Content-Type: text/html Civic Journalism in the 2000 U.S. Senate Race in Virginia Civic Journalism in the 2000 U.S. Senate Race in Virginia by David Kennamer and Jeff South School of Mass Communications Virginia Commonwealth University Richmond, VA 23284-2034 Paper submitted to the Civic Journalism Interest Group 2001 AEJMC Convention Washington, DC Comments or questions should be addressed to the first author at (804) 828-2660 or [log in to unmask] Civic Journalism in the 2000 U.S. Senate Race in Virginia ABSTRACT Civic Journalism in the 2000 U.S. Senate Race in Virginia By David Kennamer and Jeff South Virginia Commonwealth University School of Mass Communications Richmond, VA 23284 (804) 828-2660 The Norfolk Virginian-Pilot is a proponent of civic journalism; the Richmond Times-Dispatch is not. Content analysis of the papers' coverage of Virginia's 2000 U.S. Senate election reflected the divergent newsroom philosophies. The Times-Dispatch stories were more likely to be triggered by campaign-managed events, to focus on the election "horse race" and to use political establishment sources. The Pilot's stories were more likely to result from independent or enterprise reporting, to address issues and to use "real people" sources. Civic Journalism in the 2000 U.S. Senate Race in Virginia Civic Journalism in the 2000 U.S. Senate Race in Virginia Objective This paper compares coverage of last fall's U.S. Senate race in Virginia by the Richmond Times-Dispatch and Norfolk-based Virginian-Pilot to see if their divergent journalism philosophies made a difference in their news stories. The Virginian-Pilot is a well-known proponent of "public" or "civic" journalism,[1] a relatively new and controversial approach to covering news, while the Times-Dispatch adheres to a traditional approach to journalism. The Senate race presents a unique opportunity to see how these representatives of different strains of journalism cover the same important, complex and high-profile event. The Philosophical Basis for Civic Journalism In the late 1980s, several editors, commentators and media critics began to question the traditional values and practices that have guided print journalism in the United States for at least the past half century. They argued that the ritualistic adherence to values embodied in the journalistic concept of "objectivity" has resulted in journalism that is largely reactive; driven by events; dominated by conflict, crisis and scandal; oriented toward political, social and economic elites; and increasingly distasteful and irrelevant to large numbers of the public. These critics said such journalism has turned off both readers and voters: that it has been responsible for the decline in newspaper circulation over the past several decades and, more importantly, for the apparent alienation of increasing numbers of the public from the American political system. Arthur Charity writes, for example, that veteran journalists "were troubled by the low quality of much of their own work [and] by evidence that the public they had intended to serve distrusted newspapers and increasingly didn't even read them. Most importantly, they saw that the very problems they had come to journalism to help solve still weren't being solved, or even being very intelligently addressed" (1995, p.1). The proposed solution, then, is civic journalism. Jay Rosen has defined it as "an approach to the daily business of the craft that calls on journalists to 1) address people as citizens, potential participants in public affairs, rather than victims or spectators; 2) help the political community act upon, rather than just learn about, its problems; 3) improve the climate of public discussion, rather than simply watch it deteriorate; and 4) help make public life go well, so that it earns its claim on our attention" (Rosen, 1999, p. 22). An intellectual basis for civic journalism is found in the work of Jurgen Habermas and John Dewey. Habermas is a German scholar whose ideas concerning public opinion have become widely infused in a number of social science disciplines in recent decades. He theorizes the development of a "public sphere," a metaphorical "space" in which free public discourse can take place. Communications media developed in this space and served to expand it, feeding democratic impulses and democratic institutions. Thus, communication processes are key to the development and maintenance of the public sphere and to democratic discourse and processes. Proponents of civic journalism argue that in recent years, the news media have not supported this public sphere; rather, they have contributed to its degradation. Civic journalists argue that if this slide is to be reversed, journalism must change its ways to provide the public with information that is more useful, more relevant to public problems, more encouraging of civil and productive discourse, and more inclusive of diverse perspectives. Dewey, a philosopher of democratic ideals of the 1920s and 1930s, said that a fully informed public "would emerge only if politics, culture, education and journalism did their jobs well" (Merritt and Rosen, 1995). Rosen (1999) summarizes Dewey's approach to journalism this way: The newspaper of the future will have to rethink its relationship to all the institutions that nourish public life, from libraries to universities to cafes. It will have to do more than "cover" these institutions when they happen to make news. It will have to do more than print their advertisements. The newspaper must see that its own health is dependent on the health of dozens of other agencies which pull people out of their private worlds. _ Every town board session people attend, every public discussion they join, every PTA event, every local political club, every rally, every gathering of citizens for whatever cause is important to the newspaper - not only as something to cover, but as the kind of event that makes news matter to citizens (p. 20). Practitioners of Civic Journalism Researchers generally consider the Wichita Eagle the first newspaper to engage in civic journalism. Like many journalists, Davis Merritt, the Eagle's editor, was concerned about the low turnout and voter participation in the 1988 presidential election. So in directing the coverage of the 1990 Kansas gubernatorial campaign, Merritt adopted a civic journalism approach: The Eagle used focus groups and public forums to create a citizen agenda to guide election coverage, rather than let it be managed and manipulated by politicians and campaign strategists. To promote public participation, the newspaper invited ordinary people to question the candidates - and even held a voter registration drive in its lobby. Civic journalism gained momentum in 1992 when The Charlotte Observer teamed with the local ABC affiliate, WSOC-TV, to cover that year's elections from the citizens' perspective. In 1993, the Pew Center for Civic Journalism was established to promote this emerging newsroom philosophy. The center gave news organizations grants to launch civic journalism experiments and presented awards for the best projects. (The Batten Awards are named for the late James Batten, who as chairman of Knight-Ridder Inc. was an early proponent of civic journalism.) Competition for the grants and awards reflects growing interest in civic journalism. The number of grants increased from three in 1994 to 16 this year. Overall, the Pew Center has funded more than 100 civic journalism projects involving more than 200 news organizations. (A project typically involves a newspaper partnering with a television station, radio station, online operation or other organization.) Entries in the Batten Awards contest rose from 90 in 1997 to 116 last year. Newspapers of all sizes have done civic journalism. The Philadelphia Inquirer, the nation's sixth largest daily newspaper with a Sunday circulation of about 800,000, shared the Batten Award last year for "Citizen Voices '99," a yearlong "civic dialogue" about the city's mayoral election. The Dallas Morning News (circulation 785,000), San Francisco Chronicle (circulation 570,000) and Seattle Times (circulation 500,000) have received Pew grants to conduct civic journalism experiments. And the Baltimore Sun (circulation 475,000) shared the Batten Award in 1998 for a campaign to improve elementary students' reading skills. But most of the newspapers recognized for practicing civic journalism are much smaller. They include at least 15 papers with circulations between 100,000 and 300,000 - such as the Rochester Democrat and Chronicle, the Dayton Daily News, the Wisconsin State Journal and the Spokane, Wash., Spokesman-Review. More than 30 newspapers with circulations below 100,000 also have received Pew grants or Batten Awards. These publications include the Bradenton Herald in Florida, the Anniston Star in Alabama, the Bronx Journal in New York, the Lewiston Morning Tribune in Idaho and the La Grande Observer in Oregon. Civic journalists have taken on a variety of topics, from downtown development to public safety. In doing so, they have emphasized not just exposing social problems but engaging readers in the search for solutions. For example, the Savannah Morning News in Georgia shared the Batten Award last year for its use of community forums, polls, focus groups and Web interaction for a yearlong series about the community's aging population. In 1998, the Portland Press Herald and Maine Sunday Telegram mobilized study circles on alcohol abuse, and the St. Paul Pioneer Press organized book clubs and discussion groups on poverty and welfare reform. Civic journalism also has found its way into university curricula. Killenberg and Dardenne, for example, have changed the title of their course at the University of South Florida from "Advanced Public Affairs Reporting" to "News Coverage of Public Life." They write: Stories most often found in the daily newspaper or on the evening newscast reinforce prevailing mainstream thinking because they frequently come from traditional, official sources, or they use scattered comments from people with whom journalists have spent too little time. Our students, like many practicing journalists, find it difficult to break away from rounding up the "usual suspects" and getting a quick quote from a bystander. Such conventional journalism is easy and efficient because official sources are accessible, usually comfortable dealing with the press and therefore often quotable, and authoritative and credible, at least to a journalist (1997, 52-53). Critics of Civic Journalism Many civic journalism efforts have met heated resistance from journalism traditionalists. They maintain that civic journalists become far too involved in the issues they are covering; that they participate in and create news, instead of dispassionately observing and chronicling it; and that they place reporters in the position of being community boosters and issue cheerleaders. In short, traditionalists argue that civic journalism advocates have abandoned a cherished value of journalism - objectivity. Journalists of all stripes might agree that American politics and society may be going to hell in a hand basket. But while civic journalists believe they should step in and try to stop the slide, traditionalists say journalism's role is to chronicle the descent as thoroughly and objectively as possible. As a former executive editor of The New York Times and critic of civic journalism wrote, "Leave reforms to the reformers" (Frankel, in Rosen, p. 220). Many critics from the newspaper business also have dismissed civic journalism, ironically, as simply "good, solid journalism." They have argued that the best way to improve journalism is to simply do it better! Thoughtful traditional journalists are not particularly satisfied with the current performance of their colleagues. However, they say this is because journalists have strayed too far from traditional norms and practices, not because they are sticking too closely to them. Robert J. Haiman, in a Newspaper Research Journal commentary titled, "Hey editors: Just stop the nonsense," lists nine "wrongs" with journalism. He writes, for example, "Attention all journalists: Stop using anonymous sources. Just stop it! Stop letting yourself be spun by the spin doctors. Stop letting yourself be used by the leakers. Stop picking up the anonymously-sourced stories of others. Stop the speculation and the unfounded conclusions and the gossiping and the rumor-mongering" (p. 3). He also admonishes journalists, "Stop being celebrities and clowns and capitalists." To publishers, general managers and executives of journalism organizations, he says, "Stop being so mindlessly focused on slashing resources to news departments and so brutally determined to increase profits. Stop paying young reporters such niggardly salaries that some have to live near the poverty line. Stop exploiting broadcast students by offering only unpaid internships" (p. 5). Some critics of civic journalism have persuasively argued that much of what is wrong with modern journalism can be traced directly to this last point - that the new corporate owners of newspapers and broadcast outlets have slashed budgets to the point that newsrooms are chronically understaffed. It follows that an overworked reporter on a deadline may be particularly likely to craft a story quoting the "usual suspects" easily reached by phone rather than head out the door to a civic association meeting, shopping mall, beauty salon or street corner. Rosemary Armao, a former president of Investigative Reporters and Editors Inc. and a self-described "refugee from public journalism," blames any current "disconnect" between journalists and the public on "MBA management types." She adds that if a newsroom is adequately staffed, the reporters will be able to find out what the public is thinking. "We were a part of the community and we were read. And then the MBAs started coming into our newsrooms and they cut the number of reporters, and they cut the amount of space they had in the newspaper. And, whoa, now we're unconnected from our public" (in Corrigan, p. 116). Although civic journalism's "theory" has been widely studied and promoted in academia, it also has come under considerable fire from scholars. They have argued that: 1) Civic journalism is based on unwarranted assumptions about the nature and function of public discourse in democracy. 2) It is unclear if public discourse and the resulting political processes actually are in such precipitous decline. 3) If they are in decline, there are multiple causes, with journalism's alleged misdeeds perhaps not the major one. 4) Civic journalism exhibits a simplistic understanding of the role mainstream journalism has played in the past in encouraging constructive public discourse. Illustrating several of these points, Pauly (1999) notes, "In particular, public journalism continues to work with a truncated account of the origins of journalism's crisis, a dubious sense of the daily newspaper's exceptional role in American public discourse, and an overly rationalized conception of the social relations that democracy requires" (p. 139). In addition, he writes, "To read the literature on public journalism, one would think that Americans always embrace democracy, that the press always supports the common good, and that all that remains is to revive a tradition of solidarity that has only recently disappeared" (p. 145). The underlying purpose of civic journalism, according to Glasser (1999), seems to be procedural: to promote democratic processes while not promoting any particular outcome - i.e., to be a fair-minded referee of sorts. It assumes a basic fairness, civility and adherence to certain values in the public that could come to the fore, if only journalism could set aside its fixation on crime, scandal, violence and conflict. Civic journalism advocates make much of a poll conducted for the Times Mirror Center for the People and the Press (1994) showing that most Americans believe the news media are a hindrance to solving problems. However, Glasser proposes the following paradox of civic journalism: "It fails, specifically, to address the predicament newsrooms face, to take an uncomfortably familiar scenario, whenever communities act intolerantly. What does a 'fair-minded' press do when a community consensus calls for a book burning? What is an appropriate response from a 'fair-minded' press when a popular vote yields a racist mayor?" (1999, p. 9) As Pauly (1999) and others have also pointed out, the mainstream commercial press has not always been the champion of democracy. Pauly writes that the United States has been in a process of becoming a democracy over the two and a quarter centuries of its existence, with each gain the result of hard-fought, sometimes violent, battles. The mainstream commercial press has not always, or even often, been on the progressive side, and has often been perceived as the enemy by those seeking to expand the franchise, to protect and expand civil liberties, and to enhance society's tolerance for diversity. Evaluating Civic Journalism's Impact After about a decade of experiments and activities said to reflect the characteristics and philosophy of civic journalism, what impact has it had? As numerous commentators have noted, it is not always clear what civic journalism is; therefore, it is difficult to know it when one sees it. Some of its promoters have even disavowed some of the activities done in its name. The clearest measure of outcome would be in the news content itself, since civic journalism's stated purpose is to reform journalism. This would involve changing the "culture" of journalism as reflected in everyday coverage, not just special "civic journalism projects." This culture is hard to change, given the entrenched ways of many journalists and their often open hostility to what they perceive to be civic journalism. Meyer puts it very plainly as follows: If the desired end result of public journalism is increased social capital, and if that result is achieved, it ought to surface somewhere in the observable world. A necessary first step for empiricists is to prove that there are any results of public journalism at all. Such proof might come in three progressive steps: 1) Is public journalism real? 2) If so, does it have any visible output? 3) If it does, do those outputs make any difference in the minds, hearts, or observable behavior of the citizenry? (1998, p. 258). Meyer presents five operationalizations to illustrate what the visible output of civic journalism might be: ù Citizen-based vs. campaign-based sources ù Fewer stories on campaign tactics and strategy and more concerning issues ù Less emphasis on the horse race ù Less emphasis on conflict and more on areas of agreement ù Use of polls to illustrate issues rather than documenting the horse race A recent study of the degree to which civic journalism has changed the culture of a paper, rather than simply being limited to a particular project, listed a number of the same operationalizations of outcomes that might be expected from civic journalism efforts. Blazier and Lemert (2000) expected that civic journalism efforts in The Seattle Times would exhibit the following characteristics, relative to traditional journalism: 1) The sources for news stories will be representatives of citizen organizations or be unaffiliated individuals more often or more prominently. 2) Civic journalism will provide more context, background and history of issues, rather than treating stories as discrete events or episodes. 3) Civic journalism will more often discuss solutions to problems, rather than simply present problems. 4) Civic journalism will more often provide mobilizing information - e.g., information that will facilitate citizen action and involvement. 5) Civic journalism will concentrate less on conflict and more on reasoned debate. 6) Election stories will focus on issues and candidate record, rather than the horse race and candidate character. Blazier and Lemert were unable to find clear evidence for most of these changes in the everyday news pages of The Seattle Times, despite that paper's public adherence to the principles of civic journalism. The results of a study of Kansas newspapers by McMillan, Guppy, Kunz and Reis were also mixed. They found that the civic journalism newspaper - the Wichita Eagle - focused more on issues and candidate records and less on the "horse race," and included a higher frequency of stories with "mobilizing information," compared with the more traditional newspaper, the Topeka Capital-Journal. One important hypothesis, that the civic journalism paper would include more citizen and unaffiliated sources, was not supported. The Virginian-Pilot and the Richmond Times-Dispatch The Pilot and the Times-Dispatch are the two dominant newspapers based in Virginia, each with a quarter-million circulation. The Times-Dispatch is the flagship of publicly held Media General Inc. It primarily serves the capital of Richmond, its suburbs and the surrounding Central Virginia region (an area with a population of about 1 million people). The Times-Dispatch maintains bureaus throughout Virginia and is considered a statewide paper. The Pilot is the flagship of the privately held Landmark Communications Inc. It circulates primarily in South Hampton Roads, including Norfolk, Virginia Beach, Portsmouth and Chesapeake (an area with about 1.1 million people). Both Media General and Landmark own other newspapers throughout Virginia and the Southeast. The Pilot and the Times-Dispatch are well-known within the journalism community as polar opposites on the issue of civic journalism. The Pilot was an early adopter of civic journalism and asserts that it has tried to integrate civic journalism into its newsroom culture. The newspaper received a Pew grant in 1996 "to explore the difference between deliberative and market-based polling and use the deeper issues that evolve to frame political coverage" (Pew Center Web site, Projects page). The Pilot has complemented its adoption of civic journalism with other changes - for instance, replacing a traditional beat structure with topic teams. While some journalists have praised The Pilot's approach to journalism, others have criticized the paper as faddish (Chambers, 1997). It was The Pilot that Rosemary Armao fled. The Times-Dispatch has a reputation for taking a traditional "paper of record" approach to news; its editors have openly criticized civic journalism. "We're not a buzzword company and never will be," said J. Stewart Bryan III, the chairman, president and chief operating officer of Media General (Chambers, 1997). The publisher of the Tampa Tribune, a Media General newspaper, has noted that Bryan "is skeptical of what has been called public journalism" (Cunningham, 2000). The Times-Dispatch and The Virginian-Pilot have been held up as exemplars of different ways to approach journalism: Charity's textbook did so several years ago in comparing the papers' coverage of then-Governor George Allen's quest for parole reform (Charity, 1995, pp. 41-45). Lambeth describes The Pilot's approach as emphasizing "not projects but day-to-day public framing of issues and storytelling. The focus is on people as citizens and as human beings trying to make civic sense of their public involvements. To meet these goals, the newspaper management created a special 'Public Life Team' of reporters and editors" (1998, p. 20). Postings on JOURNET, an e-mail discussion list for journalists and journalism faculty members, have underscored the reputations of the Times-Dispatch and The Virginian-Pilot as solid newspapers with distinct approaches to journalism. In 1998, Christer Lundquist, a Norwegian journalist, was planning a trip to the Eastern United States, wanted to visit local newspapers and asked JOURNET subscribers for suggestions. Bill Chronister, a copy editor at The Plain Dealer in Cleveland, suggested visiting, in North Carolina, The Charlotte Observer for its public journalism and the Raleigh News & Observer for its Internet operations. Then he added: "A bit farther north is the Norfolk Virginian-Pilot, another 'hot-bed' of civic journalism. Also in Virginia worth visiting is the more traditional Richmond Times-Dispatch." Hypotheses Regarding Coverage of the 2000 U.S. Senate Race Assessments of public journalism in practice have been done in a number of cities, but few opportunities have existed to compare competing newspapers covering the same event. One exception is the study, referred to earlier, comparing the Topeka and Wichita papers on local and state election coverage. The fall of 2000 offered a chance to compare how The Pilot and the Times-Dispatch covered an important, complex and high-profile event: the race for the U.S. Senate between Democratic incumbent Charles Robb and Republican challenger George Allen. From extensive reading and discussion concerning civic journalism, we can propose the following hypotheses about the two newspapers' campaign coverage: 1) A smaller proportion of total news coverage in The Pilot will reflect campaign-managed events than in the Times-Dispatch. 2) A smaller proportion of total news coverage in The Pilot will focus on the campaign "horse race," process and strategies of candidates than in the Times-Dispatch. 3) Conversely, a larger proportion of total news coverage in The Pilot, as opposed to the Times-Dispatch, will focus on issues, policies and candidates' positions. 4) The number and range of sources will differ in the two newspapers, with a larger proportion of Times-Dispatch sources representing political elites than in The Pilot. The Pilot's stories may also contain more sources than the Times-Dispatch's stories. Methodology All stories relating to the 2000 U.S. Senate race in Virginia were collected between Sept. 2, 2000 - the start of Labor Day weekend, the traditional beginning of the campaign season - and Nov. 8, 2000, the day after Election Day. These stories were collected from the front, metro and Virginia sections of the two newspapers. A few stories were also taken from the Sunday commentary sections; these were generally comparisons of the candidates on issues and their responses to reporters' interviews in Q-&-A format. No editorials, opinion pieces or columns were included. A few stories about the Senate election appeared in other sections, i.e., sports and business. The researchers decided to include only the articles from the general news sections and commentary sections, since these dealt most directly and fundamentally with the campaign. This process produced 95 stories from the Times-Dispatch and 120 stories from The Pilot. After a review of previous studies and of a number of stories in this collection, coding schemes for story origin, story content and source type were developed. Story origin, reflecting Hypothesis 1, consisted of three categories: Campaign/event-driven; Independent/enterprise; Elements of both. Story content, reflecting Hypotheses 2 and 3, also consisted of three categories: Candidate issues, policies, proposals, attitudes and character; Campaign process, horse race or insider point of view, Elements of both. To test Hypothesis 4, each story source was categorized into one of 11 categories, based on the following definitions: ù Campaign/party sources (candidate, campaign spokesperson or operative, political party source, candidate family member) ù Political establishment sources (current or former office holders at local, state or national level; well-known political/public figures) ù Non-elected members of former state or U.S. administrations (state agency head or retired military officers) ù Political pundits, academic sources, think tanks and interest groups (an organization itself or a spokesperson) ù Members of the public made available by the campaign, generally to represent a campaign issue or make a political point ù Unaffiliated members of the public, apparently not connected to the campaign; a "person on the street," perhaps present at a campaign event, but not part of the event's "program" ù Advertisements, including television, radio and print ads and billboards ù Polls ù Government agencies or spokespersons, documents and statistics ù Anonymous sources ù Other As in the implementation of any coding system, not every story or source seemed to fit cleanly into one category. Rules were established for coding these ambiguous cases. The key factor in coding a story's origin as campaign/event-driven was a "trigger" mentioned in the article. These typically were references to a speech, press conference, meeting, campaign tour or press release generated by a candidate or campaign. Stories were coded as independent/enterprise if they appeared to be the reporter's own idea - not based on an event that "required" coverage in the same sense as a candidate's speech. These articles included analyses of issues; profiles of candidates; "ad watch" stories, in which advertising content was dissected for its veracity; and stories based on government documents, such as campaign spending reports, obtained by journalists. Some stories clearly were triggered by a campaign activity or event, but enough "digging" or independent investigation was done to w arrant a code of "elements of both." In terms of story content, careful decisions were made when an article contained information about both issues/policies and the campaign process/horse race. For example, a story about the release of a poll might include significant content about how the candidates' stances on issues ostensibly were driving the poll numbers. Likewise, many stories started out discussing the mechanics of the campaign - where the candidate was appearing, what he was doing and why - and then reported on the candidate's speech or press conference. In fact, relatively few stories were "pure" - only issue-related, or only about the horse race. However, most articles could be categorized as primarily one or the other. If the content was nearly evenly divided, the story was assigned to the "elements of both" category. To help clarify these coding decisions, the researchers have included as exhibits two stories published on Oct. 27. Exhibit A is a Times-Dispatch story about the latest opinion polls in the Senate race. The story is driven by campaign events (the release of a poll) and focuses exclusively on the campaign process/horse race; it does not mention issues in a significant way. To civic journalism proponents, this story exemplifies a traditional approach to covering politics. Exhibit B is a Virginian-Pilot story based on interviews with 56 people from Fairfax to Virginia Beach, in which they discuss the issues that have drawn them to Allen or Robb. This story is based on independent/enterprise reporting and focuses almost entirely on issues. As such, it exemplifies a civic journalism approach to covering politics. The coding of sources involved the most effort to resolve ambiguity - for example, in the case of an appointed agency head in a former state administration serving as an official in a current Senate campaign. In an instance of this sort, the individual was placed in the code representing his or her current position and/or the code representing the role most closely associated with the campaign. Thus, this person was coded as a "campaign/party source." Unidentified sources were particularly troublesome. Several stories quoted "a campaign aide" or "an Allen spokesman" or "a Robb staff member," without giving the source's name. In such cases, the source was coded as a "campaign/party source." If a source was unidentified but was not connected to a campaign, it was coded as "Anonymous." In analyzing the data, the researchers further subdivided the anonymous sources into "Anonymous officials," such as "one former Capitol Hill staffer who asked to remain anonymous"; and "Anonymous 'real people,'" such as an unnamed "woman in denim shorts and a tank top" quoted at a political rally. Polls and advertisements often were problematic. Several stories attributed information to "a recent poll" or "polls say" without naming an opinion survey anywhere in the text. Such source references were coded as polls, despite the lack of specificity. Similarly, some stories attributed information to advertisements that were not specifically identified; these sources were coded as ads. One of the authors was responsible for coding the Times-Dispatch stories, the other for the Virginian-Pilot stories. Each author read and coded his set of stories twice. Then the authors together reviewed all stories and sources in both papers for consistency of coding. When they disagreed, they reviewed the article in light of similar stories and reached an agreement. Most cases were clear; changes usually involved moving a story origin or story content code into or out of the "elements of both" category. Difficulties in source coding were concentrated in dealing with various unidentified sources. As a final step, each researcher took a random sample of about 10 percent of stories from the paper he did not originally code and coded both stories and sources. This new coding was compared with the previous coding done by the other researcher and the outcomes were compared. Using the Holsti formula (1969, in Wimmer and Dominick, 1987), reliabilities were figured. For the sto ry origins, the reliability was 95 percent; on story topic, 90 percent; and on sources, 94 percent. The data from the above content analysis were entered into Microsoft Excel. For each newspaper, the researchers created two tables. One table listed each story. Each record included a story identification number, the headline, date, page, story origin code, story content code and other basic information. The other table listed each source in each story. Each record in this table included the story identification number; the original source citation; the source's name (standardized as "Last name, First name"); the source's code; and possible notes. Several hours were spent cleaning the data and checking their integrity. These tasks involved fixing misspelled names and ensuring consistency in coding. The tables then were imported into Microsoft Access, a database manager. Access can link the tables according to the story identification number - so that, for example, one could isolate all of the sources cited in campaign/event-driven stories. The data analysis was done in b oth Excel, especially with its PivotTable feature, and Access, with queries that involved joining, grouping and cross-tabulations. Results The results of these analyses are presented below at they relate to each of the hypotheses. Hypothesis 1: Campaign/event-driven stories compared with independent/enterprise stories The results of this analysis are shown in Table 1. Of the 120 stories from the Virginian-Pilot, about 44 percent were coded as campaign/event-driven, compared with about 75 percent of the 95 stories from the Times-Dispatch. Conversely, 55 percent of the Virginian-Pilot articles were independent/enterprise stories, compared with about 25 percent of the Times-Dispatch stories. It seems clear that the first hypothesis is supported. A higher percentage of the Senate race coverage in the civic journalism newspaper (The Pilot) was independent of campaign events - i.e., not the direct result of press conferences, speeches, press releases and other events staged by the campaigns or candidates. Table 1: Story Origin - Campaign/Event-Driven or Independent/Enterprise Norfolk Virginian-Pilot Richmond Times-Dispatch Story origin code Number of stories As percent of all stories Number of stories As percent of all stories Campaign/event-driven 53 44.2% 71 74.7% Independent/enterprise 66 55.0% 24 25.3% Campaign/event & enterprise 1 0.8% 0 0.0% Total stories 120 100.0% 95 100.0% Hypothesis 2: Issue coverage compared with "horse race" coverage Table 2 shows these results. It is clear that the Virginian-Pilot devoted a larger proportion (about 74 percent) of its coverage to the candidates' issue positions, policy statements, attitudes and character than did the Times-Dispatch (about 56 percent). The difference between the two newspapers is less if one includes stories that contained substantial elements of both issue positions and campaign horse race: About 83 percent for the Virginian-Pilot's stories were primarily or partly about issues, vs. about 72 percent of the Times-Dispatch stories. About 17 percent of The Pilot stories were mostly or completely about the campaign horse race or process, compared with about 28 percent of the Times-Dispatch stories. So, again the hypothesis is supported. The civic journalism newspaper devoted more of its Senate race coverage to issues than did the traditional paper, which focused somewhat more on the campaign process or horse race. Table 2: Story Content - Issues vs. "Horse Race" Norfolk Virginian-Pilot Richmond Times-Dispatch Story content code Number of stories As percent of all stories Number of stories As percent of all stories Process/horse race 20 16.7% 27 28.4% Issues 89 74.2% 53 55.8% Issues & process 11 9.2% 15 15.8% Total stories 120 100.0% 95 100.0% The differences between The Virginian-Pilot and the Times-Dispatch are even clearer when the story origin and story content codes are cross-tabulated, as show in Table 3. Slightly more than half of The Pilot's stories had an independent/enterprise origin and focused primarily on issues - the quintessential story by the standards of civic journalism. In contrast, less than 16 percent of the Times-Dispatch stories fell into this category. Table 3: Story Origin and Content - Campaign-Driven Process Stories vs. Independent/Enterprise Issue Stories Norfolk Virginian-Pilot Richmond Times-Dispatch Story origin and content code Number of stories As percent of all stories Number of stories As percent of all stories Campaign/event-driven & issues 27 22.5% 38 40.0% Campaign/event-driven & process/horse race 17 14.2% 18 18.9% Campaign/event-driven & both issues and process/horse race 9 7.5% 15 15.8% Independent/enterprise & issues 61 50.8% 15 15.8% Independent/enterprise & process/horse race 3 2.5% 9 9.5% Independent/enterprise & both issues and process/horse race 2 1.7% 0 0% Both campaign/event-drive & independent/enterprise; & issues 1 0.8% 0 0% Total stories 120 100.0% 95 100.0% Hypothesis 3: Number and types of sources The Times-Dispatch had more sources per story than The Virginian-Pilot, 5.84 compared with 4.46. The hypothesis predicted the reverse, that the civic journalism paper would include more sources. However, further analysis showed that while the Times-Dispatch used more sources in an average story, it more often cited the same sources. Across the entire set of stories for each paper, The Pilot had more unique sources (279) than did the Times-Dispatch (242). The Pilot cited a larger number of distinct sources and did not turn to the same sources as frequently as the Times-Dispatch did. An analysis of individual sources indicates this clearly. Table 4 shows the frequency with which campaign officials and spokesmen, the candidates themselves and academic political commentators were quoted. While the two candidates were quoted slightly more often in The Pilot, all of the other sources were quoted more frequently in the Times-Dispatch and accounted for higher percentages of the total source citations. For example, the spokesman for the Robb campaign, Mo Elliethee, was quoted 29 times in the Times-Dispatch, representing 5.2 percent of all of the paper's source citations, and 16 times in The Virginian-Pilot, or 3 percent of all its source citations. The Allen campaign equivalent source, Tim Murtaugh, was quoted 28 times in the Times-Dispatch, 5 percent of its total source citations, and 13 times in The Virginian-Pilot, 2.4 percent of its total source citations. The use of several academic sources was also quite different. The Times-Dispatch quoted five Virginia academic sources (Larry Sabato, Robert Holsworth, Stephen Farnsworth, Mark Rozell and Scott Keeter) a total of 32 times, or 5.8 percent of all its source citations, compared to 15 times in The Virginian-Pilot, or 2.8 percent of all its source citations. Table 4: Citing Candidates, Campaign Officials and Commentators Norfolk Virginian-Pilot Richmond Times-Dispatch Source Number of citations As percent of all sources Number of citations As percent of all sources Charles Robb 47 8.8% 48 8.6% George Allen 42 7.9% 36 6.5% Robb campaign spokesman 16 3.0% 29 5.2% Allen campaign spokesman 13 2.4% 28 5.0% Robb campaign manager 0 0% 10 1.8% Allen campaign manager 4 0.7% 12 2.2% Subtotals: Candidate, campaign spokesmen and campaign managers 122 22.8% 163 29.4% Academic commentators 15 2.8% 32 5.8% Total source citations in all stories 535 100.0% 555 100.0% This illustrates that the main thrust of Hypothesis 3 is supported, that a larger proportion of the Times-Dispatch sources fall in the campaign, party and political establishment categories. This analysis is extended in Table 5, which shows the results for all source categories. The percentages for all political establishment categories - campaign and party, other political establishment, former administration and unidentified officials - are higher for the Times-Dispatch; and when all these categories are combined, the difference between the papers is quite striking. Together, the four categories accounted for about 58 percent of all source citations in the Times-Dispatch, compared with about 43 percent of all source citations in The Virginian-Pilot. In addition, a key component of civic journalism is the use of non-elite, general public, "real people" sources. These accounted for about 12 percent of The Pilot's source citations, compared with about 3 percent of the Times-Di spatch's sources. The Times-Dispatch also referred more frequently to poll results, perhaps reflecting a more "horse-race" orientation. Table 5: Frequency of Source Categories Norfolk Virginian-Pilot Richmond Times-Dispatch Source category Total number of references As percent of all sources Total number of references As percent of all sources Campaign/party sources 167 31.2% 207 37.3% Political establishment sources 45 8.4% 73 13.2% Polls 14 2.6% 46 8.3% Pundits, academics, think tanks and interest groups 81 15.1% 89 16.0% Former administration members 8 1.5% 25 4.5% Unidentified officials 8 1.5% 16 2.9% Ads 51 9.5% 44 7.9% Government agencies/documents 46 8.6% 11 2.0% Public, made available by campaign 8 1.5% 14 2.5% Unaffiliated public (real people) 60 11.2% 15 2.7% Unidentified "real people" 5 0.9% 0 0.0% Other (other media, etc.) 42 7.9% 15 2.7% Total source citations in all stories 535 100.0% 555 100.0% To summarize, the results show strong support for the hypotheses generated for this study. The elements of civic journalism are clearly more likely to be present in The Virginian-Pilot than in the Richmond Times-Dispatch. Discussion Civic journalism remains controversial. Implementation of civic journalism principles often has been inconsistent and limited to special projects, even in those papers that support it. The Virginian-Pilot claims to have made civic journalism a part of routine news coverage, thus changing the newspaper's "culture." If that is so, then daily news coverage by The Pilot and the Times-Dispatch should be quite different. The Senate race presented an opportunity to see, in effect, if The Pilot practices what it preaches - and whether the results are truly different from the coverage by its traditional-newspaper, intra-state rival. We conclude from these data that The Pilot indeed is distinctive. Its coverage of the Senate election is clearly different from that of the Times-Dispatch in terms of story choice and framing as well as source selection. That being said, these differences may not all be due to the adoption of civic journalism principles by The Pilot. While the papers show some similarities in the size of their circulations and the communities they serve, the communities themselves are quite different. As the state capital, Richmond is the seat of the Virginia political establishment, which, along with state employees, constitutes a significant segment of the Times-Dispatch readership. Thus, the Times Dispatch, with the kind of coverage indicated here, may be serving its audience. Indeed, if the Times Dispatch were to adopt civic journalism principles, it still might provide more coverage of political elites than The Virginian-Pilot. The Pilot serves the Hampton Roads region of Virginia, a more transient and diverse area, with its economy based in the military and tourism. It is not as attached to the political elites of Virginia, and its readership may not be as deeply rooted in the state's traditions or as interested in the inner workings of the state's political machinery. Therefore, not all the differences between these two newspapers, as illustrated here, may stem from the adoption of civic journalism by one and its rejection by the other. However, the difference in the papers' approach is clear in the way they reported who won the Senate seat. On the morning after the election, the Times Dispatch announced Allen's victory over Robb with a 51-paragraph story: Only one paragraph was devoted to issues - and that one in a horse-race context. The Pilot's story about Allen's election was 28 paragraphs long, with seven paragraphs devoted to the issues that were important to the campaign. References Blazier, T.F., and Lemert, James B. "Public journalism and changes in content of The Seattle Times." Newspaper Research Journal. 21(3), 2000, 69-80. Chambers, Paula Peters. "The Paper Chase." Virginia Business. December 1997, 10-15. Charity, A. Doing Public Journalism. New York: Guilford, 1995. Corrigan, Don H. The Public Journalism Movement in America. Westport, CT: Praeger, 1999. Cunningham, Brent. "A Publisher's Life: In the Lab." Columbia Journalism Review. May/June 2000. Dewey, John. The Public and Its Problems. Denver: A. Swallow, 1927. Glasser, Theodore L. The Idea of Public Journalism. In T.L. Glasser (Ed.) The Idea of Public Journalism. New York: Guilford, 1999, 3-18. Habermas, J. The Structural Transformation of the Public Sphere, trans. T. Burger and F. Lawrence. Cambridge: MIT Press, 1989. Haiman, Robert J. "Hey, editors: Just stop the nonsense." Newspaper Research Journal. 19(4), 1998, 2-12. Killenberg, G. Michael, and Dardenne, Robert. "Instruction in News Reporting as Community-Focused Journalism." Journalism Educator, 52(1), 1997, 52-58. Lambeth, Edmund B. "Public Journalism as a Democratic Practice." In E.B. Lambeth, P.E. Meyer and E. Thorson (Eds.), Assessing Public Journalism. Columbia: University of Missouri Press, 1998, 15-35. McMillan, Sally J., Guppy, Macy, Kunz, William, and Reis, Raul. "Public Journalism: What difference does it make to editorial content?" In E.B. Lambeth, P.E. Meyer and E. Thorson (Eds.), Assessing Public Journalism. Columbia: University of Missouri Press, 1998, 178-190. Merritt, Davis, and Rosen, Jay. "Imaging Public Journalism: An Editor and Scholar Reflect of the Birth of an Idea." Roy W. Howard Public Lecture in Journalism and Mass Communication Research. School of Journalism, Indiana University, Bloomington, IN. April 13, 1995. Meyer, Philip E. "If it works, how will we know?" In E.B. Lambeth, P.E. Meyer and E. Thorson (Eds.), Assessing Public Journalism. Columbia: University of Missouri Press, 1998, 251-273. Pauly, John J. "Journalism and the Sociology of Public Life." In T.L. Glasser (Ed.) The Idea of Public Journalism. New York: Guilford, 1999, 134-151. Pew Center Web site, Projects. . March 2001. Rosen, J. What are Journalists For? New Haven: Yale University Press, 1999. Wimmer, R., and Dominick, J.R. Mass Media Research (2nd edition). Belmont, CA: Wadsworth, 1987. [1] For the most part, the terms "public" and "civic" journalism are used interchangeably in the research and commentary on the subject. We will use "civic journalism" here, except in quoted material.