Television Audiences Across the World

Deconstructing the Ratings Machine

J. Bourdon, C. Méadel, J. Bourdon, C. Méadel

• Publisher: Palgrave Macmillan
• Year: 2014
• ISBN:
  9781137345103
• Topic: Media & Performing Arts
• Subtopic: Film & Video

This book is the first to deal with the world composition of television ratings. It focuses on the peoplemeter, a 25 year old technology which succeeds in homogenizing very different populations and television practices. It provides a fascinating account of the production of figures on which the whole world of popular culture depends.

Part I
Inventing Measurement

1

The Politics of Enjoyment: Competing Audience Measurement Systems in Britain, 1950–1980

Stefan Schwarzkopf

What does it mean to ‘measure’ audiences?

The measurement, analysis and research of broadcast audiences have a history reaching back to the origins of commercial radio during the 1920s. With the arrival of television during the 1930s, methods that had been developed in the radio industry were quickly adapted to measure audiences for television broadcasts (Beville, 1996; Buzzard, 1999; Robinson, 1947). Early practitioners of broadcast audience measurement systems realized that their activities had a bearing on the foundations of a new type of media-oriented democracy, since programme ratings could enable or disable consumer choice between programmes and broadcasting stations, and promote or distort the democratic process by channelling advertising sponsorship into the hands of broadcasting stations that held specific political views. Given the recent debates over the measurement of online audiences for informational content and entertainment, it is worth revisiting the technical complications and socio-theoretical rivalries that accompanied the introduction of audience measurement systems from the 1930s onwards (Balnaves and O’Regan, 2002; Bermejo, 2009;Napoli, 2003: pp. 6–10; Webster, Phalen and Lichty, 2000: pp. 1–27).

The development of television audience measurement systems in Britain offers a number of insights into both the intended and unintended consequences of the introduction of, and the competition between, different systems of audience measurement for a modern media environment. The British Broadcasting Corporation (BBC) began to broadcast televised programmes on a regular basis from a transmitter in north London (Alexandra Palace) in November 1936. Britain was thus the first country in the world to have a public television channel (Briggs, 1965: p. 594). Research on the few hundred members of the public who owned a television set began in the same year. The four decades or so between 1936 and 1981 witnessed the emergence of the first publicly financed audience research programme in the world, the introduction in 1955 of a commercially financed and more quantitatively oriented rival audience research programme, and finally the merger between these two forms of television audience measurement systems. The history of television audience research in Britain provides the circumstances that epidemiologists and economists would require for a ‘natural experiment’. The case allows media sociologists to gain unique insights into the effects that the act of measurement itself has on those who measure, on those who use the produced data, and on those who are being measured.

From the mid-1930s, the licence fee-funded BBC became very proactive and innovative in developing audience measurement methods, first for radio and later for television. At the forefront of these methodological and technical developments was the Head of BBC Audience Research, Robert J. Silvey. His personal background as a Fabian socialist and the organizational identity of the BBC explain why the corporation collated television audience figures in a specific way. Instead of focusing only on the number of TV sets that were tuned in and the number of people who watched a particular programme, Silvey and the BBC wanted to know whether viewers understood and appreciated the programmes they watched. For Silvey, audience measurement methods had to transcend purely quantitative measurements and also use qualitative research to establish whether or not viewers actually enjoyed a programme or rejected its content, tone and implications.

While the BBC saw broadcasting essentially as a public service, its main rival from 1955, the advertising-financed television network ITV, saw broadcasting as a business model. To serve the needs of this commercial network, private research companies developed an audience measurement system which produced only quantitative data on the number of TV sets that were tuned in to the ITV channel at a given time (‘ratings’). The British case thus represents two ideal types of broadcasting and audience research ontologies, one that attempted to meet the needs of a public broadcaster and the other system meeting the needs of a set of organizations linked together through market transactions.

Today, the audience research community is fully aware that the purely quantitative measurement of audiences (i.e., audience sizes and programme ratings) does not provide a full picture of the relationship between audiences and broadcast content. Audience stratification, audience behaviour, audience reactions, subjective enjoyment and the individual uses of media are now at the heart of contemporary media research (Gunter and Wober, 1992; Roberts and Lemieux, 1981; Schrøder et al., 2003: pp. 26–32; Vorderer, Klimmt and Ritterfeld, 2004). This focus on subjective audience reactions to broadcast content, a focus normally associated with the uses and gratifications school of communication studies of the 1970s, first emerged as a public research programme at the BBC (Mattelart and Mattelart, 1998: pp. 123–7; Schrøder et al., 2003: pp. 37–8). During the 1950s, this audience-centred outlook of the BBC’s research department became sidelined by a market-driven focus on the mere quantification of television audiences who came to be seen as made up of passive viewers.

Origins of television audience research in Britain, 1936–1955

Although financed by a license fee and thus formally independent of the imperatives of the commercial market for audiences and advertising, the BBC faced a tremendous virtual or indirect competition from advertising-sponsored continental-European broadcasting stations. Because foreign radio stations like Radio Normandy and Radio Luxembourg needed to assure advertisers of their attentive audiences, they broadcast ‘light’ music, dance music and variety shows on Sundays, when the BBC featured mostly religious programmes, choir music and generally more serious offerings (Briggs, 1965: pp. 54–5, 234–8; Scannell and Cardiff, 1982). The British press, which by the mid-1930s had largely turned against the BBC, used this disparity in order to castigate the BBC for its allegedly aloof public service philosophy (Street, 2006: pp. 138–46). Newspapers presented the BBC as an organization out of touch with what the public ‘really wanted’.¹ As a response to such public attacks, and to forestall any government-led restructuring (i.e., privatization) of the broadcasting industry, the BBC felt obliged to introduce methods to assess audience size and attitudes. Internal critics of audience research feared that the data obtained from panel surveys and daily interviews would end up driving programme policy, and thus override the authoritative judgement, experimental instincts and discerning connoisseurship of the BBC’s programme makers.²

Because of the strong internal resistance against audience research, it took until late 1936 before the BBC finally began to develop a coherent audience research programme by setting up a Listener Research unit under the former market researcher Robert J. Silvey (Silvey, 1974). With his arrival, the BBC began to investigate audience sizes and listener opinions proactively and on a continuous basis. In January 1936, a first questionnaire was sent out to radio licence holders in order to find out about their listening habits and their particular preferences for programme types, like dance music, talks, religious programmes, choir music,and so on.³ From October 1937, the BBC used a panel of 2000 listeners (later expanded to 4000 listeners), who recorded their listening in diaries which were then analysed on a weekly basis by the BBC.⁴ In December 1939, a system to measure the amount of listening was introduced which was based on 800 interviews conducted on a daily basis to assess how many people had listened to which programmes the day before.⁵

Because of the small number of private owners of television sets in the country, television audience research had its beginnings merely as a poor cousin of radio audience research. Nevertheless, the BBC began television audience research at virtually the same moment as it started its regular surveys of radio listening habits. In December 1936, the BBC appealed to owners of television sets to make themselves known to the Listener Research unit so that a form could be sent to them, asking owners to state when they had bought their set, which brand of set they owned, how and when they watched the BBC television programmes, and whether sound and picture quality were acceptable. In early 1937, the feedback thus received was compiled in the form of a report, which showed that about three-quarters of all returned questionnaires approved of the light entertainment programmes (cabaret, variety shows, music, ballet), but people did not think that the studio demonstrations of cooking, washing and ironing were of any use. Most viewers asked for more and longer films, for cartoon films, and they expressed dislike of the high number of repeated programmes.⁶

Two years later, in spring 1939, a second television enquiry was organized. Following up on the earlier survey, viewers were again asked whether the proportion of light entertainment was balanced and whether the present systems of arranging repeats was acceptable. Crucially, this time the questionnaire also asked viewers to point out programme types which they ‘liked’, and the questionnaire in general contained for more specific questions regarding the length of dramatic productions; how often and in which slots the programmes should be repeated; whether the average length of the evening programme was sufficient, too short or too long; how often and for how long there should be intervals between programmes or within plays; whether viewers preferred male or female announcers; whether announcers spent too long or not long enough on describing programmes; how many people watched television on their set ‘fairly regularly’, and many more detailed questions.⁷ By April that year, the BBC had received 4000 questionnaires of the 4800 it had distributed to owners of sets, of whom 63 per cent thought the service was satisfactory and 78 per cent thought it was also getting better. Also for the first time, Silvey’s research unit broke down the answers they had received on the questionnaires by professional status and separated respondents into the following groups: engineers, company directors, civil servants, traders, and salesmen.⁸ Although the 1939 questionnaire contained a lot more detailed questions and for the first time made use of segmentation as an analysis tool, viewers were not asked about specific programme items (only types) and were also not asked to rate or rank specific programmes over others. The questionnaire was clearly designed to present the BBC as an organization interested in viewers’ opinions, attitudes and behaviours. In other words, viewers were asked whether they generally liked what they saw – not what they would like to see.

At the time of the outbreak of war in September 1939, the BBC regularly reached between 20,000 and 25,000 television homes with their programmes (Briggs, 1965: p. 620; Henry, 1986: p. 27). After the interruption of television services during the war, the BBC resumed its television broadcasts in 1946, at which time Silvey immediately began to petition his superiors to provide funds to enable him to extend the survey techniques developed for radio audiences, including panels to study the amount of listening and appreciation, to the study of television audiences.⁹ This request marks a significant moment in the history of television audience research in Britain, for it shows how even in the complete absence of any form of competition by other television stations, a measurement regime that was developed for one broadcast medium – radio – became extended to cover another medium.¹⁰ In spring 1946, Silvey planned to create a panel of up to 1000 television viewers which would have allowed him to break viewers’ behaviour down into categories of gender, age and what he called ‘taste groups’. Silvey and the Director of the BBC’s Television Services explained that the panel research was necessary in order to provide creative feedback to the makers of television programmes, to avoid their ‘mental isolation’ from the viewing public and a feeling of being pleased with themselves as final judges of their output.¹¹

Silvey’s attempts at creating a television audience panel, however, fell on deaf ears and it was not before 1948 that he was allowed to launch a study into television set ownership and viewing behaviour under the title ‘The Viewers’ Vote’.¹² Two years later, in 1950, a second comprehensive television audience research exercise was conducted which followed up on the findings of the earlier research.¹³ The two television audience research reports of 1948 and 1950 represent the BBC’s preparation for and response to the Beveridge Committee on Broadcasting, which sat between 1949 and 1950. The task of the Committee was to review the licence of the BBC and the nature of broadcasting in Britain in general. Although the Committee recommended to parliament a renewal of the BBC’s charter and extension of the BBC’s monopoly in radio and television broadcasting, a minority report was filed which criticized the BBC’s broadcasting monopoly and its lack of responsiveness to changes in popular taste, especially regarding television.¹⁴ Thus, by 1950, Silvey’s audience research unit had once again become a tool for the BBC to create data on levels of audience satisfaction, in order to fend off the organization’s many critics and produce favourable evidence for the numerous parliamentary committees of enquiry which were to follow the Beveridge Committee.

The 1950 audience survey for the first time made full use of the panel technique, which Silvey had developed for the research of radio audiences a decade earlier. Members of the TV panel, who remained on the panel for 12 weeks before being replaced, recorded in a weekly questionnaire to what extent their TV sets were used as well as their opinions of the programmes they had watched.¹⁵ The results of the weekly panel data were used by Silvey to construct a ‘Reaction Index’ which measured not so much how many people watched the TV programmes, but to what extent they liked them.¹⁶ The key differentiation between audience size and levels of audience appreciation was methodologically supported by a well-designed system of viewer/listener panels, viewer/listener diaries and regular viewer/listener interviews.¹⁷

By the time Silvey was given the go-ahead for his next large-scale television audience survey in early 1952, the panel and the diary method in particular had developed into a mainstay of the methodological spectrum of social and market research. Silvey’s newly constructed panel of viewers, who had been selected from 57,000 families who volunteered for the research, rotated every 12 weeks and consisted of 2000 viewers in total.¹⁸ The total volume of listening and viewing (i.e., audience sizes) was measured through the continuous Survey of Listening and Viewing, which involved the face-to-face questioning of 2250 people each day by some 150 part-time survey workers all over Britain. The technique used by the interviewers, known as ‘One-day Aided Recall’, involved randomly selected members of the public who were asked what they had w...

Table of contents

• Cover
• Title
• Copyright
• Contents
• List of Tables and Figures
• Acknowledgements
• Notes on Contributors
• Deconstructing the Ratings Machine: An Introduction
• Part I Inventing Measurement
• Part II Appropriating Audience Figures
• Part III Confronting Changes
• Bibliography