A long list of television news studies points to network consonance, specifically similarity in news topics (Whitney et al., 1989; Atwater, 1987), sources (Atwater, 1989; Whitney et al., 1989), and formats (Atwater, 1989). This is due in part to factors that influence news selection and presentation that are not unique to one network. They include: profit motivations (Epstein, 1973), constraints of current technology (Epstein, 1973), news gathering routines (Sigal, 1973), and training and values of news personnel (Gans, 1979).
Although network news similarity is assumed to extend to presidential campaign coverage (Robinson & Sheehan, 1983), election coverage comparisons are usually made between media--television and wire services (Robinson & Sheehan, 1983), television and newspapers (Johnson, 1993; Johnson, 1993b; Johnson et al., 1996; Tedesco et al., 1996), television, newspapers, and news magazines (Kerbel, 1997), newspapers, local television stations, and national television networks Oust et al., 1996)--rather than between networks. Studies of media and elections either use one network to represent television (Robinson & Sheehan, 1983), examine coverage from different networks without looking at differences between them (Barnhurst & Steele, 1997; Johnson, 1993b; Johnson et al., 1996), or make some network comparisons without statistically testing for--or drawing general conclusions from--these comparisons (Johnson, 1993b; Just et al., 1996; Kaid et al., 1996; Lichter & Smith, 1996; Lowry & Shidler, 1995).
Results from the few campaign studies that assess network differences disagree over the extent of television news homogeneity. Some studies illustrate similarities in the networks' coverage of election themes (Kerbel, 1997), tone Oust, 1997), horse-race focus (Johnson, 1993), and issues (Rudd & Fish, 1989). Others point to network differences in the amount of coverage without exploring the nature of that coverage (Johnson, 1993b; Johnson et al., 1996; Lichter & Smith, 1996). Some studies have demonstrated significant differences among networks' campaign coverage in the use of sound bites (Lowry & Shidler, 1995; Russomanno & Everett, 1995) and political "ad watches" (Kaid et al., 1996; Tedesco et al., 1996). As yet, though, no comparisons have been made of how public opinion is covered by the network news shows during a campaign. This study examines network news consonance by focusing on the use of the public as a source in and subject of presidential general election coverage. Specifically, frequency of public opinion coverage, ways the public is represented, types of people included, and topics they discussed are compared across ABC, CBS, and NBC.(1)
Despite the media focus on the race between the candidates (Patterson, 1993), coverage of the public during campaigns is both prevalent and important. As Marion Just and her co-authors (1996, p. 43) remind us, "Campaigns begin and end with voters." The extensive coverage of citizens (in 1992, second only in frequency to candidates and their aides) demonstrates the central role played by the public in television news' conception of presidential election campaigns (Kerbel, 1994, p. 31). In addition, since coverage of public opinion can influence public opinion (Noelle-Neumann, 1984; Traugott, 1992), it is important to examine news representations of democracy's ultimate decision makers.
There are reasons to suspect that there were network differences in public opinion coverage during the 1996 presidential election. Just as there is greater product differentiation in large local television markets (Atwater, 1984), increased competition in the national media marketplace of the 1990s has led to diversification of general news coverage (Bagdikian, 1991). NBC executives claim to have made their evening news show distinctive by covering more of what they deem relevant to people in the "heartland"--an approach they refer to as "populist" (Tucher, 1997). …