). Limits of competition. In Picard, R.G.,
Winter, J.P.,
McCombs, M.E., & Lacy, S. (Eds.), Press concentration and monopoly: New
perspectives
on newspaper ownership and operation (pp. 71-87). Norwood, NJ:
Ablex Publishing.
Vermeer, Jan P. (1995). Multiple newspapers and electoral competition: A
county-level
analysis. Journalism Quarterly, 72, 1, 98-105.
Weaver, David H. & Mullins, L.E. (1975). Content and format characteristics of
competing daily newspapers. Journalism Quarterly, 52, 257-264.
[1] Any story mentioning either candidate or the specific district was
coded. If a
story included information on more than o
ne race, only that portion of the story
explicitly dea
ling with the race or races of interest was coded. Most of the state
legislative districts considered here fell completely with
in the overlapping
circulation zones of the two newspa
pers. State Senate districts, which are twice as
large
as State House districts, were more likely to be partially included in t
he area
of competition. No Congressional district fell
completely within the circulation zone
of both papers
. The Congressional districts used in the analysis, however, did have
large portions of their jurisdictions falling within the
area of competition.
[2] Chi-square provides a test of the independen
ce of nominal-level variables. It
assumes that if the
variables are independent--which is the null hypothesis--cases will
be
evenly distributed throughout the cells of the table. If the actual distr
ibution is
significantly different from this expected distribution, the
n the null hypothesis of
independence is rejected and
it is assumed the two variables are related.
[3] Second section refers
to the Tribune's "Metro Northwest" section after
compe
tition; its "Chicagoland" section before competition; and the Daily Heral
d's
"Neighbors" section during both periods. These sec
tions contained much of the local
news for both papers
.
[4] The lead story was determined by headline size and placement, ei
ther across the
top of a section front page or in the
"drop-right" position, which means occupying the
two r
ight most columns of the page. If no story in one of these positions had
a
dominant headline, no lead story was chosen for that
page.
[5] Type of race was entered as a set of dummy variables, with s
tate house race
serving as the base variable.
|
