This paper was presented at the Association for Education in Journalism
and Mass Communication in Toronto, Canada, August 2004.
If you have questions about this paper, please contact the author
directly. If you have questions about the archives, email
[log in to unmask] For an explanation of the subject line, send email to
[log in to unmask] with just the four words, "get help info aejmc," in the
body (drop the "").
Community Newspapers As Instruments of Social Control:
Updating Community Conflict and the Press
Michael L. Thurwanger
6415 N. Suffolk Drive
Peoria, IL 61615
Phone: (O) 309-677-2366
E-Mail: twanger@ bradley.edu
Newspapers As Instruments of Social Control
Community Newspapers As Instruments of Social Control:
Updating Community Conflict and the Press
Nearly a quarter century ago, the Minnesota research team of
Tichenor, Donohue and Olien (1980) published their study of local
newspapers in community conflict, concluding that they "are not the
independent, self-styled social agents that either they or members of the
public imagine them to be" (p. 217). Among their findings was that local
newspapers perform a social-control function through the management of
information provided to the community and this role is most evident in
small, homogeneous communities. Echoing the findings of Janowitz (1967),
they concluded that local newspapers avoided reporting internal conflict in
the interest of maintaining harmony while presenting an image of
consensus. Such content decisions serve as a form of control that "has
far-reaching implications for what the community will hear about, think
about, and talk about" (1980, p. 20).
They called for additional research to validate their findings in the
face of ongoing social and media industry changes. Among those changes are
trends reflecting a decline in local autonomy as state and federal
governments, major businesses and outside industries have an increasing
influence on local policies and economies. At the same time, mass media
usage and media industry practices have undergone significant change in the
face of ownership consolidation and a proliferation of new media choices.
This study proposed to build on the work of Tichenor, Donohue and
Olien and explore the social control function played by local newspapers
serving communities covering a significant issue having the potential to
cause conflict within the community. The study also attempted to validate
several of the research team's earlier findings with regard to the
correlation between community pluralism and evidence of social
control. Finally, the 24-year span of analysis covered in this
study offered an opportunity to test for evidence of change in the social
control function due to societal and media industry changes over time.
The focal point for this study was the issue of prison site-selection
in the state of Illinois. In the past, corrections facilities were often
opposed by local residents, however some communities have begun to court
these facilities as sources of economic growth. The clash between
individual concerns and community desires provided a likely source of
conflict in which the local newspaper could be expected to play an active role.
Prison issue. During the past quarter century, this nation
experienced an unprecedented boom in its prison population and in the
construction of the prison infrastructure necessary to house it. In 2001,
state and federal prison populations in the United States exceeded 2.1
million inmates ("Illinois prison population," 2003, July 28). Prison has
become big business and an industry unto itself (Hallinan, 2001; Schlosser,
1998). Hallinan (2001) raised concern over the lack of national debate
surrounding this critical social and policy issue. "At the time the prisons
were being built, almost no one publicly questioned the consequences of
this growth. No one asked what happens when prison becomes an industry,
like steel or coal, or when large numbers of free people are given an
economic stake in the imprisonment of others" (p. 85).
Illinois reflects national trends, having nearly tripled its
inventory of state correctional facilities during a period stretching from
1980 through the most recent site selections announced in April
2001. Illinois corrections construction typified national patterns in
which the vast majority of new prison sites were built in or near small,
rural communities. This pattern of site selection served as a form of
economic subsidy for communities in depressed regions while meeting the
state's need for expediency by locating traditionally undesirable
facilities where they were least likely to face organized opposition
(Gibbons & Pierce, 1995; Schlosser, 1998; Shicor, 1992; Takahashi & Graber,
1998). Prisons have typically elicited a "Not-In-My-Backyard" reaction in
the past with local residents expressing concerns about such issues as
personal safety, threats to community integrity represented by an influx of
prison staff and inmate families, and a "prison town" stigma (Carlson,
1988, 1990; Krause, 1992; Popper, 1981). While larger communities with
more diversified economic and political bases continue to oppose such
proposals, small economically-challenged communities now actively compete
for selection as a prison site. Local leaders work within these
communities to garner support and present an image of community consensus
fully supportive of the prison proposal (Schlosser, 1998).
Community newspapers. Janowitz (1991) identified the mass media's
traditional role in democratic society as threefold: contributing to a high
level of participation, stimulating meaningful deliberation upon which
citizens arrive at voting decisions, and operating to preclude one side
from gaining undue advantage through the press. To the extent that the
press failed to achieve these ideals, the democratic process does not
represent true consent but becomes "an exercise in mass pressure" (p. 245).
Tichenor, Donohue and Olien (1980) observed that the traditional
image of the community newspaper has been one of self-determination and
independence, though they argued that this is not an accurate
image. Analysis of community newspaper content has consistently shown that
internal community conflict was avoided and controversy received little
emphasis in news or editorial content when compared to larger newspapers
(Breed, 1958; Janowitz, 1967, 1991; Tichenor, Donohue & Olien,
1980). Community newspapers have been found to perform a social
integration function in which regular readers use newspaper content to
establish ties with their community, resulting in higher levels of
community cohesion, identification and participation (Davidson & Cotter,
1997; Edelstein & Larson, 1960; Janowitz, 1967; Merton, 1949; Park, 1925,
Park (1925) was one of the first to break with the traditional image
of the press as independent, portraying the relationship between the local
press and community as interdependent. Janowitz (1967) characterized
newspapers as a sub-set of a larger community system where the community
audience "conditions its content, determines its appeal, and facilitates
its impact" (p. xxi). Others have noted the economic imperative for the
newspaper to create a place for itself within that community identity
because the success of its product depends on the interest and acceptance
of community members (Edelstein & Schulz, 1963; Kaniss, 1991). Banfield
and Wilson (1963) argued that the local newspaper's interest in economic
growth "inclines it toward boosterism, and inclines it also on occasion to
'play down' or even suppress news that would put the city in a bad light"
Janowitz (1967) identified the functions of the editor and publisher
as central to the study of community newspapers, suggesting that the
publisher served as an index to understanding how conflicts between
competing interest groups are resolved at the local level. Webster (1987)
described the publisher's position as an ethical dilemma—the demands for
journalistic standards of objective detachment versus the responsibility as
a local leader to support community economic and social development. A
number of researchers have commented on this dual expectation. Publishers
and editors must represent the newspaper and observe standards of
objectivity and balance but, as educated and influential members of the
community, they are expected to be active participants in the community's
administration (Byerly, 1961; Fitzgerald, 1996; Gaziano & McGrath, 1987;
Janowitz, 1967; Sneed & Riffe, 1991; Tuchman, 1978).
Others focused on business responsibilities, noting that publishers
and editors must maintain close ties with other business leaders and those
ties are critical to the newspaper's economic survival (Byerly, 1961;
Janowitz, 1967; Kennedy, 1974). Breed (1955) noted that the influence of
the publisher on newspaper policy is constrained by journalistic norms, but
he found that policies regarding news coverage of political, economic and
labor issues were subtly shaped by omissions in coverage, preferential
selection and placement.
In spite of references to publishers and editors as members of the
community elite, Donohue, Tichenor and Olien (1995) argued that the
community newspaper and its representatives are not integrated as equals in
the community power structure due, in part, to the newspaper's dependence
on members of the power structure for access and information. This
reflected Janowitz' (1967) view of community newspapers, not as part of the
power structure, but as an intermediary between individuals and major
community institutions. Donohue, Tichenor and Olien (1995) described the
role of the local newspaper as presenting a sense of community consensus,
distributing information on behalf of local elites. Far from a watchdog
that scrutinizes the actions of those in power, newspapers in small
communities were described as a "sleeping guard dog" protecting the status
quo and serving the interests of the local power structure (p. 116).
Social control. Warren (1978) defined social control as the
"process through which a group influences the behavior of its members to
conform with its norms" (p. 10-11). This concept has undergone
considerable change in terms of its definition and focus (Janowitz, 1975,
1978, 1991; Roucek, 1978; Sumner, 1997). During the early part of the past
century, social control was primarily concerned with a benign model
exploring social organization and how groups regulate themselves (Janowitz,
1991). Early sociologists such as Mead, Dewey, and Park all addressed a
need for society to construct and maintain shared systems of values and
social norms. Dewey (1927) and Park (1925) suggested that the press was
one of the institutions that served this function.
Martindale (1978) noted that social control was initially presented
as a natural order that society pursued instinctively with individuals
requiring some degree of external control to overcome self-interest and
join in a common effort to attain order. However, research shifted to an
exploration of the means by which social controls were applied, often in
connection with the related concepts of socialization and deviance. This
shift sought answers to identify who exercises social control and for what
purpose? (Roucek, 1978).
Gamson (1968) viewed social control as a means of eliminating or
minimizing the influence of competing constituents. Local leaders have two
options to deal with constituents—appeasement by meeting their demands, or
control through persuasion and other means. Both Dahl (1982) and Stone
(1989) addressed the costs and benefits of social control in terms of
political and economic capital. They noted that individuals and groups
might influence policy-making processes by increasing the costs of
maintaining social control through such tactics as activism and
litigation. Conversely, Dahl (1961) observed that where competition has
been removed by effective social control, political entrepreneurs can more
readily exert their infleunce. Gamson (1968) listed a number of ways that
authorities could exert control, including efforts to influence public
opinion on an issue and the regulation of access to decision-makers and to
Social control function of newspapers. McCombs (1997) cited local
media sources as playing a critical role in setting a community agenda that
emphasizes arriving at and maintaining consensus on community
issues. Janowitz (1967) found that community newspapers served to maintain
local consensus through an emphasis on shared values rather then on the
resolution of conflict. Internal conflicts were viewed as divisive while
conflicts with external forces helped unify the community against an
outside threat and legitimized the role of local leaders in representing
community interests. Breed (1958) observed that local news coverage often
conveyed a "chamber of commerce attitude" (p. 111) in which newspapers
served as local boosters to ensure community support and attract economic
growth. This was reflected in editorial decisions against covering
failures or social aspects that undermined the community image. Breed
conducted what he termed "a reverse content analysis" (p. 112) to identify
those elements of news and information that are regularly omitted from
local coverage. The most frequent omission was related to what Breed
labeled the "undemocratic power of business elites" (p. 111). He concluded
that these omissions in coverage served to preserve citizen confidence in
the community and its institutions.
Among the types of social power identified by Galbraith (1983) is
conditioning power, the ability to win submission without
coercion. Control of information has been identified as a form of
conditioning power and Mott (1970) identified information control as a
vital resource in establishing social control. He attributed that power to
mass media organizations by virtue of their role in identifying, defining
and articulating social problems. Gamson (1968) identified news management—
the selective withholding of information—as a technique used in maintaining
social control. A number of theorists support the view that the community
newspaper plays a significant role in maintaining social control by
selectively reinforcing, legitimizing and endorsing accepted views
(Altschull, 1995; Carey, 1979; Donohue, Tichenor & Olien, 1995; Janowitz,
1978, 1991; Lasswell, 1949; Shoemaker & Reese, 1996; Tichenor, Donohue &
Tichenor, Donohue and Olien studied community newspapers, concluding
that they "exercise a major social-control function" (1980, p.
79). Whether by choice or circumstance, the authors observed that the
selective pattern of media attention paid to issues reinforces certain
social norms and values. They theorized that local newspapers serve a
control function through two types of information control. In its feedback
information control function, the local press facilitates conflict
resolution by increasing problem awareness, defining the issue and
encouraging public discourse, even as it defines the accepted boundaries of
that debate. This is not an independent function, but is influenced or
controlled by members of the community power structure who often use public
reaction to gauge acceptance, refine their position, and call for action
(Tichenor, Donohue & Olien, 1980, p. 85). The distribution information
control function serves to maintain the community social system by
disseminating general information for routine consumption with no
expectation of active participation by readers. Hindman (1996) observed
that editorial decisions not to cover an issue also served a distributive
information control function.
Studies have found that newspapers serving larger and more complex
communities carry more conflict coverage than those serving smaller, more
homogeneous communities (Hindman, 1996; Tichenor, Donohue & Olien, 1970,
1980). Hindman (1996) also found that conflict is most frequently covered
in local mass media when it has been initiated or endorsed by community
leaders. Tichenor, Donohue and Olien (1980) found coverage reflected the
views of "the dominant group based on the structure and distribution of
social power in the community" (p. 218). They also noted that in the
American experience, this trend means that local media generally reflect
the outlook of the business community.
Community structural pluralism. Tichenor, Donohue, and Olien (1980)
identified the degree of community pluralism as a useful predictor of the
degree of social conflict within a community and the extent of conflict
coverage in its newspaper. Conflict is accommodated in larger, more
pluralistic communities through a variety of formal "pressure valve
mechanisms" such as public hearings, grievance procedures and citizen
review boards, all of which receive press coverage (Hindman, 1996). In
smaller communities, open and public expressions of differences may disrupt
the normal functioning of the community and, as a result, the mention of
such differences is avoided (Tichenor, Donohue & Olien, 1980).
The social-control function of the community newspaper was found to depend
largely upon the social structure of the community served by the
newspaper. Newspapers serving less pluralist communities were found more
likely to focus on the distributive function, while newspapers in more
complex communities focused a greater portion of coverage on the feedback
function (Donohue, Tichenor, & Olien, 1973; Tichenor, Donohue & Olien,
1970, 1980; Viswanath & Arora, 2000).
Changes in community and mass media. Tichenor, Donohue, and Olien
(1980) expressed concern regarding the external validity of their findings
over time and in the face of social change. They expressed particular
concern about emerging social trends, noting that the autonomy of rural
communities was likely to weaken in the face of increasing influence from
external institutions and agencies. The growing power of state and federal
government and of large outside businesses to impact on community revenue,
regulation and employment threatened the status quo with regard to the role
of local elites. At the same time, increasing ownership of local
newspapers by external media corporations, trends toward reliance on
alternative media sources for news and information, and the promise of new
media technologies all suggested that the role of community newspapers and
the local publisher would undergo significant change. These changes
forecast increased pluralism even in smaller communities, resulting in a
weakening of social control by local elites and of the information control
function exercised by community newspapers.
Hypothesis one addressed the opening presentation of the issue in
newspaper coverage and provided a preliminary measure of the social-control
function of the local newspaper in terms of initial access to information
and evidence of distributive control versus feedback control.
H1: Initial news coverage introducing the prospect of the community seeking
consideration as a prison site will present the initiative as a decision
taken by community leaders.
To test this hypothesis, the first day's coverage of the issue in each
newspaper was identified and the content evaluated in terms of whether the
issue was presented as a decision already taken or one yet to be
decided. This is consistent with the findings of Tichenor, Donohue and
Olien (1980; Donohue, Olien & Tichenor, 1985) that portray the local
newspaper as primarily serving a distributive information function in such
circumstances after local elites arrive at consensus among themselves
Previous studies have found that in smaller, more homogeneous
communities, the local newspaper presents the majority of coverage in
potentially contentious issues as routine information that is provided to
inform readers, portraying the issue in terms of consensus in which no
response is expected and reflecting a continuation of the distributive
H2: News and editorial coverage of the prison-siting issue by the
community newspaper will frame the issue in terms of community consensus
more frequently than in terms of community conflict.
As in the previous hypothesis, each paragraph of coverage was analyzed to
determine whether the prison-siting issue was described in the context of
community consensus or conflict.
The third hypothesis employed community structural pluralism indices
identified by Hindman, Ernst, and Richardson (2001) to test relationships
between community structure and the newspaper's social-control function.
H3: Newspapers serving communities having greater structural pluralism
will provide a higher proportion of conflict coverage than communities with
lesser structural pluralism.
Conflict content and the direction of coverage were analyzed to determine
the strength of correlation with community structural pluralism.
Finally, Tichenor, Donohue, and Olien (1980) expressed concern
regarding the external validity of their findings with regard to their
applicability over time and in the face of social change. This leads to
the last hypothesis, which exploits the quarter century of local news
coverage in this study to compare conflict coverage across this period to
determine whether a significant change in patterns of coverage could be
H4: More recent community newspaper coverage of the prison site-selection
issue will reflect a greater proportion of conflict coverage than coverage
published earlier in the period under study.
Once again, conflict content and the direction of coverage were examined to
determine the strength of correlation with the dates of publication.
Sample. The sample for this study included 24 weekly, semi-weekly and
small daily newspapers serving Illinois communities that had successfully
sought selection as the site of a state correctional facility constructed
between 1980 and 2001. Local news coverage of the prison site-selection
process began in 1977. Communities ranged in population from less than 500
to more than 88,000 residents.
Of the newspapers serving these communities, 16 were dailies while
eight were published less frequently. Five communities did not have a
newspaper published within the community itself so the newspaper published
closest to the prison site was used. Newspaper circulation totals varied
from the largest with a daily circulation of 51,920 down to the smallest
with a weekly circulation of 2,150. The mean circulation figure for all
newspapers used in this study was 12,697 copies. The composition and size
of the newspapers and the characteristics of the communities they serve fit
the criteria developed by Byerly (1961) in defining community
newspapers. (See appendix A for a list of communities and newspapers used.)
A census was conducted of all news and editorial coverage directly
related to the prison site-selection issue. The period of coverage was
defined as beginning with the first article reporting the state's
initiation of a selection cycle or the first article suggesting regional or
community interest in seeking selection. The period continued through
official announcement of the site selection and for two weeks thereafter.
The period following the selection announcement was included to account for
follow-up activities by the community and state as well as to include
coverage of public reaction following the decision.
Content Analysis. The sampling units for the study were individual
issues of the community newspaper published during the defined period for
that community. The primary source of newspaper articles were microfilmed
copies accessible through the Illinois State Newspaper Archives. All
articles and other items of newspaper content referring to the prison issue
were identified. Articles reporting the meetings of local government,
business and civic groups were reviewed for individual paragraphs related
to the prison proposal even if the primary focus of the article addressed
other, unrelated issues or items of business. Only the body text of news,
editorials (including op-ed pieces and columns) and letters to the editor
were coded. Letters to the editor were included for content analysis
because they represent content subject to editorial policy including the
decision to publish and to edit letters for length or content. A total of
1,019 items were identified and copied. Articles found to report on
corrections issues but without specific reference to the community in
question were eliminated, resulting in a final sample of 897 articles
included in the content analysis.
For the purposes of this study, individual paragraphs were used as
the unit of analysis. This was considered appropriate for thematic
analysis related to the direction of content and depiction as consensus or
conflict. Coders counted individual paragraphs and analyzed each as a
separate unit of meaning in accordance with the various categories and
rules provided in the coding protocols.
Thematic analysis was conducted for the analysis of the direction of
content and portrayal of the issue as conflict or consensus. Holsti (1969)
emphasized the importance of an identifiable subject in defining themes and
for this study that subject was defined as the prison site-selection issue
in the context of the community being evaluated. The first step in the
coding process, then, was to determine whether the paragraph being
evaluated referred to the prison issue and whether the reference was
related to the specific community in question. Both Krippendorf and Holsti
recognized the usefulness of thematic units because these address the
intent of the communicator, however, both also raise concerns regarding
issues of reliability in coding these units (Riffe, Lacy & Fico, 1998).
Coding. Definitions and coding categories for each stage of the
content analysis were developed from the related literature. Coding
categories, rules and procedures were tested and refined over several
months of training and coding sessions to improve reliability. Throughout
this initial process, coders worked together to review and clarify coding
categories and protocols. Coding sheets were prepared and modified to
support the protocols. These sheets required the coder to verify
identification and content information for the articles including the type
of content, date of publication, and number of paragraphs.
Initial Status. The first day's coverage of the prison-siting
proposal in the community newspaper was analyzed to address hypothesis one
and to determine whether the issue was presented to the public as a
decision already made or as a proposal still open to public discourse and
debate. Coding for this hypothesis was conducted separately from the rest
of the study and used only articles published in each newspaper's first day
of coverage addressing the community's entry into the site-selection
process. First day's coverage was defined as the first article to appear
in each community newspaper that addressed the prison issue specifically in
relation to the community in question.
Each paragraph of the articles was analyzed regarding how the prison
issue was presented. Coding categories consisted of the following:
• Decided content presented the prison issue as a decision already made or
action already initiated by local governments, community leaders or
• Undecided content presented the issue as an option to be considered or as
an issue yet to be decided before action is taken. Examples included
calls for additional information, plans for local polls or hearings to seek
• Neutral content was defined as paragraphs that referred to the issue but
did not address the status of the decision or for which a status could not
be identified. These provided background information without addressing
the status of the issue.
• Unrelated content was defined as those paragraphs included in an article
that were not in any way related to the community in question or to the
Direction. All coverage was coded to evaluate the direction of the
paragraph's content relevant to prison site-selection. Coders evaluated
whether content portrayed prison facilities and the decision to seek
site-selection in a favorable or unfavorable light. This required coders
to analyze the paragraph's content to determine whether (1) prisons in
general, (2) the prospect of attaining a prison facility in the vicinity or
(3) the impact of a prison is portrayed as positive or negative outcomes.
If the information presented in the paragraph was worded in neutral terms,
used a mix of both positive and negative references, or if there was no
clear position presented in the paragraph, the unit was coded as neutral.
Conflict versus consensus. Coders reviewed each paragraph to
determine whether it presented the issue in the context of internal
conflict or general consensus. For conflict content, definition and
categories were developed from that used by Donohue, Olien and Tichenor
(1985; Olien, Donohue & Tichenor, 1968). "Local conflict was coded as
(coverage) devoted to reporting about manifestly differing positions or
statements about a public issue from at least two persons, factions or
interest groups in the community. The question was not whether an article
reported about a subject of controversy, but whether that article made
clear that a controversy was involved"(1985, p. 492). Examples of conflict
coverage included reports of public disagreements, demands for additional
information or reconsideration, and concerns or objections at any level of
community including individual objections raised in letters to the editor.
No similar definition was available for consensus content. For this
study, community consensus content was defined as content that portrayed
the majority of the community in general agreement on the issue, that
characterized actions as being taken in response to a sense of broad
community support, that made a direct appeal for public unity and support,
or that conveyed such appeals from community organizations, officials or
If the paragraph addressed the prison issue but could be coded as
neither consensus nor conflict, it was coded as neutral
content. Paragraphs that addressed the prison issue in the context of the
experiences or bids of other communities or that recounted reactions of the
community in a previous selection cycle were also coded as
neutral. Paragraphs coded as unrelated to the site-selection issue were
not coded for conflict/consensus.
Coder training and reliability. A primary coder was responsible for
coding all articles with a secondary coder analyzing a random sample of
content to measure intercoder reliability. Coders underwent an extensive
training process, beginning with a detailed review and discussion of the
coding protocols including procedures, categories, and definitions. Once a
consistent application of the rules and procedures was developed and
desired intercoder reliability levels were achieved, the primary coder
completed the coding of all content while the secondary coder analyzed the
A random sample of the content was analyzed to assess intercoder and
intracoder reliability. Lacy and Riffe (1996) suggest that the size of
reliability samples should be calculated to provide valid reliability
estimates and a known degree of confidence. Their formula was used to
calculate the appropriate size of the reliability sub-sample, resulting in
a total of 125 articles randomly selected from the 897 articles included in
the study, well within the 10 to 25 percent range suggested by Wimmer and
Dominick (2003). Both intercoder and intracoder reliability was well above
the 75 percent level established prior to the study with direction of
content for intercoder reliability the only category falling below the 80
percent mark. (See appendix B.)
Community structural pluralism indices. To address hypotheses three,
a measure of community pluralism was developed using the multiple measure
community pluralism indices employed by Hindman, Ernst and Richardson
(2001). Using data obtained from the 1990 U.S. Department of Commerce
census, this analysis used city and county populations; percentage of
county workforce not involved in agricultural, forestry or fishing
occupations; and the percentage of county residents who have obtained a
bachelor's degree. The 1990 U.S. Census data represent the approximate
mid-point of the period being analyzed and were viewed as the most
appropriate to the focus of this study. (A listing of community
structural pluralism indices is provided in appendix C.)
Review of the 24 community newspapers included in the study yielded a
total of 1,019 items. Of these, 897 articles (88%) met the study's
criteria for inclusion in the content analysis. Individual paragraphs were
used as the unit of analysis with 10,996 examined. Of that total, 1,761
paragraphs (16%) were coded as content unrelated to the prison issue. (See
appendix D for distribution of content.)
H1: Initial coverage. A total of 26 articles were identified as
meeting the study's definition of opening-day coverage. Of those, 24 were
locally prepared news stories and two were editorials. Many of the news
stories reported discussion, votes or actions taken by local government
bodies in the context of routine business resulting in a total of 137
paragraphs (30.4%) found to be unrelated to the prison issue. Of the
remaining content, more than half (54.3%) were evaluated as neutral in
terms of the status of the decision to pursue selection as a prison site at
the community level. These neutral paragraphs did relate to the prison
issue, providing information and context, but did not address the status of
a decision on the issue.
Those paragraphs that did address the status of the decision at the
community level supported the hypothesis. Paragraphs coded as decided
units totaled 121 (38.7% of content related to the issue) while only 22
undecided units were identified (7.0% of relevant units). Paragraphs that
framed the issue as decided occurred more than five times as frequently as
those suggesting that a decision remained to be made or that actively
sought public input on the issue. (See appendix E.) In light of this
imbalance, hypothesis one is accepted.
This pattern of coverage is consistent with the observations of
Tichenor, Donohue and Olien (1980), who found that newspapers seldom led
community discourse but that coverage was begun after the initial decision
had been made. The resulting coverage reports these as routine decisions
and presents them as a focus of general community consensus. The bulk of
the relevant content presented was neutral and clearly served a routine
distributive information role but the great majority of the information
that did address the status of the decision treated the issue as routine
distributive information as well. Only seven percent of the relevant
content was presented as pertaining to an undecided issue, suggesting the
need for public response or feedback.
Though overall coverage was predominantly decided, five of the 24
communities presented more undecided units of content in their initial
coverage than decided content. Three of these involved state proposals to
convert state mental centers to prisons. In all three cases, the mental
health center was long established, a major employer and closure was
opposed by significant segments of the public. In these cases, community
consensus had not been reached and the timing of the state initiatives and
announcements preempted the local decision-making process. Two other
community newspapers presented coverage that was predominantly
undecided. The first had sought a prison in an earlier cycle but was not
selected, in part because of local opposition. This past opposition was
noted in the story and provided context for the mayor and city council's
apparent indecision at this early stage in the process. The second had not
previously sought a prison in past selection cycles, but the article
suggested that the question had met with a negative response when
considered during earlier selection cycles. In both of these cases, the
newspaper performed a feedback control function in which community leaders
used the initial article to gauge community response and solicit public
feedback. Thus, these exceptions are consistent with the social control
models and predicted patterns as reported by Tichenor, Donohue and Olien
H2: Conflict versus consensus. Nearly two-thirds (65.7%) of the coded
units were found to be neutral on the question of conflict versus
consensus. This was to be expected as coding of conflict and consensus was
specifically tied to the community in question and to the current prison
site-selection cycle. Paragraphs that defined terms, that described the
process and bids by other communities, or that provided history and context
surrounding the issue are all examples of the types of content that would
typically be coded as neutral. The definition applied in this element of
coding also limited the definition of consensus to references that
encompassed the community or a broad segment of it. Statements that
expressed the views of a single individual or a small group may have been
coded as positive in direction but were coded as neutral in terms of
consensus. Individual statements of opposition and those that addressed
concerns of a small group, however, were coded as conflict. This
distinction was considered an appropriate interpretation of the definition
of community conflict and consensus as used in this study and it served to
enhance the reliability of the study. This distinction may raise concerns
regarding potential coding bias but the impact of such a bias, if present,
results in a more conservative test of hypothesis three and reduced the
likelihood of finding support for it.
Of the content (3,167 units) not coded as neutral, those units coded
as community consensus were almost twice as likely to occur. Consensus
paragraphs accounted for 22.0 percent of overall content (2,036 units)
compared to 12.3 percent (1,131 units) for conflict coverage. (See
appendix F.) In light of this relationship, hypothesis three is accepted.
Further analysis of content by type offers additional insight into
how the issue is portrayed in terms of conflict and consensus and supports
the validity of this element of the study. Local news stories approximated
the overall pattern of coverage with a slight increase in neutral content
and were more than twice as likely to cover the issue as a matter of
consensus rather then conflict. News wire coverage provided a
significantly higher percentage of neutral coverage but still portrayed the
issue as a point of consensus almost twice as often. Editorials largely
took the form of endorsements or congratulations to the community after
selection. Consistent with that observation, editorials were found to
convey a consensus focus six times as often as conflict. The only category
of content that exhibited more conflict than consensus surrounding the
prison issue was letters to the editor. (See appendix G.)
The pattern of coverage with respect to consensus is consistent with
previous research that found that local coverage downplayed contentious
aspects of an issue to portray it as a focal point for community
consensus. Tichenor, Donohue and Olien (1980) related this, once again, to
the local newspaper's distributive function.
H3: Structural pluralism. Previous studies found a relationship
between community structural pluralism and the extent of conflict
coverage. In testing this hypothesis, the community structural pluralism
indices were applied and the relationship was tested against the percentage
of content coded as related to conflict coverage. The direction of content
was also used in this analysis as an indication of content diversity. (See
appendix H.) While direction and consensus are separate issues, both
relate to the context in which the newspapers presented the issue to their
The analysis provided mixed results. (See appendix I.) Moderate
positive correlations were found for both negative direction and conflict
coverage. For direction, significant correlations were identified for all
three population indices, with the county figure yielding the strongest
relationship. The percentage of those employed in non-agrarian jobs also
indicated a moderate positive correlation. Use of the percentage of county
residents holding a bachelor's degree did not prove a useful
indicator. These correlations suggest that newspapers in communities
having a higher degree of pluralism were more likely to present a balanced
view of the prison issue in terms of positive versus negative
direction. The coefficient of determination (r2 values) for the
relationship between direction of coverage and community structural
pluralism indices ranged from 0.16 to 0.44, depending on the indicator
The same pattern held for conflict coverage, though the correlations
were slightly weaker and only held for two pluralism indicators. (See
appendix I.) Moderate positive correlations were indicated in the case of
county and newspaper community populations. While a correlation was
indicated between the prison community population and proportion of
conflict coverage, the correlation was weak and failed to achieve the p <
.05 level of significance.
The remaining pluralism indicators for levels of education and
employment also failed to indicate a significant relationship. The lack of
correlation across these three indices may be explained, in part by the
limited variance in the sample. The correlations identified in this
portion of the analysis suggest that newspapers in more pluralistic
communities, at least those with larger populations, were more likely to
provide balanced coverage of the prison issue in terms of community
conflict and consensus. Significant correlations for conflict coverage
yielded r2 values in the 0.20 to 0.33 range.
Although the results suggested by the various pluralism indices were
mixed, there was sufficient evidence to support acceptance of hypothesis three.
H4: Conflict coverage over time. The prediction of increased
conflict coverage is intended to test past findings for their external
validity over time. The assumption was that social and mass media changes
would result in an increasing proportion of conflict coverage in more
recent coverage. As with hypothesis three, both the proportions of
conflict units and negative direction units to overall coverage were
The data did not support hypothesis three and it was rejected. No
correlation was indicated for either conflict coverage or negative
direction, suggesting that there has been little change in either aspect of
coverage over time (See appendix J.) This suggests that the concerns of
Tichenor, Donohue and Olien have not been realized, at least in the context
of the communities and prison issue addressed in this study. A plausible
explanation may well be that the trends causing concern had not reached
those communities or that the period of time has not resulted in measurable
Discussion. As a replication of prior research by Tichenor, Donohue
and Olien, this study tested their earlier findings regarding the role of
local newspapers in community conflict. The results of the study provided
strong evidence in validation of past findings with regard to the role of
the community newspaper in conflict situations. In the first hypothesis
tested, content analysis of initial coverage of the prison issue at the
community level provided clear support for previous observations that
newspapers serve a largely distributive information role, reporting or
endorsing decisions already taken by local elites. Several communities
were found to be exceptions to this pattern but provided additional support
in their consistency with the underlying concepts of Tichenor, Donohue and
Consistent with previous findings of the Minnesota research team,
patterns of newspaper coverage analyzed in this study found newspapers more
likely to portray the prison issue in a positive light and as a focus of
community consensus. Furthermore, the newspapers continued to serve a
distributive information role far beyond initial coverage, extending this
maintenance function through the entire prison site-selection process.
Also consistent with the conclusions of Tichenor, Donohue and Olien
were the moderate correlations found between content diversity and
community structural pluralism. There was evidence that newspapers serving
more pluralistic communities were more likely to present greater content
diversity and more conflict coverage with local population figures serving
as the most consistent pluralism indicator. This relationship can be
explained in the context of a mixed community power model in which the
dominant power structure in more complex communities splits into multiple
competing power centers resulting in greater diversity of opinion and a
higher likelihood of conflict (Schulze, 1957; Woods, 1998).
The final hypothesis of this study addressed the concerns of
Tichenor, Donohue and Olien (1980) that changes in community identity and
autonomy coupled with alterations in the mass media industry would modify
the patterns they had previously identified in local news and editorial
content, including an increase in conflict coverage and content diversity
over time. No significant correlation between the year of publication and
the levels of conflict coverage and content diversity was found to support
Limitations. Although this study was ambitious in its attempt to
conduct a census of coverage related to the prison issue, the resulting
sample remains purposive. Only communities successful in their bid for
site selection were analyzed, resulting in an identifiable sample that fit
the purpose of the study. While the findings of this study are consistent
with previous findings, the external validity of the study is weakened by
the sample selection and one should be cautious in generalizing the results
of this study to other samples or issues.
A concern common to all studies using content analysis is the
potential for coding categories to reflect the researcher's biases. To
address this, particular care was taken to incorporate definitions from
previous research where available. Where areas of potential bias were
identified, the more conservative definition or coding procedure was
applied. These efforts to ensure reliability while strengthening internal
validity introduce a threat to external validity. Riffe, Lacy and Fico
(1998) raised this potential weakness of content analysis, noting that the
meanings derived from content by trained coders formally evaluating content
may be far different from those developed by the typical reader.
Future research. Several avenues for future research are suggested
by the limitations sited previously. Further replications using other
samples and issues are appropriate to build support and further define the
model of local newspapers and their social-control function in communities.
Implications. This study leads to the conclusion that the community
newspaper is an agent of social control and that it serves the interests of
those in power. This validation may be reassuring from the theoretical
perspective of validating Tichenor, Donohue and Olien's previous findings
regarding the social control function of the community newspapers, but this
provides further evidence undermining our traditional ideal of an
independent local press.
Altschull, J. H. (1995). Agents of power: The media and public policy
(2nd Ed.). White
Plains, NY: Longman Publishers.
Banfield, E. C., & Wilson, J. Q. (1963). City politics. Cambridge,
MA: Harvard University Press and MIT Press.
Breed, W. (1955, May). Social control in the newsroom: A functional
analysis. Social Forces, 33 (4), 326-335.
Breed, W. (1958, December). Mass communication and social-cultural
integration. Social Forces, 37 (2), 109-116.
Byerly, K. (1961). Community journalism. Philadelphia: Chilton Publishers.
Carey, J. W. (1979). Foreword. In H. Hardt (Ed.), Social theories of the
press: Early German and American perspectives. Beverly Hills,CA: SAGE
Carlson, K. A. (1988). Understanding community opposition to prison
siting: more than fear and finances. Corrections Today, 50 (4), 84-90.
Carlson, K. A. (1990, June). Prison escapes and community consequences:
Results of a case study. Federal Probation, 49 (2), 36-42.
Dahl, R. A. (1961). Who governs? New Haven, CT: Yale University Press.
Dahl, R. A. (1982). Dilemmas of a pluralist democracy. New Haven, CN:
Yale University Press.
Davidson, W. B., & Cotter, P. R. (1997, April). Psychological sense of
community and newspaper readership. Psychological Reports, 80 (2), 659-666.
Dewey, J. (1927). The public and its problems. Denver: Alan Swallow.
Domhoff, G. W. (1998). Who rules America: Power and politics in the year
2000 (3rd ed.). Mountain View, CA: Mayfield Publishing Company.
Donohue, G. A., Olien, C. N., & Tichenor. P. J. (1985, Fall). Reporting
conflict by pluralism, newspaper type and ownership. Journalism Quarterly,
Donohue, G. A., Tichenor. P. J., & Olien, C. N. (1973, Winter). Mass
media functions, knowledge and social control. Journalism Quarterly, 50
Donohue, G. A., Tichenor, P. J., & Olien, C. N. (1995, Spring). A guard
dog perspective of the media. Journal of Communication, 45 (2), 115-132.
Edelstein, A. S., & Larson, O. N. (1960, Autumn). The weekly press'
contribution to a sense of urban community. Journalism Quarterly, 37 (4),
Edelstein, A. S., & Schulz, B. (1963). The weekly newspaper's leadership
role as seen by community leaders. Journalism Quarterly, 40 (3), 565-574.
Fitzgerald, M. (1996, Nov. 23). Building newspapers and communities: 1996
National Newspaper Association. Editor & Publisher, 129 (47), 26-27.
Galbraith, J. K. (1983). The anatomy of power. Boston: Houghton Mifflin.
Gamson, W. A. (1968). Power and discontent. Homewood, IL.: Dorsey Press.
Gaziano, C. & McGrath, C. (1987, Summer-Autumn). Newspaper credibility and
relationships of newspaper journalists to communities. Journalism
Quarterly, 64(2-3), 317-328, +345.
Gibbons, S. G. & Pierce, G. L. (1995). Politics and prison development in a
rural area. The Prison Journal, (75), 3, 380-389.
Hallinan, J. T. (2001). Going up the river: Travels in a prison
nation. New York: Random House.
Hindman, D. B. (1996, Autumn). Community newspapers, community structural
pluralism, and local conflict with nonlocal groups. Journalism and Mass
Communication Quarterly, 73(3), 708-721.
Hindman, D. B., Ernst, S., & Richardson, M. (2001). The rural-urban gap
in community newspaper editors' use of information technologies. Mass
Communication and Society, 4 (2), 149-164.
Holsti, O. R. (1969). Content analysis for the social sciences and
humanities. Reading, MA: Addison-Wesley Publishing Co.
Illinois' prison population falls as nation's rises. (2003, July
28). Associated Press.
Janowitz, M. (1967). The community press in an urban setting. Chicago:
The University of Chicago Press.
Janowitz, M. (1975). Sociological theory and social control. American
Journal of Sociology. 81 (1), 82-106.
Janowitz, M. (1978). The intellectual history of "social control." In J.
S. Roucek (Ed.), Social Control for the 1980's: A Handbook for Order in a
Democratic Society (pp. 20-45). Westport, CT: Greenwood Press.
Janowitz, M. (1991). On social organization and social control. Chicago:
University of Chicago Press.
Kaniss, P. (1991). Making local news. Chicago: University of Chicago
Kennedy, B. M. (1974). Community journalism: A way of life. Iowa State
University Press: Ames.
Krause, J. D. (1992). The effects of prison siting practices on community
status arrangements: a framework applied to the siting of California
prisons. Crime & Delinquency, 38 (1), 27-55.
Lacy, S., & Riffe, D. (1996, Winter). Sampling error and selecting
intercoder reliability samples for nominal content categories. Journalism
& Mass Communication Quarterly, 73 (4), 963-973.
Lasswell, H. D. (1949). The structure and function of communication in
society. In W. Schramm, (Ed.), Mass Communications (pp.
102-115). Urbana: University of Illinois Press.
Martindale, D. (1978). The theory of social control. In J. S. Roucek
(Ed.). Social Control for the 1980's: A Handbook for Order in a Democratic
Society (pp. 46-58). Westport, CT: Greenwood Press.
McCombs, M. (1997). Building consensus: the news media's agenda-setting
roles. Political Communication, 14, 433-443.
McCombs, M., & Shaw, D. (1972). The agenda-setting function of mass
media. Public Opinion Quarterly, 36, 176-185.
McFarland, R. (1999). Bacon's newspaper directory 2000 (48th
ed.). Chicago: Bacon's Information, Inc.
Merton, R. K. (1949). Social theory and social structure. Glencoe, IL:
The Free Press.
Mott, P. E. (1970). Power, authority and influence. In M.T. Aiken & P.E.
Mott, (Eds.). The structure of community power (pp. 3-16). New York:
Olien, C. N., Donohue, G. A., & Tichenor. P. J. (1968, Summer). The
community editor's power and the reporting of conflict. Journalism
Quarterly, 45 (2), 243-252.
Park, R. E. (1925). The natural history of the newspaper. In R.E. Park,
E. W. Burgess & R. D. McKenzie, (Eds.). The city (pp. 8-22). Chicago:
University of Chicago Press.
Park, R. E. (1929). Urbanization as measured by newspaper
circulation. American Journal of Sociology. 25, 60-79.
Popper, F. J. (1981, April). Siting LULUs. Planning, 4, 12-15.
Riffe, D., Lacy, S., & Fico, F. G. (1998). Analyzing media messages: Using
quantitative content analysis in research. Mahwah, NJ: Lawrence Erlbaum
Roucek, J. S. (Ed.). (1978). Social control for the 1980's: A handbook for
order in a democratic society. Westport, CT: Greenwood Press.
Schlosser, E. (1998, December). The prison-industrial complex. The
Atlantic Monthly, 51-77.
Schulze, R. O. (1957). The role of economic dominants in community power
structure. In M. Aiken & P. E. Mott (Eds.), The structure of community
power (pp. 31-45). New York: Random House.
Shichor, D. (1992). Myths and realities in prison siting. Crime &
Delinquency, 38 (1), 70-87.
Shoemaker, P., & Reese, S. (1996). Mediating the message: Theories of
influence on mass media content. New York: Longman.
Sneed, D., & Riffe, D. (1991). The publisher—public official: Real or
imagined conflict of interest? New York: Preaeger.Spies, S., Murdock, S.
H., White, S., Krannich, R., Wulfhorst, J. D., Wrigley, K., Leistritz, F.
L., Sell, R., & Thompson, J. (1998, March). Support for waste facility
siting: Differences between community leaders and residents. Rural
Sociology, 63 (1), 65-93.
Stone, C. N. (1989). Regime politics: Governing Atlanta,
1946-1988. Lawrence: University Press of Kansas.
Sumner, C. (1997). Social control: The history and politics of a central
concept in Anglo-American sociology. In R. Bergalli & C. Sumner
(Eds.). Social control and political order : European perspectives at the
end of the century (pp. 1-33). Thousand Oaks, CA : Sage.
Takahashi, L. M., & Graber, S. L. (1998). Controversial facility siting
in the urban environment: Resident and planner perceptions in the United
States. Environment and Behavior. 30 (2), 184-216.
Tichenor, P. J., Donohue, G. A., & Olien, C. N. (1970). Mass media flow
and differential growth in knowledge. Public Opinion Quarterly, 34, 159-170.
Tichenor, P. J., Donohue, G. A., & Olien, C. N. (1980). Community
conflict and the press. Beverly Hills, CA: SAGE Publications, Inc.
Tuchman, G. (1978). Making news: a study in the construction of
reality. New York: The Free Press.
U.S. Department of Commerce (1992, June). 1990 Census of Population and
Social and Economic Characteristics for Illinois (1990 CPH-2-15, Section 1
of 2). Washington, D.C.: U.S. Government Printing.
U.S. Department of Commerce (1993, September). 1990 Census of Population
and Housing: Summary of Social, Economic, and Housing Characteristics for
Illinois (1990 CPH-5-15). Washington, D.C.: U.S. Government Printing.
Viswanath, K., & Arora, P. (2000, Winter). Ethnic media in the United
States: An essay on their role in integration, assimilation, and social
control. Mass Communication & Society, 3 (1), 39-56.
Warren, R. L. (1978). The community in America (3rd ed.). Lanham, MD:
University Press of America.
Webster, J. (1987, Fall/Winter). The publisher and civic activity: Civic
activism dilemma. Journal of Mass Media Ethics, 2 (1), 41-47.
Wimmer, R. D., & Dominick, J. R. (2003). Mass Media Research: An
Introduction (7th ed.). Belmont, CA: Wadsworth Publishing.
Woods, M. (1998, December). Rethinking elites: Networks, space and local
politics. Environment and Planning, 30 (12), 2101-2119.
Community Pop. Title Circulation
Canton 13,922 Daily Ledger (e) 5,706
Centralia 14,274 Centralia Sentinel (e)14,145 (S) 14,145
Danville 33,828 Commercial-News (e)18,695 (S) 20,345
Decatur 88,885 Herald & Review (m)38,035 (S) 46,947
Dixon 15,144 The Telegraph (e) 9,100
East Moline 20,147 The Dispatch (e) 27,870 (S) 34,673
East St. Louis 40,944 News-Democrat (m)51,920 (S) 62,279
Galesburg 33,530 Register-Mail (e) 17,300 (S) 16,800
Grayville 2,043 The Carmi Times (e) 2,970
Harrisburg 9,289 Daily Register (e) 4,808
Hopkins Park 601 Daily Journal (e) 26,961 (S) 32,008
Jacksonville 19,324 Journal-Courier (m)14,597 (S) 14,403
Kewanee 12,969 Star Courier (e) 6,365
Ina 489 Register-News (e) 10,500
(Mount Vernon) (16,988)
Robinson 6,740 Robinson Daily News (e) 6,528
Taylorville 11,133 Breeze Courier (e) 6,700 (S) 6,700
Community Newspapers –Weeklies/Semiweeklies
Community Pop. Title Circulation
Hillsboro 4,400 Hillsboro Journal (M/Th) 6,164
Mount Sterling 1,922 Brown County (Tu) 2,700
Murphysboro 9,176 Murphysboro American (M/Th) 11,565
Pinckneyville 3,372 Democrat (W/Th) 3,000
Rushville 3,212 The Rushville Times (W) 3,500
Sumner 1,083 The Sumner Press (W) 2,150
Savanna 3,819 Northwestern Ill. Dispatch (W) 8,250
Tamms 750 Gazette Democrat (Th) 5,200
Percent intercoder agreement
Percent intracoder agreement
Community Structural Pluralism Indices
% with bachelors degrees
State of Illinois
East St. Louis
1 % Non-Ag.
% with bachelors degrees
Notes: 1 % of non-agrarian workforce is defined as the percentage of the
workforce not employed in agriculture, forestry or fisheries positions.
2 Communities involve two counties in the site-selection
values were used as indices for these communities. Population figure
average of two county population figures. Workforce and education figures
represent weighted averages of combined counties.
(U.S. Department of Commerce, 1992,1993)
Distribution of Data by Type of Content
Types of content
Letters to the editor
Op-Ed pieces/ columns
Distribution of Initial Coverage by Decision Status
Total units of analysis
a The term, relevant units, refers to the total number of coded
paragraphs (450) minus
those paragraphs coded as unrelated (137). Relevant units totaled
Distribution of Content by Conflict Versus Consensus
Conflict Versus Consensus by Content Type
Letters to the editor
Distribution of Content by Direction
Correlation Matrix for Pluralism Indices and Proportion of Conflict Coverage
Media community population
Percent with bachelor's degree
Percent employed in non-agriculture/forestry/fishing
* p < .05
Correlation Matrix for Direction of Coverage
and Proportion of Conflict Coverage by Year
Year of coverage
Note: Statistical significance not indicated. No p values < .05.
 Copies of the coding protocols and the coding sheets used in the
study are available from the author on request.