A spate of recent scholarship on the contemporary crisis in public broadcasting has highlighted the economic, political and technological challenges faced worldwide by non-commercial broadcasters (Avery, 1996). Within this mediascape critics have noted a fundamental change in the nature of public broadcasting, away from its educational, service-driven origins toward an audience- and funder-driven orientation, in which public broadcasters target demographically upscale segments of the potential audience (Aufderheide, 1991; Avery, 1993; Behrens, 1996b; Croteau, Hoynes and Carragee, 1996; Friedland, 1995; Goodman, 1993; Hoynes, 1994; Jarvik, 1997; Lapham, 1993; Ledbetter, 1997; McChesney, 1993; Mifflin, 1997; Raboy, 1995; Schiller, 1989; Stavitsky & Gleason, 1994; Tolan, 1996). These audience segments include viewers and listeners most likely to support public television and radio stations financially, people who are also attractive to corporate underwriters. This changing conception of audience raises timely and important questions about what constitutes the public to which the public broadcaster is beholden, and through which the industry claims its legitimacy in tenuous times.
One way to illuminate the relationship between public broadcaster and audience is by examining audience research as practiced in U.S. public television. As Hurwitz has noted in the context of commercial broadcasting, "the adoption of increasingly sophisticated research methodologies has been more a product of broad economic, social, political and cultural motives than of the pursuit of truth or better service" (1984, p. 208). This argument wields additional power in the realm of non-commercial media, in which the broad application of audience research reflects the tension between the imperatives of those who fund the service, whether they be consumers or corporations, and traditional conceptions of public broadcasting's social role.
Public broadcasters today are actively engaged in studying their viewers and listeners; networks and many stations employ audience researchers, and a cottage industry of consultants has emerged. Many public television program directors pore over the "overnight" ratings derived from Peoplemeter data, just as their commercial counterparts do. Such research, however, has been a lightning rod for criticism in terms of encroaching commercialization (Stavitsky, 1993; 1995). This study will explore the evolution, diffusion and application of audience research in U.S. public television from its origins through the 1970s, the formative period for the system.
Researching Audience Research
Audience research is fundamental to decision-making in U.S. commercial broadcasting. Programming decisions have long been grounded in audience interests and desires, as perceived by a variety of research methodologies. Research data, reported as "ratings," provide the institutional knowledge used for the sale of advertising time, the industry's economic base, as well as providing criteria for program selection. Accordingly, most scholarly attention paid the subject of audience measurement has focused on functional, methodological and historical issues of the commercial research industry (Beville, 1988; Webster & Lichty, 1991; Webster & Phelan, 1997). More recently, however, scholars have viewed audience research in terms of the complex accommodations among broadcasters, advertisers, policy-makers, academics, and members of the audience. Rowland's (1983) study of the role of effects research in the television violence debate, Buzzard's (1990) history of the ratings services, Rogers' (1994) description of the roots of communication study, Ettema and Whitney's (1994) edited volume on how media "create" audiences, and Ang's (1991) comparison of U.S. and European approaches to objectifying audiences all typified audience research as a social construction with important political and economic implications.
Most of the foregoing discussion focused on commercial media, in which the function of audience research is unquestioned. But what of non-commercial media, which generally view their purpose for broadcasting in social and cultural (as opposed to financial) terms? How do research models designed for marketplace applications serve the sometimes divergent, avowedly sociocultural, purposes of public broadcasters? Some recent studies have examined these questions in international contexts. Eaman (1994) traced the history of audience research by the Canadian Broadcasting Corporation and argued that commercial-style ratings proved inadequate for ascertaining public priorities in the area of broadcasting. Similarly, Ang wrote of Western European public-service television that "knowing the size of the audience alone is not sufficient to gauge the success or failure of public service television's communicative efforts, not least because success and failure are a normative rather than material issue here" (1991, p. 30). In these contexts, however, different funding structures (i.e. higher, albeit falling, levels of public subsidy) create different research imperatives than in the U.S. case.
Two historical studies examined audience research in U.S. non-commercial radio. An analysis of two pioneering educational radio(1) stations -- WOSU in Columbus, Ohio, and WHA in Madison, Wisconsin -- found evidence of audience studies of various kinds as early as the 1920s (Stavitsky, 1993). These included maps that noted locations from which listener mail had been received; analyses of mail response; and on-air solicitations of listener reaction to certain programs. While many studies conducted by educational broadcasters involved pedagogical concerns (i.e. the effectiveness of radio in teaching), attempts to measure audiences and their reactions to programs continued on a sporadic basis in educational radio for decades, though limited primarily to the larger, university-based stations. As Ralph Johnson, a former WHA manager noted: "It's not that the interest wasn't there; the money wasn't" (personal communication, June 30, 1989). The National Association of Educational Broadcasters (NAEB) convened a Research Committee in the 1950s, which expressed interest in hiring an audience researcher, lamented the lack of funds for audience research, and discussed purchasing commercial ratings data (Stavitsky, 1993, p. 16). However, NAEB President Harry Skornia, years later, reflected the view of many educational broadcasters that "great care must be taken in the application of findings, if we are to avoid the shortcomings found in the commercial media application of ratings" (1966, p. 37).
Another study found the use of audience research in non-commercial radio became a contentious issue after the Corporation for Public Broadcasting (CPB)(2) began providing funds for studies of the radio audience in the 1970s (Stavitsky, 1995). While some station managers welcomed audience research as valuable feedback, many "greeted the methods, paradigms, and proponents of research with open hostility and disdain," according to a prominent research consultant (Giovannoni, 1991, p. 3). However, after enduring flat or decreasing tax-based revenues through the 1980s, weathering National Public Radio's 1983 fiscal crisis, and attending CPB-sponsored seminars on the function and value of research, public radio managers came to accept audience research as an essential management function. Nonetheless, for some critics, audience research had come to symbolize the transformation of public radio. "I think there has been an influx of commercial people.... Guys in suits with charts and pages of numbers," public radio personality Garrison Keillor said. "I think that this is a pretty dreadful development" ("Thoughts from Lake Wobegon," 1994, p. 58).
Educational Television and the Audience
The contemporary U.S. public television system emerged from what had been a loose confederation of "educational" television (ETV) stations, founded during the 1950s and 1960s (the first of which, KUHT in Houston, went on the air in May 1953 [see Hawes, 1996]) and licensed primarily to educational institutions and non-profit community groups (see Day, 1995, and Robertson, 1993, for histories of the development of the public television system). ETV producers and managers, however, knew little about their audiences, for several reasons. First, commercial research services (such as Nielsen) did not routinely measure ETV viewing, in the absence of clients to pay the costs of collecting the data, and customized studies were expensive (see Schramm, Lyle, & Pool, 1963, pp. 18-21). Second, from the early days of radio, educational broadcasters were interested in more from research than simply "counting the house"; they wanted answers to pedagogical questions such as "what are audience members learning from our programs?" and such questions were not asked by the commercial firms (Lazarsfeld, Cantril, & Stanton, 1939). In addition, in the years before underwriting and viewer "memberships" became commonplace, stations received the bulk of their funding from their institutional licensees, or, in the case of community licensed stations, from foundations and other grantors. Accordingly, some ETV managers felt more beholden to their university and foundation colleagues than to unseen, uncounted and little-understood viewers.(3)
The Ford Foundation's Fund for Adult Education, which supported the early development of the ETV system, funded an early audience survey, in 1957 for KRMA-TV, the school-district ETV station in Denver (Public opinion survey, 1957). The survey measured how many people tuned to KRMA and how often; their demographics; what programs they watched; and the "value" of those programs (in terms of "broadening their backgrounds in knowledge and culture"). Similar studies were also conducted at some stations licensed to research universities, generally conducted by graduate students under the...