Rosse's Model Revisited: Moving from Linearity to Concentric Circles to Explain
Janet A. Bridges and Barry Litman
with Lamar W. Bridges
Janet A. Bridges
Associate Professor of Communciation
University of Southwestern Louisiana
Box 43650 107 Burke Hall
Lafayette LA 70504-3650
Professor of Telecommunication
Michigan State University
Lamar W. Bridges
Professor of Journalism
Texas A&M Univesity at Commerce
Submitted to the Media Management and Economics Division, AEJMC, Chicago, 1997
Rosse's Model Revisited: Moving from Linearity to Concentric Circles to Explain
Janet A. Bridges and Barry Litman
with Lamar W. Bridges
Competition in the newspaper industry is no longer explained by the linear
umbrella model of competition proposed by Rosse in the 1970s.
Changes in the newspaper industry suggest a more fluid model of concentric
circles is appropriate. The proposed model retains the four Rosse layers,
incorporates a fifth, and illustrates changing conditions in the newspaper
industry that make suburban and satellite dailes more competitive.
Submitted to the Media Management and Economics Division, AEJMC, Chicago, 1997
Rosse revisited, page
This paper suggests a modification of the common interpretation of Rosse's
umbrella model of newspaper competition from the generally demonstrated linear
composition to one of concentric circles.
Predicated on the assumptions that newspapers compete in two markets, (1)
for circulation and (2) for advertising dollars, and that segmentation within
the local market is crucial to a newspaper's survival, Rosse's seminal analysis
of the newspaper industry resulted in the umbrella model of newspaper
competition (Rosse, 1975; Rosse & Dertouzos,1978). Using an economic model,
Rosse & Dertouzos (1978) explained that the newspaper (or product) is developed
for broadest appeal in its local market. The model suggests that in a given
geographic region competition among newspapers exists among four layers that
linearly subordinate both different-sized communities and the newspaper
organizations within these communities.
Based on circulation penetration data from the San Francisco and the New
York City areas, Rosse suggested that at the highest level, three layers of
newspapers are subordinated to an overreaching "umbrella" layer of a
central-city metropolitan daily that broadens its coverage and circulation, in
effect becoming a regional or even a state paper. The densest circulation of the
"umbrella" newspaper is in the central city itself. In each succeeding layer
the newspaper is expected to serve a decreasing geographical area; the interests
of the target audience (the circulation base) are assumed to become more local
or narrow; and the potential advertising also becomes more restricted.
Rosse suggested that the layer #1 daily is attractive to national and
regional advertisers and therefore has a high quality content because of its
economic status, but that the layer #1 daily doesn't cover local news beyond the
metropolitan area very well.
The second layer of the umbrella is what Rosse labeled "satellite" city
newspapers, which have a dense circulation in their respective core cities but
which also circulate, although less densely, in a regional area abutting the
satellite community itself. The circulation of each satellite-city newspaper is
discrete from circulation of dailies within other satellite cities. Using the
umbrella terminology, each satellite city newspaper is shaded by the central
city daily (ies) and in turn shades a third layer of daily newspapers, the
The suburban city newspapers of the third layer share a characteristic of
the satellite dailies in that their circulation areas do not overlap, but they
operate in a much more restricted circulation zone. Rosse showed no competition
within the layers, only between the layers. Under the Rosse model, the suburban
newspaper is competing with its relevant satellite newspaper and its relevant
central city newspaper, but there is little or no overlap in circulation with
other suburban newspapers, hence the linear model.
A fourth layer of newspapers includes nondailies, noncontrolled circulation
newspapers, shoppers without news, and other formats. Many of these serve
communities without a suburban daily newspaper, but others operate within the
broader layers of the satellite and/or central city newspapers. Unique to the
fourth layer is special-interest content that identifies these newspapers
Owen (1975) and Picard (1989) demonstrated Rosse's umbrella model
-------------- Figure 1 about here -------------
The graphic model implies that the fourth layer is also exclusive within
the third layer, which is appropriate for those publications that focus on a
geographic area. But, as Rosse and Dertouzos (1978) said, special-interest and
weekly newspapers may serve overlapping geographic areas or may circulate within
larger communities already served by metropolitan or satellite newspapers.
Rosse & Dertouzos (1978) illustrated the umbrella model using circulation
penetration in the San Francisco Bay area, where two dailies competed in the
central city layer. Their figures showed a decline in metropolitan newspaper
penetration in the larger city (San Francisco), while penetration of smaller
dailies in their respective satellite and suburban communities was increasing.
They later explained the central-city decline with four factors: (1) the
general-interest content of the metropolitan daily newspapers had been subsumed
by television news--competition from another medium; (2) "input" costs, notably
newsprint, had increased, causing increases in circulation and content to become
more expensive, a problem that increases with the size of the paper; (3) labor
costs had risen in the central cities; and (4) urban demographics had changed as
population moved to the suburbs and was followed by potential retail and
Rosse & Dertouzos's analysis of the New York City market showed the
expected decrease in central city penetration and the domination of the central
city by three major daily newspapers, but not the predicted exclusive, limited
geographic circulation in the second and third layers. However, they felt the
umbrella model was supported because the three suburban daily newspapers had
stronger penetration in their circulation areas than did the central-city
dailies, and the individual New York City newspapers still circulated most
densely in the central city. Even so, suburban penetration of the central-city
dailies was substantial and specific to certain suburban communities.
Rosse (1975) had suggested that advertising competition would not be
"effective" within layers. He predicted that the capability of television to
carry national advertising to a broader audience than possible with individual
metropolitan newspapers would result in geographic selectivity in circulation as
the metropolitan dailies took advantage of economies of scale and pulled back
their reach. This selectivity would shrink the reach of these first layer papers
in the umbrella and encourage smaller-layer papers to broaden their content.
Underlying the umbrella model seem to be four assumptions, both stated and
1) circulation penetration will be strongest in the newspaper's home
2) analysis of penetration will substitute for an analysis of advertising
[Rosse & Dertouzos indicated that other factors besides circulation affect
advertising, but said "audiences are necessary before advertising can be sold"
and presented circulation as evidence (1978: 148)]
3) as newspapers move farther from layer #1, advertising will become
4) as newspapers move farther from layer #1, content will become
increasingly local and of lesser "quality." Quality was not defined.
Since Rosse presented his umbrella model, other researchers have attempted
to duplicate Rosse's results in other geographic areas or have used his model as
a base for additional research. Most of the research is based on one of the
Circulation-based studies: Tillinghast (1988) tested the umbrella model in
southern California where two metropolitan newspapers in Los Angeles competed
with two satellite and 20 suburban dailies. The suburban newspapers had heavy
competition from the two higher layers in their retail trading zones, less so in
their city zones. There was little competition among the suburban dailies during
the week, although on Sundays those with Sunday editions had substantial
circulation in the city zones of their suburban competitors without the Sunday
edition. Still, the metropolitan Times had even higher penetration in those
suburban communities. Tillinghast concluded that in this southern California
market the umbrella model was supported by the inter-layer competition.
Devey (1989) examined circulation in the Boston area from 1945 to 1985 when
Boston had as many as three competing metropolitan newspapers. She found that
circulation increased faster at the lower umbrella levels than at the
metropolitan levels, although Globe (the leading paper) circulation increased
faster than the lower-level newspapers combined. The trailing metropolitan
papers lowered the metro figures overall.
More importantly, circulation in satellite and suburban papers increased
proportionately to increases in population. Suburban newspapers grew faster than
their population; therefore, Devey suggests that competition is strongest
between satellite and suburban newspapers when the metropolitan layer has
newspapers that are concentrating on competing between or among themselves.
Although she did not replicate Rosse & Dertouzos's penetration analysis in each
community, the findings suggest that full umbrella inter-level competition is
not a factor in the Boston area, because circulation as a proportion of the
existing population is apparently not affected by the metropolitan newspaper
situation. However, Rosse's prediction that lower layer papers would grow if
metropolitan circulation was pulled back was supported.
Using 900 suburban newspapers from a broader geographic base, an earlier
study (Niebauer, et al., 1988) examined the effects of three conditions of
central-city competition on both daily and weekly suburban newspapers, and the
results were similar to Devey's. The strongest predictor of presence of a
suburban newspaper was the population of that suburb. The authors were not
testing the umbrella model, but a very weak link between circulation of the
metropolitan dailies and absence of a suburban daily newspaper in a community
indirectly supports Rosse's model of inter-layer competition, regardless of the
intra-layer competition status of the metropolitan newspaper.
Lacy and Davenport (1994) and Lacy and Dalmia (1991) tested the umbrella
model using county penetration, loosening the criteria that determine
competition. The 1991 study was restricted to Michigan and partially supported
the umbrella model. The examination of Michigan newspapers identified some
intra-layer competition, contrary to expectations of the umbrella model, and
identified a continually changing industry. The 1994 study extended the Lacy and
Dalmia concept to the national level and replicated the results of the Michigan
study. Nearly half of the counties had daily newspapers circulating from two or
more layers and nearly half had two or more newspapers from the same layer. The
two studies used a new definition of competition; Rosse's standard of
competition was much more restrictive. Lacy and Davenport suggested that
newspapers could pursue regional competition, and Lacy and Dalmia suggested that
the density of population in an area may affect the umbrella model.
Lacy and Dalmia (1991) suggested that two additional layers should be added
to Rosse's model: (1) isolated dailies and (2) isolated weeklies that are too
far from the metropolitan area to be reached by a metropolitan newspaper. Lacy
(1988) had earlier suggested in a review that a broad layer of national
newspapers overlays the first metropolitan layer suggested by Rosse and added a
layer of grouped nondailies. Lacy also noted that group-owned suburban daily
newspapers form a separate competitive environment between the suburban and
Advertising-based studies: Using national advertising figures to determine
elasticity of demand, Busterna (1987) refuted Rosse's and Owen's assertions that
competition for national advertising in newspapers comes from competing national
media. Busterna did not examine individual markets.
Combined Advertising and Circulation-based studies: If predictions of
newspaper executives can substitute for circulation and advertising statistics,
a regional study indicated that weeklies (the fourth layer) perceive more
competition from suburban than from metropolitan dailies (Lacy, 1984). Suburban
publishers predicted that removal of the metropolitan daily newspaper from their
umbrella would bring more circulation but not affect advertising; weekly
newspaper publishers predicted the reverse. Metropolitan daily executives were
more likely to predict an increase in inter-layer competition than were
publishers in the lower layers, counter to Rosse's prediction (as Lacy noted).
Lacy did not separate the satellite and suburban layers. Further analysis of the
data (Lacy, 1985) indicated that lower-layer publishers under competitive
metropolitan dailies perceived more advertising competition than those under
monopoly metropolitan dailies. The reverse was true for competition for
circulation. Advertising competition was seen as more of a threat when the
second-layer daily was within 20 miles of the first-layer community.
Content Studies: Content studies have been based on the premise that
substitutability of content (both news and advertising) will affect competition
and on a concern for quality of content. Using an indirect measure of content,
Lacy (cited in Lacy, Fico and Simon, 1988) found that intercity competition was
related to the proportion of the newspaper given to news and the proportion of
the newshole allocated for local news. Although not specifying where competition
fit into the layers, Lacy, Fico and Simon (1989) used path analysis with 21
large newspapers to determine whether competition between cities was related to
quality of content. Assuming that a smaller workload for reporters would mean
more attention to individual stories and therefore higher quality content, they
found that competition from other cities was positively related to fairness in
stories but negatively related to reporter workload and by extension to story
Attempting to determine cross-elasticity of demand, Lacy and Sohn (1990)
compared content about suburban areas in Detroit and Denver metropolitan
newspapers and content of nondaily newspapers in the same suburbs with the
circulation of each newspaper in the relevant suburb. They also examined display
and insert advertising. Comparing similarity of correlations, they found little
evidence of substitutability of the metro and weekly newspapers in Detroit. But
in Denver, the correlations indicated that display advertising, insert
advertising, and local sports coverage were substitutable content. In both
metropolitan areas circulation of the suburban weeklies had high correlations
with local sports, local editorials and local social news, while the
metropolitan dailies had high correlations between circulation and display
Summary of Direct Tests of the Model: Some of the preceding studies
examined elasticity of demand for advertising and variations in newspaper
content under differing conditions of competition, but those studies that
directly tested the linear umbrella model have:
(1) looked at the effects of metropolitan intracity competition on the
(2) examined the inter-layer circulation competition in different
(3) proposed expanding the definition of circulation competition from the
city to the county
(4) suggested adding more layers to the model.
The effects of metropolitan intracity competition were mixed. Tillinghast
found that with metropolitan intracity competition the umbrella was supported,
but Devey's conclusions were just the opposite. Neibauer et al. also found that
the metropolitan intracity status did not affect the suburban market.
Geographically, in examinations of circulation the model was supported in
California, not supported in Boston, and mixed in Michigan and nationally.
While the Lacy and Dalmia and Lacy and Davenport studies may suggest reason
to discredit Rosse's assumption of no intra-layer competition (the authors
carefully interpret their findings as demonstrating "potential" for intra-layer
competition), they do not apply as rigorous a test as did Rosse and Dertouzos,
who tested the assumption by comparing circulation density within specific
communities (p. 39). Both Lacy and Dalmia and Lacy and Davenport tested the
assumption by noting presence or absence of the competing newspaper(s) in a
broader geographic area, the county; presence was defined by 5 percent
penetration. This measurement makes intuitive sense, but if applied to Rosse's
example, would be a reinterpretation of the San Francisco data. Alameda County
had four and Contra Costa County had three suburban daily newspapers circulating
and would under the Lacy and Dalmia and Lacy and Davenport criterion be
illustrative of intralayer competition. Thus the evidence used to demonstrate
Rosse's model would refute it. As indicated by Lacy and Davenport, the county
aggregate data do not permit conclusions about substitutability of the
individual dailies within the county. This caution would also apply to the
amount of competition between and among the dailies. While under the Rosse
measure there is also no guarantee of the influence of the local daily in a
specific community, this consideration is even lower when the county is the
standard. There is also growing evidence that in the 1990s, these intra-county
dailies may be cooperative rather than competitive (see below).
Additional layers proposed were a national layer, a separate suburban layer
of chain-owned dailies or weeklies, an isolated suburban and an isolated weekly
INDUSTRY CHANGES SINCE ROSSE
Since the late 1970s when Rosse presented his umbrella model, the newspaper
industry has focused seriously on expanding its penetration into nonlocal areas
and on making smaller daily and even weekly newspapers attractive to nonlocal
advertisers. The industry is experiencing some attempts by both satellite and
suburban papers to circulate beyond their geographic boundaries and even
attempts to attract commuters from other geographic areas, because commuters
spend much of their time in the work communities and are useful to the
advertisers there. The industry also is experiencing extensions of group
ownership that combine geographically proximate suburban newspapers into one
package for national and regional advertising, and, as Lacy had noted for
suburban chains in general, this regional package competes with the other levels
This suburban group ownership can also provide a centralized news office
that can provide news common to chain members and take advantage of economies of
scale that had formerly been restricted to larger dailies. Rosse had made the
assumption that suburban daily newspapers are primarily local in content, but
with the availability of wire services and feature syndicates, some suburban
papers have little local news as they strive to compete by providing a broader
news function for their readers. Owen (1975) had remarked that wire services and
feature syndicates affect the "'intraumbrella' effectiveness of competition"
because they can exclude a newspaper from their services (p. 53). Rosse had
predicted that if lower-layer newspapers could expand their news base, they
would compete more intensely with the upper-level dailies. Because smaller
dailies generally have lower salaries, maintaining the staff necessary to cover
and investigate local news is often a problem and therefore "canned" material is
more economical for them.
To demonstrate his umbrella model Rosse had presented an analysis of
circulation penetration proceeding linearly from the largest community to the
smallest; we are suggesting
-- that changes in technology and in the industry suggest that newspaper
competition is no longer explained by a line [in an "umbrella," a spoke] from
the lowest level newspaper through to an overriding metropolitan daily (or
competitive metropolitan dailies) that shades each succeeding level, as has been
the interpretation of Rosse's umbrella. Rather we are suggesting that the
newspaper market operates as a series of concentric circles - still under an
shading through overlap - but without the linearity (see Figure 2 below).
-- that advertising must be examined separately from circulation. Picard
(1993) indicated that newspaper revenue comes from two individual streams, 65 to
80 percent from advertising and 20 to 35 percent from circulation. In addition,
as we will demonstrate below, local circulation may not be the determining
factor for advertising.
--that circulation analyses will demonstrate that both inter- and
intra-layer competition are common and that this competition is not limited to
specific linear sublayers.
DATA FOR THE CIRCULAR MODEL
Although newspaper competition comes from nonprint media as well as other
newspapers, at least one study (Lacy, 1988) found little impact of intermedia
competition on allocation of resources used in the news process. Therefore, this
discussion will concentrate on newspapers as a separate competitive system. When
circulation is discussed, daily circulation will be the base of analysis.
Circulation: As mentioned above, studies have tested Rosse's model in
California, Colorado, Michigan and nationally. They used both county penetration
and city penetration. We purposively attempted to find a competitive situation
in a different geographic area where satellite newspapers operated, and located
competing large dailies in the Tampa Florida region. Audit Bureau of
Circulations (ABC) data indicate that the Tampa region has two large daily
newspapers, The Tampa Tribune and the St. Petersburg Times, competing against
each other. The Times maintains a separate SRDS listing under Tampa as well as
under St. Petersburg. No finite definition separates metropolitan daily
newspapers from satellite newspapers, but the size of these newspapers compared
to metropolitan daily newspapers such as The Dallas Morning News, the San
Francisco Chronicle, or even the Atlanta Journal/Constitution lead us to
categorize the two Florida dailies as smaller, satellite dailies. The Miami
Herald (which circulates in Hillsborough county where Tampa is located) provides
weak metropolitan "shade." The Orlando Sentinel is also a smaller presence in
As Lacy (1988) suggests, a national layer exists.USA Today and the Chicago
Tribune circulate in Hillsborough and Pinellas counties (St. Petersburg is in
Pinellas county), creating a separate national layer. Thus under our definition,
the layers in the Tampa area are national, metropolitan, satellite, suburban and
weekly. Because the retail trading zone associated with the Tampa newspaper
covers a finite area, we are interested here in the relationship between and
within the satellite and suburban layers.
ABC weekday city-penetration figures for the four suburban daily newspapers
in the Tampa designated market are presented in Table 1. If we look only at
the raw penetration percentages, the linear interpretation of Rosse's umbrella
model is supported. But a closer look indicates that the St. Petersburg Times is
making inroads into the Tampa city circulation. Although showing only 5 percent
penetration, this is one-fifth the amount that the Tampa Tribune has for itself
in its home community. The suburban layer also has a second interpretation if we
look at the overlap in the Sarasota Herald-Tribune and the Bradenton Herald.
Although the Bradenton daily has minimal penetration in Sarasota (less than 2
percent), the Sarasota daily has 14 percent penetration in Bradenton, where the
Bradenton Herald has 31 percent itself. This is almost half as much penetration
as the local daily. Although we could say these figures are not as high as the
city-based daily, the penetration is strong and these figures suggest that there
is indeed competition for circulation within the suburban layer.
Inter-layer competition is most prevalent in Lakeland (for the Tampa
Tribune) and in Crystal River (for the St. Petersburg Times). In Crystal River,
the St. Petersburg newspaper has almost half as much penetration as the local
-------TABLE 1 ABOUT HERE -------
Table 2 presents ABC analysis of proportions of county circulation of daily
newspapers in the same Florida market. The Tampa Tribune increases its
proportion of circulation in Citrus County; there is a minor amount of
circulation from the suburban dailies in the satellite layer; but overall the
patterns of circulation competition do not change.
Advertising: For advertising competition, we start with general information
from SRDS that demonstrates changes in the industry's approach to attracting
advertising. The move is to cooperation, especially within layers for nonlocal
advertisers, rather than competition.
First, in the early 1980s the newspaper industry adopted the Standard
Advertising Units (SAU), a mechanical device that signaled a cooperative
attitude among competing daily newspapers, regardless of their placement in the
umbrella model. The significance of the SAUs was that newspapers were committing
to standardizing their formats to six-columns and standardizing their
advertising sizes to accommodate potential advertisers. The individualized
formats used prior to the SAU agreement meant that advertisers had to reformat
ads to fit into various structures if they wanted to reach readers of different
newspapers. Under the SAU format, an ad of one size would fit the same way into
all newspapers using the SAUs, thus reducing first-copy costs for advertisers
who might wish to advertise in smaller newspapers. The industry also cooperates
nationally in six common retail categories to place ads through the National
Newspaper Network in combinations of their member daily newspapers. These
newspapers operate at all levels of the "umbrella."
More important is the regional focus of the suburban "groups." Newspaper
analyst Morton (1997) calls this proliferation of regional groups "clustering,"
a term he says will be the buzzword of the 1990s. The industry has always had
groups of smaller newspapers, but as Morton explains, groups are buying and
trading with regionalism as the goal. As the clustering becomes more and more
regional, the potential increases for production economies of scale that were
available only to the larger newspapers in Rosse's analysis. Consolidated
production and administrative activities can reduce costs and enable the group
to present an advertising package to retailers whose buying decisions have been
becoming more and more national and regional. Morton also suggested that the
clustering would improve news content through shared coverage and features.
What these trends imply is that at least at the suburban level competition
for advertising across layers is resulting in cooperative ventures within layers
beyond the local market. This cooperation creates overlap in the pool of
potential advertisers, a contradiction of the assumption behind the linear
interpretation of the umbrella model. The inter-layer competition that is the
end result of this cooperation will make the regional package competitive as a
unit rather than on a one-on-one basis with larger newspapers. The demographics
of the relevant suburban areas could make these clusters very competitive with
the overall market of larger dailies. Delivery and newsprint costs still
restrict indiscriminate circulation growth. Demographics rather than density may
Returning to the Florida markets, SRDS indicates that four daily and two
weekly newspaper regional groups were operating in the state at the end of 1995.
Seven weekly newspapers are part of the Tampa Suburban group, providing
competition to the Tampa region from the fourth layer of Rosse's model. Five
daily newspapers are clustered in the Gainesville area, where metropolitan daily
circulation is minimal. The largest proportion in Alachua county (Gainesville)
is 2.2 percent for the Orlando Sentinel. The Gainesville-area group presented
paid daily circulation of 158 thousand and asked $110.29 per inch for a
black-and-white ad, or .0698 cents per thousand. This compares to the
metropolitan Orlando Sentinel , which presented 281 thousand circulation and
asked $204 per inch for a black-and-white ad, or .0726 cents per thousand.
Demographics and other market factors should have more consideration in this
competitive situation than circulation per se.
The Florida groups and group rates indicate that the competition for
advertising is not an attempt to eliminate another suburban newspaper but is
focusing instead on the ways daily newspapers can cooperate to become more
attractive for advertisers outside of the local market. The local market is
weak; retailers have been consumed by large chains, and the large chains want
more than a small local market for their advertising dollars. By responding to
these advertiser needs, the suburban newspapers are moving away from the linear
umbrella model and becoming a fluid, changing layer that competes differently
under differing regional conditions.
The larger daily newspapers have been experimenting with on-line news
delivery and are beginning to see financial returns. When advertising becomes a
major component of these web sites, the fluidity of competition will change
THE CIRCULAR MODEL
Although the Florida figures are not definitive, combined with prior
studies they suggest that the linear model no longer fits the industry as a
whole. The proposed circular model accounts for the changing nature of the
newspaper market, as compared to the false sense of order and proportion between
the layers portrayed by the linear umbrella model. The market is more fluid than
it was in 1975, and the area of natural advantage for each newspaper will depend
both on the geographic location and on the role the newspaper has chosen to play
in its particular market. (Lacy and Dalmia, 1991, for example, had noticed that
in Michigan the newspaper industry was continually changing.) The newspaper's
choice of a role is a major consideration, as this model accommodates the
behavior of the newspaper rather than just its location. This role will also
determine the content or product differentiation portrayed by each newspaper.
The proliferation of chain ownership has also changed the market, providing the
potential for economies of scale for smaller newspapers and reallocation of
resources among the larger dailies.
As Figure 2 demonstrates, the circular model retains Rosse's layers, but
incorporates them into concentric rings rather than assuming the linearity of
layers of larger and larger umbrellas.
---------------------- Figure 2 about here -----------------
In Figure 2 the outer ring would include national newspapers (Lacy, 1988,
had also suggested a national layer) of both newspapers with national content
such as USA Today and specialized/other newspapers that may circulate nationally
such as The New York Times, The Wall Street Journal and the Christian Science
Monitor. Also included in this ring would be newspapers that circulate in a
broad geographic area of the country.
The second ring includes the major metropolitan newspapers that circulate
throughout the state or a lesser region than those in the first ring. These
newspapers may have zoned editions or may circulate daily with constant
material. The rings would support Rosse's assumption that the metropolitan
newspapers would be more global in content and would have more circulation
density in the metropolitan city.
The third and fourth rings are made up of smaller city daily newspapers,
but their area of competition is not restricted. The satellite newspapers could
compete either intralayer or interlayer and are expected to expand their reach
beyond the satellite central city (note the data for Tampa and St. Petersburg in
this study, where the St. Petersburg Times is beginning to compete in Tampa, for
The suburban daily newspapers also show competition within their layer, in
this study demonstrated by the movement of the Sarasota Herald-Tribune into the
Bradenton city area. Mobile commuters who are commuting more between suburbs
than to central cities (Stepp, 1996) reduce the stability of reader interest in
the suburban cities. This commuting trend and geographic clustering of the
suburban chains (Morton, 1997) are expected to make the suburban competition
even more fluid. Lacy (1988) had suggested that grouped suburban dailies
operated between two rings, but that distinction is implied by the fluidity of
A fifth ring includes weekly newspapers, shoppers, and ethnic and other
specialized nondaily newspapers operating within any geographical area
throughout the state or region, a concept articulated by Rosse. Newspapers in
this ring generally have control within the area they choose to serve. A shopper
may operate in a single geographic area, while an ethnic newspaper may circulate
regionally or even farther. They generally have little competition in their own
niche. Lacy and Dalmia (1991) had addressed the isolated weekly, and Lacy (1988)
had noticed the effect of grouped nondailies, but both are incorporated in the
nondaily, specialized ring.
The ring model presumes a fluidity of competition that responds to changes
in either advertising prices and/or circulation, regardless of the local area of
natural advantage. Rosse had predicted many of the changes in the newspaper
industry and had articulated scenarios that would result in other changes. This
concentric circle model of newspaper competition incorporates these changes.
Audit Bureau of Circulations. "Audit Report: Newspaper": Bradenton Herald
Audit Bureau of Circulations. "Audit Report: Newspaper": Citrus County
Chronicle (July 1996).
Audit Bureau of Circulations. "Audit Report: Newspaper": The Ledger (April
Audit Bureau of Circulations. "Audit Report: Newspaper": St. Petersburg
Times (August 1996).
Audit Bureau of Circulations. "Audit Report: Newspaper": Sarasota
Herald-Tribune (May 1996).
Audit Bureau of Circulations. "Audit Report: Newspaper": The Tampa Tribune
Audit Bureau of Circulations. "County Penetration Report": Florida (October
Busterna, John C. "The Cross-Elasticity of Demand for National Newspaper
Advertising." Journalism Quarterly 64 (1987): 346-51.
Devey, Susan M. "Umbrella Competition for Newspaper Circulation in the
Boston Metro Area." Journal of Media Economics (spring 1989): 31-40.
Lacy, Stephen. "Competing in the Suburbs: A Research Review of Intercity
Newspaper Competition." Newspaper Research Journal 9 (1988): 69-76.
Lacy, Stephen. "Competition among Metropolitan Daily, Small Daily and
Weekly Newspapers." Journalism Quarterly 61 (1984): 640-644, 742.
Lacy, Stephen. "Effect of Intermedia Competition on Daily Newspaper
Content." Journalism Quarterly 65 (1988): 95-99.
Lacy, Stephen. "Monopoly Metropolitan Dailies and Inter-City Competition."
Journalism Quarterly 62 (1985): 640-644.
Lacy, Stephen, and Shikha Dalmia. "Michigan Newspaper Competition from 1980
to 1986: Expanding the Geographic Application of the Umbrella Model." Paper
presented to the Association for Education in Journalism and Mass Communication,
Lacy, Stephen, and Lucinda Davenport. "Daily Newspaper Market Structure,
Concentration and Competition." The Journal of Media Economics 7 (1994): 33-46.
Lacy, Stephen, Frederick Fico, and Todd Simon. "Relationships among
Economic, Newsroom, and Content Variables: A Path Model." Journal of Media
Economics (1989): 51-66.
Lacy, Stephen, and Ardyth B. Sohn. "Correlations of Newspaper Content with
Circulation in the Suburbs: A Case Study." Journalism Quarterly 67 (1990):
Morton, John. "A Flurry of Deals to Create Clusters." American Journalism
Review (1977 March), 52.
Niebauer jr., Walter E., Stephen Lacy, James M. Bernstein and Tuen-yu Lau.
"Central City Market Structure's Impact on Suburban Newspaper Circulation."
Journalism Quarterly 65 (1988): 726-732.
Owen, Bruce M. "Newspapers" in Economics of Freedom of Expression: Media
Structure and the First Amendment (Cambridge MA: Ballinger Publishing Co.,
Picard, Robert G. "Economics of the Daily Newspaper Industry" in Alison
Alexander, James Owers and Rod Carveth, eds., Media Economics: Theory and
Practice (Hillsdale NJ: Erlbaum, 1993): 181-203.
Picard, Robert G. Media Economics: Concepts and Issues. Newbury Park: Sage,
Rosse, James N. "Economic Limits of Press Responsibility." Studies in
Industry Economics No. 56, Paper presented at Duke University Center for the
Study of Communications Policy, January 1975.
Rosse, James N. and James Dertouzos. "Economic issues in Mass
Communication." Proceedings of the Symposium on Media Concentration, Vol. 1,
presented at the Bureau of Competition of the Federal Trade Commission, December
Stepp, Holly E. "Public Transit Use Down, Study Finds: Many Commuting
between Suburbs rather than to Cities, Census Data Show." The Dallas Morning
News (August 16 1996), p. A1, A14.
SRDS Newspaper Advertising Source 77:12 (December 1995).
Tillinghast, Diana S. "Limits of Competition," in Robert G. Picard, James
P. Winter, Maxwell E. McCombs and Stephen Lacy, eds., Press Concentration and
Monopoly (Norwood NJ: Ablex, 1988): 71-87. Table 1: Daily Penetration of Tampa
Designated Market Cities, 1996 ABC data.
St.Peters- Braden- Crystal
City: Tampa burg ton Sarasota Lakeland River
Tribune 25.45 1.85 1.76 .45 10.24 1.86
Times 5.05 17.92 1.64 .22 n.a. 4.47
Herald n.a. n.a. 31.10 1.69 n.a. n.a.
Herald-Trib. n.a. n.a. 14.40 41.81 n.a. n.a.
Ledger n.a. n.a. n.a. n.a. 67.23 n.a.
Citrus Cty. n.a. n.a. n.a. n.a. n.a. 9.24
Table 2: Proportion of Daily Circulation in Tampa Designated Market Counties,
1996 ABC data (in percents).
Hills Pinnel- Mana- Sara-
County: borough las tee sota Polk Citrus
Tribune 40.6 6.9 2.6 1.4 9.7 11.6
Times 5.7 61.0 2.0 .2 n.a. 21.3
Herald .0 n.a. 42.4 .8 n.a. n.a.
Herald-Trib. n.a. n.a. 19.2 58.0 n.a. n.a.
Ledger .2 n.a. n.a. n.a. 45.3 n.a.
Citrus Cty. n.a n.a n.a. n.a. n.a. 42.0
Today 1.6 1.1 1.3 2.1 .8 .8
Herald .2 n.a. n.a. .2 .0 n.a.
Sentinel .1 n.a. n.a. n.a. 1.5 .8
Tribune .2 n.a. n.a. n.a. n.a. n.a.
Puerto Rico .0 n.a. n.a. n.a. n.a. n.a.
Sun-Herald n.a. n.a. n.a. 1.5 n.a. n.a.
Star-Ban'r. n.a. n.a. n.a. n.a. n.a. 2.6 Figure 1: Rosse's Linear Umbrella
Model of Newspaper Competition
from Owen (1975), p. 51
from Pickard (1989), p. 30 Figure 2: Concentric Circles: The Ring Model of
 . Information attributed to SRDS is from the December 1995 volume. ABC data
are from 1966 audit reports.
 . These city-penetration figures were calculated from ABC zipcode
distribution data by adding all one-day gross distribution attributed to the
relevant cities by ABC and dividing by the 1995 projected number of households
in the community, as provided by ABC. These figures may tend to overestimate the
penetration, because gross distribution figures may include nondelivered
newspapers and a few zipcodes cover more than one community. However, the
delivered newspaper figures were estimates.
 .Lacy and Dalmia (1991) cited an unpublished paper by Shikha Dalmia, "Ring
Theory as an Alternative to the Umbrella Model of Newspaper Competition: A
Study" but did not expound on the information. The paper was unavailable.