In the past 25 years, the Big Three broadcast television networks, ABC, CBS, and NBC, have experienced a significant decline in the share of the prime-time viewing audience. In 1980, more than 90% of television viewers were tuned in to one of these three networks during prime time. By 2005, the season ending average prime-time share of the Big Three networks had fallen to 32%. This means that during the 2004-2005 television season, fewer than one in three households using television during prime time were tuned to ABC, CBS, or NBC (Head, Sterling, & Schofield, 1984, p. 105; Nielsen Media Research, 2005; Owen & Wildman, 1992).
Explanations for the decline vary. In the early 1990s, network executives denied that a change in viewing was taking place. Instead, they insisted that Nielsen's new PeopleMeter was underestimating the size of network audiences (Piirto, 1993). The more common explanation for the decline of Big Three network shares of the television viewers was competition for viewers from new cable networks, new broadcast networks, and home viewing of VCR/DVDs (Carman, 1999; Carter, 1992; Dimmick, 2003; Henke & Donohue, 1989; Kaplan, 1978; Krugman & Rust, 1987, 1993; Lin, 1994; Ross, 1999). The penetration of remote control devices into the majority of television households during this time period also made it easier for viewers to casually surf through the new channels (Ferguson, 1994). In addition to these technological explanations, one might also ask about the social changes that accompanied the technological changes (Parsons, 2003).
This paper is an analysis of technological and social factors associated with the 25-year decline in the prime-time shares of the top three broadcast television networks.
The Substitution Hypothesis
The decline of the Big Three's prime-time shares indicates that those using television during prime-time are watching less ABC, CBS, and NBC programming, and are instead viewing more programming offered via other television sources, discussed below, such as: new broadcast networks; new networks available exclusively on multichannel video program distributors (MVPD) such as cable television, DBS, and home satellite dishes (HSD); or videocassettes/DVDs (FCC, 2006; Krugman & Rust, 1987, 1993; Lin, 1994).
Over the same period that broadcast shares have fallen, new broadcast networks have been established: Fox; ION Television Network, formerly PAX TV; CW Network, formerly WB and UPN; and MyNetworkTV, formerly News Corporation-owned UPN stations that were stranded by the UPN-WB merger (Romano, 2006a; Seid, 2006). These new networks earned a season-ending average 16 share in 2004-2005, which represents half the 32 share of the Big Three networks (Nielsen Media Research, 2005). Among Spanish-language viewers, Telemundo, Univision, and Azteca have also become formidable competitors to the Big Three (Romano, 2005).
Cable television, in combination with other multiple video program distribution (MVPD) providers such as DBS, has been growing steadily since 1980, and in 2005 reached over 85% of television households (FCC, 2006). Among subscription-based MVPD sources, cable television has dominated. Cable penetration, measured as the percentage of U.S. television households with cable, has increased from just over 20% of households in 1980 to just over 60% in 2005 (Brown, 2004; FCC, 2006). Both cable and the VCR have displaced broadcast television on the time spent dimension (Dimmick, 2003).
The growth in cable penetration accelerated following the FCC's 1972 Open Skies order, the production of affordable downlink dishes by Scientific Atlanta, and the growth in program providers such as Home Box Office and Ted Turner's Superstation, later called WTBS (Parsons, 2003). As more cable networks were born, cable started to differentiate itself from broadcast TV, and became the "television of abundance" (Bates & Chambers, 2004). Cable penetration peaked in 1998 at 66.8%, and has declined slowly since. Satellite television penetration has grown since the mid 1990s on the success of Direct Broadcast Satellite (DBS) services (FCC, 2006).
What this all means is that the Big Three networks faced a very different competitive environment in 2005 than they did in 1980. In 1980, American homes had an average of 10.2 television channels from which to select (Compaine, 2000). In 2004, the majority of subscribers had between 54 and 90 channels available (Warren, 2004). Videocassettes, DVDs in combination with multiple sets within households, and multimedia remote control devices, have further contributed to the "proliferation of choices" that has led to a decline in broadcast network shares (Ferguson, 1994; Henke & Donohue, 1989; Krugman & Rust, 1993).
One might argue that the decline in Big Three prime-time television network shares is the result of increasing use of home computers. But, the "share" measurement is based only on those individuals or households using television during the time period, and so a computer user who is not simultaneously watching television would not count against the Big Three network share. It is a different matter altogether when a computer user is watching a download of a broadcast or cable television network program. In that case, ratings reports will eventually include all forms of television program viewing, including computer downloads, via cell phones, PDAs, and even Play Station Portables (Dickson, 2006; Gerbrandt, 2006). Regardless, evidence of time displacement between computers and television is mixed. Both substitution and complementary effects have been documented (Dutta-Bergman, 2004; Kayany & Yelsma, 2000; Lin, 2004; Robinson, Barth, & Kohut, 1997).
It stands to reason that simple economic forces are at work in the decline of Big Three prime-time shares. The substitution hypothesis has a long history in media research (for a review, see Dutta-Bergman, 2004). Among those competition-based theories is the principle of relative constancy which suggests that consumer and advertiser spending on mass media is relatively constant, and what changes with the introduction of new media is simply the way the consumers' resources are distributed (McCombs & Eyal, 1980).
Contrary to the principle of relative constancy, time spent with media has increased since 1980. In 1980 the average American home had a television on for 6.60 hours per day, but in 2004, this number had increased to 8 hours, 11 minutes per day (FCC, 2006). This is consistent with Wood's (1986) and Wood and O'Hare's (1991) re-examination of the data from McCombs and Eyal's (1980) study in which the authors rejected the notion that new media necessarily displace old media. Wood (1986) argued that economic growth and lower prices for media hardware may hide an increase in demand for new media, even as media spending may show a declining proportion of consumer spending. Time spent with media also increased in the 1950s. Additional time spent with new media came at the expense of reading, radio listening, and a number of other daily activities, including sleep (Owen, 1999, p. 11).
The network share is a good indicator of the relative commitment of consumers to one form of television over another, even as the total time spent with television has increased. One might argue that the Big Three networks offer a unique form of television that emphasizes mass, as opposed to niche, audiences. Evidence that the Big Three networks are increasingly abandoning niche programming in favor of mass audience programming is shown by the steady decline in program diversity of prime-time programming between 1954 and 2003 (Einstein, 2002).
Attention to a particular channel is a zero sum game in that one channel's viewers comes at the expense of another's (Owen & Wildman, 1992, p. 165). Thus, the average television audience share would be expected to decline when the number of available channels increases (Picard, 2002). A negative relationship between measures of multiple video program distribution and broadcast network shares would be an indication of substitution between the two media (Krugman & Rust, 1987, 1993; Levy & Pitsch, 1985, p. 66). Thus:
[H.sub.1]: The greater the penetration of MVPD into households, the lower the Big Three share of prime-time viewers.
The most common explanations for the precipitous decline of the Big Three prime-time shares are technological. But one might also ask about social conditions that provide the climate in which those technologies can thrive. In other words, what changes occurred during the past 25 years that were also associated with the network decline?
The Social Differentiation Hypothesis
As the nation grows, it becomes more diverse in ethnicity, occupations, interest groups, lifestyles, and myriad other observable social categories. This increase in social heterogeneity and complexity associated with population growth is called social differentiation.
Social differentiation, a theory that can be traced to Spencer's (1860/1891) applications of organic analogies to social change, has implications for a number of phenomena, including information and entertainment preferences, media organizational characteristics, citizen behaviors, and media coverage (Demers, 1996; Hindman, 1996; Tichenor, Donohue, & Olien, 1980). Social differentiation, also called structural pluralism, is defined as the degree of heterogeneity along institutional and specialized interest group lines, in a way that determines the potential sources of organized social power (Tichenor et al., 1980, p. 16). Social systems with greater social differentiation have more potential sources of political and economic power than do more homogenous social systems (Clark, 1968; Hindman, Littlefield, Preston, & Neumann, 1999).
Focusing on the individual level of analysis, researchers explain media choice in terms of how well each medium satisfies an audience member's expectations, or gratifications sought, from a repertoire...