The Production of Television’s Female Audience: Early BBC Audience Research and Gender Classification
by: Sarah Arnold , November 9, 2018
by: Sarah Arnold , November 9, 2018
In 1991, Ien Ang observed,
‘Quite obviously, before there was television, there was no such thing as a television audience. The television audience then was not an ontological given, but a socially-constituted and institutionally-produced category’. (Ang 1991: 3)
It is possible to see how the female audience became such an institutionally produced category at the BBC through analysis of its audience research carried out from 1936 onwards. The BBC first began collecting and studying data on its audiences with the establishment of the Listener Research Department (later, in 1950, Audience Research Department). The department deployed methods and techniques of measurement and classification that claimed truths and produced facts about the female viewing public. The tools and techniques used to assess the viewing public — shaping them into the ‘audience’ — were regarded as objective and scientific. (Meehan 1984 & Napoli 2014) Thus, the audience produced by such research was rendered as a quantifiable ‘fact.’
Nonetheless, this supposed scientific objectivity at the BBC relied on social discourses of identity, as well as on behavioural norms. When collecting information on the audience, the BBC interpreted it according to selected demographic classifications such as gender, class and age, and so the public became pre-defined according to certain criteria as opposed to others (i.e. gender rather than height). In addition, early BBC television audience measurement drew its interpretations about the audience from samples that were made up of those in possession of a television set, in other words, those who could afford this expensive new device. Thus, the knowledge that the BBC produced about its female audience was based on quite a narrow segment of the female population. Those early methods of collating data, in turn, informed how the BBC undertook television audience research in the post-war years, particularly in regards to its emphasis on the daytime female audience. According to the BBC, the female audience was imagined as a daytime one. However, it was exclusive of those working or otherwise occupied women who formed a significant part of the potential audience. The female audience, then, was less a fact and more a product of gendered institutional practices within audience research.
The BBC’s production of the female audience, as well as its use of such a category in programming and scheduling practices, points to some of the issues that have been raised about the concept of the audience more generally and the means by which it is put to work. Lisa Gitelman, for example, suggests that data is never ‘raw’ or neutral; rather it is ‘cooked’. (2013: 3) The process of segmenting the public according to gender already imposes a certain interpretive logic on the data. In addition, the more large-scale and quantitative the measurement is used, the more removed from social reality this institutional ‘knowledge’ becomes. In other words, when knowledge is formed through the use of the categories of ‘audience’ as well as such classifications, as for example, ‘male’ or ‘housewife,’ what is lost are the complexities, nuances and personhood of individuals.
Philip M. Napoli confirms, ‘efforts to enhance knowledge, predictability, and control in relation to the audience have … been accompanied by the kinds of analytical simplification that have historically been associated with the process of rationalization’. (2014: 7) For example, when audience research and measurement begins with the collection of data on women, it already assumes shared characteristics in this group of the population resulting in preconceived correlations between their viewing habits. It may seem inevitable that the viewing public might be understood and acted upon in these terms. Since gender is among the key markers of identity in society, the prevalence of these categories has materialised through decades of refinement of audience research methods.
We can note from the early, experimental years of television — between 1936 and 1939 when the war resulted in the cessation of broadcasts — that there was initially at least a distinct separation of audience research and programme planning and production. In other words, very early audience research did not have a direct impact on what was produced and scheduled for the ‘audience’ as constructed by the audience research department. Between 1936 and 1950 — at which point audience researchers began producing audience research reports on individual episodes of programmes — there was a more informal relationship between programme production departments and the audience research department. Indeed, Head of Listener Research at the BBC, Robert Silvey, had maintained that such a distinction was necessary:
‘From the very beginning of my time with the BBC, I constantly stressed that audience research’s function was limited to providing the decision-makers with information upon which they could act — or not act — as seemed to them right. Map-making and navigating were quite different functions. Ours was map-making’. (1974: 34)
By the 1950s, there was more communication and interaction between audience research departments and programme production departments as evident in the production of viewer reports that were issued to programme-makers from the 1950s onwards. And, later still, the relationship between audience measurement, programme planning and production became closer but was still not entirely integrated into an overall strategy for acting upon the findings of audience research reports, surveys or viewing reports. For example, if viewership of programme X fell below figure Y, then the programme would be cancelled or rescheduled.
The case of television audience research enables us to understand how the classification of the viewing public took root and was increasingly embedded in the institutional logic and practices of the BBC, firstly focusing on radio and later television. Looking back at archive materials from the 1930s through to the 1950s, it is possible to identify how and where gender became integrated into the knowledge production mechanisms of television audience research. Over time, as this knowledge was disseminated to various other planning and production departments, the ‘female audience, ‘the female viewer,’ ‘the housewife’ and ‘the mother’ come to represent what was imagined about, and inferred from, the viewing public. Early television audience research did not make sense of the audience in gendered terms, even though radio research was doing so. Instead, the measurement of gender was later introduced to television and this knowledge was put to use across the organisation, for example, in the development of women’s programmes, the scheduling of these programmes and the partitioning of programmes and audiences along gendered lines.
The analysis of the audience research carried out at the BBC reveals that audience research did not ‘discover’ that viewing habits, tastes and experiences were determined by gender. Rather the classification of gender entered into the lexicon of audience research reports and, as a consequence, extended into the institution more widely. Systems of classification did not, of course, begin with the BBC. The origins of systems and methods of categorising and classifying populations reveal that they long predated audience measurement. These systems of classification were devised by states, institutions and organisations to manage and control large populations, often in the context of health, employment and urban planning. As Michel Foucault suggests, the use of classifications brought about modes of understanding, of addressing, articulating and knowing the people and objects categorised. (Foucault 1979) The origins of such categorisations have been traced to the nineteenth century when there was a growth of
‘formal, commodified classifications and standards, both scientific and commercial. People classified, measured and standardised just about everything…Government agencies, industrial consortia, and scientific committees created the standards and category systems. So did mail-order firms, machine tool manufacturers, animal breeders, and thousands of other actors’. (Bowker & Leigh Star 2000: 17)
Anne Cronin follows Foucault to suggest that during the nineteenth century, the social world became classified according to a hierarchy of gender.
‘Gender can … be understood as a classifying matrix that facilitated, amongst other things, an ordering and hierarchizing of the social world. Women were classified and controlled according to their gender, class and race, but those taxonomies also operated as productive mechanisms to order the social field’. (Cronin 2012: 29)
The strategies for measuring audiences at the BBC and elsewhere stemmed from these earlier efforts to manage and control populations. This was evident in the way that BBC audience research produced the category of the ‘female audience’ despite the fact that the tastes and behaviours of female viewers did not always correlate with the institutionally produced ‘female audience.’
The development of audience research at the BBC from 1936 was, as Stefan Schwarzkopf notes, undertaken in response to press criticisms that the organisation was not representing the public interest and thus was not providing a valued service to its viewers. (2013) During the 1930s and early 1940s, the infant television industry was more inward-looking, working as it was to develop a form of broadcast that might suit the new medium. Any sporadic engagements with the audience were more for the purpose of using viewers as early testers of the medium. Feedback, often in the form of letters, enabled programme-makers to adapt or improve their broadcasts and acted as a barometer of tastes and interests. (Silvey 1974 & Napoli 2014: 10) As such, early attempts at audience research were less concerned with the profile of viewers and more with the reception of individual programmes and programme formats, as well as the general enthusiasm (or lack of) for television. The BBC, however, was keen to encourage viewer feedback. Through its publication, Radio Times, the BBC engaged with its audience by responding to individual reactions from the public and acknowledging the receipt of letters and phone calls, as noted an issue of the magazine from 15 April 1938. An interest in the composition of the audience would only materialise once television was viable as a mass medium and a business in the post-war years. In the pre-war years of the television service, any audience research undertaken formally and informally placed emphasis less on who the audience was and more on how the television service was received. Consequently, there was little reference to women as a category, market or social class of viewer. If daytime talks were popular in the afternoon, this was not immediately correlated with the female viewer unlike in the post-war years where the daytime schedule was determined as the primary terrain for ‘women’s programmes’.
When BBC television broadcasters began to formalise audience measurement in the post-war years, women emerged as a distinct group among the general audience. The significance of the gender composition of the audience was realised, and, in turn, it determined what sense was made of the viewing public. Subsequently, the female audience became isolated from the general audience. In other words, where programmes were viewed equally among men and women, they were understood in terms of genre. Where programmes were preferred by more women than men, they became ‘women’s programmes.’ Equally, when men and women watched at the same time, schedules contained programmes that were of interest to the general audience. However, when women watched at certain times more than men, this schedule came to be dominated by women’s programmes (instead of general interest programmes that women might have also enjoyed). While these might seem logical from the broadcaster’s point of view, it had the effect of partitioning women’s programmes from the broader schedule. It also produced a narrow understanding of the female audience, one which was interpreted as domestic and maternal, despite the fact that already back then many single or married working women did not fit this category. By 1947, for example, six million women were in the workforce and by 1951 women made up 30% of the full-time labour market. (Summerfield 2000: 13 & Elliot 1991: 86) Ultimately, the female viewer represented an institutionalised form of knowledge about how the viewing public was composed. This was based less on the real conditions of women’s domestic and non-domestic lives but on the social conventions and norms that held women as domestic and middle class. Such ideas about women formed the lens through which female viewers and data collected on them were interpreted. This had implications for what television would become including how it was scheduled, what types of programme and commercial content was broadcast and how the general public was addressed. Tracing the practices in television and radio audience research from its beginnings in 1936 to its formalisation in 1950 shows the extent to which the idea of the female audience was developed through a series of problematic surveying and somewhat inaccurate representation of the female viewing public.
During the experimental years of television any interest in, and reference to, the television viewing public was largely related to viewer responses to the quality of production, broadcast as well as image and sound. At this early stage of the television service, so little was known about the viewer that there was no attempt at segmenting an already small audience into categories such as male and female. Because of it, the BBC concerned itself more with the improvement of the technical operations of television broadcast and the development of television programmes. As the service developed, there was some concern about how the audience would respond to television and television was imagined as at the mercy of the public’s tastes. It must be remembered that during the early experimental years only approximately 100 households owned television sets, only increasing in fairly conservative numbers between 1936 and 1939. (Stanton 2012: 363) Private set ownership remained extremely low until the early 1950s. (Harper 2006: 70) The BBC was still concerned with the audience as it was eager to encourage more television sales and to develop the service. A Postmaster General’s report from January 1935, for example, referred to television being ‘put to the acid test of public opinion.’ (BBC WAC T23/108 Publicity Report on Television by the Postmaster General 31 January 1935) In an August 1936 report of the television service, the public’s reaction to television was of utmost importance, and the report suggested that the public would ultimately shape what television became. In a section on the reaction of the viewing public, it noted that
‘This is going to be extremely difficult to determine at first because we have to separate the interest due to novelty from that arising from genuine entertainment. It may be found that methods which have been laid down for ordinary broadcasting may be completely unsuitable’. (BBC WAC T23/108 Publicity Report on the Television Service from the Alexandra Palace by Controller (Engineer) 10 August 1936)
Here, the public was invested with a certain amount of power, and it was implied that their response to television would guide programme production. This comment also suggested a sense of powerlessness on the part of the BBC due to its lack of knowledge about its audience.
Such early concerns with the opinion of the viewers led to one of the first efforts to carry out audience research on the responses to television. In December 1936, the BBC put out a call to its small viewing public to engage them in a survey about its programmes. It received 74 completed responses, though the sample was largely composed of those who were already working or otherwise involved in broadcasting. (BBC WAC T1/6/1 Viewers and the Television Service: a report of an investigation of viewers’ opinions in January 1937 5 February 1937) Although carried out within the newly formed Listener Research Department, the Viewers and the Television Service study —published in 1937 — was more a survey of television in the public than who the television public was.
Nonetheless, the report provides some insight into how the BBC made sense of its audience. The report was based on a sample of respondents who were affluent enough to own a television set since the cost at this time was approximately 60-120 guineas (£63-£126) (Burns 1998: 592). This left it out of reach for those earning an annual wage of approximately £200 (BBC News 2005). The resulting survey findings, then, were based on a narrow representation of the general public.
The objectives of the survey included the following:
- To find out how many private viewing sets were in the hands of the public;
- To find out under what conditions the television programmes were being received;
- To find out viewers’ opinions on the television programmes;
- To find out the number of places where the television sets were installed for the purposes of demonstrating the service to the general public. (BBC WAC T1/6/1 Viewers and the Television Service: a report of an investigation of viewers’ opinions in January 1937 5 February 1937)
The report did not ask specific questions about individual programmes but it did allow space for comments, which was exploited by many of the respondents. Because the comments were quite broad-ranging, it was difficult for those interpreting them to produce any meaningful conclusions. In fact, their report noted that this had ‘its disadvantages as a method of obtaining clear guidance about the views of the television programmes’. (BBC WAC T1/6/1 Viewers and the Television Service: a report of an investigation of viewers’ opinions in January 1937 5 February 1937) Later research efforts would turn towards more quantitative methods to better manage the data. In the 1937 report, programmes—rather than viewers—started to be categorised. In this report responses to programmes were quantified in terms of genre and enjoyment. This was accompanied by summaries of general views about the programmes. Of particular note was the viewer feedback on the BBC’s instructional and educational programmes often centred on domestic chores, named in the report as Studio Demonstrations and Talks. Such programmes included titles like Quarter-of-an-Hour-Meals (9 December 1969), Accidents in the Home (8 January 1937) and Demonstration by the Women’s League of Health and Beauty (2 February 1937). Given their subjects and titles, the programmes might be assumed to be — but not yet identified as — ‘women’s programmes.’ The responses to them were largely unfavourable. Disapproval concentrated largely upon demonstrations of cooking, washing, ironing, etc., which were condemned as of little interest to those who could afford television sets. It was also pointed out that fashion parades were of little use given the absence of colour. (BBC WAC T1/6/1 TV Viewers and the Television Service: a report of an investigation of viewers’ opinions in January 1937, 5 February 1937) This gave the impression of an affluent television audience. Since television ownership at this time was confined to those of sufficient economic means or those with access to the public viewing rooms (where demonstrations of television programmes were provided), it is possible to assume that the demographic of the television audience was more determined by class and social status than by gender. The 1937 survey of 74 respondents was made up of both private owners of sets as well as those set owners who operated viewing rooms. Although the number of private sets was likely to be much higher than the number of survey respondents, it was far from being as ubiquitous as radio. Estimates for UK pre-war television set sales were 20,000. (Wood 2015: 345) These television owners and viewers became understood as the television audience by the BBC. In effect, then, the BBC’s first efforts at programmes were influenced by some general stereotypes about female interests including domestic chores, fashion and beauty while its post-1937 survey impression of the female audience was equally narrow: in this case limited to the affluent middle-classes.
Further television audience research continued in 1939 with a number of ad hoc measures to develop an understanding of the television audience. By that time, the Listener Research Department was already established and, although not a priority among others at the BBC, some further research was carried out on television. In the same year, the BBC held a television conference which invited 150 television viewers to ask questions of the Director of Television, Gerald Cock. The aim of this conference was to offer the public an opportunity to speak with and give feedback to BBC representatives. The two very different forms of interaction with the viewing public offer a good means of representing how and why formal audience research became the dominant means of developing knowledge of the viewing public in the post-war period. The television conference was more an enquiry of television by the public, whereas the Television Enquiry carried out by the Listener Research Department was a study of the viewers of the television service.
The television conference was held in June 1939, after the Television Enquiry interim report had been published; however, it represents some of the earlier ways that the viewing public was understood (or not understood, as the case may be). The conference was intended to provide a forum in which to inform television viewers of how television was developing and to enable them to ask the staff questions about plans for the service. Viewers were also able to make suggestions about what they expected of television. Both men and women raised questions about, and offered feedback on, the variety of programmes, the quality of television and the problems with broadcasts. While the intention of the conference was perhaps to showcase and promote the television service, the transcription of the conference suggests that it was a troublesome affair. (BBC WAC R9/9/6 Television Conference: Viewers’ Questions Monday 26 June 1939) Cock was quite defensive and dismissive when asked about the expansion of the service and the possibility of more programmes. In one exchange with a woman, who suggested the production of a ‘children’s hour,’ he agreed that this would be a valuable addition to the service but stated that it was not within the television department’s means to guarantee that this would happen. In fact, following the conference Cock wrote that future conferences should ban speeches from viewers and lamented that viewers had written to him to complain about his responses. (BBC WAC R9/9/6 Memo: Viewers’ Party and Questionnaire 28 June 1939) Such an experience perhaps made those involved in the production of television more cautious about directly engaging with audiences and it is, therefore, no surprise that further assessments of audiences took place by those directly engaged in more quantitative audience research.
The Television Enquiry of the same year, on the other hand, shifted focus from individual viewers to the viewing public and it is possible to see it the formation of the television audience in the earliest television surveys carried out. Using a sample of nearly 1,200 television set owners, the survey asked respondents a series of questions about what kind of sets they owned, their programme preference, as well as more specific questions about viewing habits. In comparison to later television surveys in 1948 and 1951, this research was again less concerned with developing knowledge of the composition of the audience and more interested in understanding general engagement in the service. While gender remained absent as a category of the viewing public, methods of quantification and segmentation were deployed. The interim report from 4 April 1939 introduced the research method and findings and quantified responses by percentage. (BBC WAC T1/11 Television Enquiry 1939 Interim Report 4 April 1939) In other words, percentages of total responses — as opposed to individual comments — became important ways of understanding audience behaviour. For example, in asking how many people watched television or how many preferred the use of intervals between programmes, the audience was classified in terms of simplified responses (yes/no). Unlike the television conference, there were only limited opportunities for viewers to account for or contextualise their responses. This represents one of the initial efforts to produce the audience using quantitative methods and to turn it into something far more manageable than the unruly viewers that participated in the television conference.
By the time the enquiry was completed and published in June 1939, classifications had entered into the system of measurement. (BBC WAC T1/11 LR/75 An Enquiry into Viewers’ Opinions on Television Programmes conducted in the first quarter of 1939, June 1939) For example, the 865 respondents were classified according to occupation, with a list of trades and professions detailed in the hand-written appendix, including the housewife.
The report was also concerned with the social class of television viewers and noted that television set ownership was not exclusively for the well-off but that ‘if the group is a fair sample, the audience is still predominantly middle class’. (BBC WAC T1/11 LR/75 An Enquiry into Viewers’ Opinions on Television Programmes conducted in the first quarter of 1939: introduction June 1939) However, while such data was collected in the appendix, it was not yet correlated with viewing interests or habits. What is of particular note in the report is the extent to which responses that would, in later years, be understood in terms of gender, were here accounted for as representative of the total population. For example, among the main preferences for programme type, the following are listed: ‘O.B.’s of Plays (or Variety) from Theatres; News Reels; ‘Picture Page’; Light Entertainment (Cabaret, Variety, etc.); O.B.’s of Sporting Events; O.B.’s of other outside events’ (BBC WAC T1/11 LR/75 An Enquiry into Viewers’ Opinions on Television Programmes conducted in the first quarter of 1939, June 1939). In later reports from 1948 and 1951, these were clearly accounted for in terms of male and female preference. However, in the 1936 and 1939 reports, they were classified only in terms of general popularity. Equally, in the 1936 report viewer requests that ‘demonstrations of cooking and fashions should be included in the afternoon and not in the evening programme’ were not interpreted in terms of gender. (BBC WAC T1/11 LR/75 An Enquiry into Viewers’ Opinions on Television Programmes conducted in the first quarter of 1939, June 1939) Where gender was raised it was in terms of the preference for male or female announcers — as in the 1939 report — rather than preferences of male or female viewers.
While it might be assumed that BBC audience research was rudimentary at this time, it is worth noting the extensive use of categorisation and classification in BBC radio listener research during the same period. In a number of listener reports between 1937 and 1942, there was an intense focus on the gender of the audience and the influence it was thought to have on viewing times and preferences. The 1937 Variety Listening Barometer, which surveyed the patterns of listening to various programmes during the day and throughout the week, gathered data on when women listened during the day as well as how this impacted on total listening figures. It also implied that women’s listening was inevitably higher than men noting that ‘the disparity is naturally considerable during the daytime’. (BBC WAC R9/9/1 A Report on the Variety (Light Entertainment) Listening Barometer October-December 1937 29 November 1937) The 1938 Variety Listening Barometer from March interim report similarly suggested that ‘naturally, since more women are at home than men, afternoon audiences are predominantly feminine’. (BBC WAC LR/67 Variety Listening Barometer Interim Report 5 April 1938) The Variety Listening Barometer April interim report of the same year noted that the ‘size of the feminine audience for daytime programmes is of special interest’. (BBC WAC LR/67 Variety Listening Barometer Interim Report No. 7. Vaudeville in National and Regional Programmes 14 March 1938) A September 1938 report on Winter Listening Habits produced much more data on the audience share by gender and correlated this with the programme preference and timing of programmes (BBC WAC LR/67 Winter Listening Habits: A Report on the First Random Sample Scheme 1 September 1938). By 1942, the BBC Listener Research Department had established specific audience panels of around 500 members per panel and aimed at gathering data on audience reactions to specific programmes. Along with two General Listening Panels and a Music Panel, this included a Women’s Panel ‘mainly for daytime programmes’. (Silvey 1974:114) Panel members were recruited not through general sampling but from responding to a radio call for volunteer panel members. Silvey himself noted the possibility of volunteer-bias since those volunteering would likely be keener listeners. (1974: 115) A 1942 report on Daytime Programme Repeats asked this Women’s Panel, rather than ones composed of general listeners, about preferences for these daytime programmes. (BBC WAC R9/9/6 A Listener Research Report Daytime Repeats 1942) In other words, in assessing the responses to daytime programmes, only women were invited to respond and these were possibly women who already enjoyed such programmes.
There are many reasons why there were differences in the extent to which radio listener research had developed where television audience research remained quite sporadic and unsophisticated. Radio broadcasting had been established for some years by the time that audience research formally commenced and had listeners in their millions. It also had a fairly regular schedule of broadcast and a consistent output of programme formats and genres. By comparison, television had a minuscule audience, a fairly sporadic schedule of programmes and, at this stage, an uncertain future. In other words, radio production and broadcast had become efficient and professionalised and, with this, the audience had equally become standardised. The radio audience was, thus, reduced to types and categories, which enabled more large-scale surveying of general trends and habits. This resulted in generalisations about audiences, nowhere more apparent in the often repeated ‘fact’ that women watched more during the daytime.
This assertion became common-sense and resulted in programme planning that ghettoised women and women’s programmes to this period of the day, despite the equally large female audience that listened at other times of the day. Television was yet to shape its audience and its programme schedule in this way. It was concerned with ‘everybody’ and what this ‘everybody’ preferred to watch. More rudimentary in its surveys, this was also more egalitarian, although it is possible to see how concerns with social composition were beginning to make their way into surveys. In the early years of television, then, the viewing public was a spontaneous, impulsive, changeable group that collectively expressed opinions and attitudes towards the television service and its programmes. Programme-makers were subject to the tastes, interests and behaviour of the audience and (middle-class) women had an equal stake as members of this viewing public. If, for example, a high percentage of the total audience favoured magazine programmes, this was not interpreted through the lens of gender rather it was significant in and of itself. As outlined further on, the more gender categorisation became a norm within audience measurement, the more this had implications for what types of audiences were valued or not. This was particularly evident in the post-war years when the television service returned and television audience research began to employ the methods and techniques of categorisation established within listener research.
The return of BBC television in the post-war years saw the organisation more eager for formal audience measurement. The Listener Research Department and the Television Service worked hard to make the case for any form of television audience research. At the same time, the Head of the Television Service and overall coordinator of television programmes, Maurice Gorham, made a number of requests for a means of gathering daily data on the patterns of viewing: a ‘television equivalent of the Daily Listening Barometer’. (BBC WAC R9/21 Memo from Maurice Gorham to Robert Silvey 29 January 1946) Gorham indicated that he would assist in the generation of a representative panel. Although Gorham and Silvey communicated about the possibility for television audience research, this was initially rejected by the Director General. Gorham’s persistence in pursuing this matter suggests the level of urgency felt within the Television Service for the need for an understanding of the television audience. Gorham implored the Director General to change his mind and suggested that it was difficult for the television department to plan productions without knowing what the audience was interested in. (BBC WAC T1/6/2 Audience Research Memos: Gorham asks DG to reconsider his refusal for television audience research 28 June 1946) In a letter to the Senior Controller, he noted that viewer letters had decreased and asked if he could include a closing announcement in a television programme to solicit feedback from viewers about their opinions on programmes. (BBC WAC T1/6/2 Gorham to Senior Controller on the Solicitation of Correspondence from Viewers 21 August 1946) In a memo to the editor of the Radio Times, he went so far as to ask whether it would be possible to have viewers publish their ratings in the magazine. (BBC WAC R9/21 Plans for Viewer Research: Memo from Head of Television Service (Gorham) to Editor, Radio Times 6 September 1946)
Similarly, Silvey began developing plans for television sample panels and worked on preliminary questionnaires for participants. (BBC WAC R9/21 Plans for Viewer Research: Draft Letter and Detail form to send to viewers willing to receive questionnaires 9 April 1946) He compiled a draft letter to potential respondents which began ‘we want to know as much as possible about our audience’. (BBC WAC R9/21 Plans for Viewer Research: Draft Letter and Detail form to send to viewers willing to receive questionnaires 9 April 1946) This knowledge would be gleaned from responses to questions about the make-up of the household as well as their age and gender composition. Unlike the earlier television surveys, newer surveys would concentrate much more on classifying the viewing public. By 1948 the Director General was finally confident enough in the future of television to permit a television audience inquiry and a number of initiatives were undertaken to survey the audience and its interest in programmes. (BBC WAC T1/6/2 Audience Research Memos: Director of Administration (Bottomley) writes to Silvey 6 July 1948)
The Listener Research Department undertook a number of reports on weekly viewing titled The ‘Viewers’ Vote’ Scheme as well as reports on the composition and viewing behaviour of the audience such as Television: Some Points about the Audience. (BBC WAC T1/6/2 Audience Research Memos: A Listener Research Report — Television: Some Points about the Audience 1 July 1948) The latter demonstrated a far more determined effort to produce a quantifiable and understandable audience where viewing activity could be read in terms of, and in relation to, social classification. Respondents were asked to identify ‘how many men, women and children usually watched television when the set was in use.’ (BBC WAC T1/6/2 Audience Research Memos: A Listener Research Report — Television: Some Points about the Audience 1 July 1948) The use of male and female categories was foregrounded in this report and it is clear to see that the gender segmentation was productive of different meanings about and interpretations of the audience. The gender of the audience was particularly important in the data on viewing times and frequency among the audience. The Frequency of Viewing table, for example, highlighted the differences in numbers of men and women watching at particular times during the day and across the week. The splitting of male and female viewers introduced a new interpretive logic to audience measurement whereby different sense would be made of some members of the audience in comparison to others. This is nowhere more evident in the data that suggested that weekday afternoon viewing was undertaken by 1.4 women in comparison to 0.4 men. In other words, daytime viewing became understood as a predominantly female activity. This resulted in the daytime audience been understood almost exclusively as the ‘female audience’ despite a significant number of male viewers watching at this time. In addition, while the number of women watching television during the daytime appeared large comparative to men, it was largely in keeping with general trends across the day. The table, thus, gives the impression that women were mainly daytime viewers when, in fact, women watched in fairly consistent numbers throughout the day and week.
The inquiry also noted that the viewer comments suggested an audience that was suburban, middle class and middle-aged. However, no effort was made to interpret the female audience in relation to these additional classifications and, for example, to consider the regional, age or class differences among the total female audience. In other words, the research gave the impression that all women watched a great deal of daytime television when, in fact, the survey sample was representative of those middle-aged, suburban and middle-class women. As Asa Briggs notes, by 1948 television set ownership was distributed among various social classes: 37% of more well-off, 12% of the population in Class 1; 34% of 20% of the middle-income population in Class 2; and 29% of the 69% lower income population in Class 3 (Briggs 1995: 230). In the BBC inquiry, those middle-class respondents came to represent women of all social classes regardless of the wider demographics of all female viewers. Equally, if the survey respondents were watching during the day, then this was taken as all women, regardless of the proportion of women that might have been elsewhere engaged in non-domestic work and not available to watch daytime television. Essentially, this erased any differences among groups of female viewers and produced a manageable, numerical object called the ‘female audience.’
This macro-level view of the audience was quite different to another report published in 1949. The Mass Observation organisation — founded in 1937 — undertook regular nationwide studies of the social life of the British public by gathering data through diaries and questionnaires completed by a panel of volunteers. (BBC WAC T1/6/2 Mass Observation Report on Television July 1949) In comparison to the BBC’s 1948 inquiry, Mass Observation’s Report on Television emphasised the diversity of television viewing amongst women as well as the similarities across social classifications. For example, it found that some housewives thought television a waste of time, where others found it a valuable educational resource. The Mass Observation report was based on a survey of 684 people among whom, as Helen Wood notes, only 2% owned a television set. (Wood 2015: 345) The survey respondents were, as Wood notes, largely ‘left-leaning and lower middle class because they would have had the time and inclination to commit to the project’. (Wood 2015: 345) Therefore, they were somewhat similar in class composition to the 1948 BBC inquiry. Despite this, the results of each were different in many respects. Among the female respondents to the Mass Observation survey, many of whom identified themselves as housewives, attitudes towards television were framed in relation to women’s identities as productive workers in the home. Some women were concerned that television would be a distraction from other duties and leisure pursuits. Others were concerned that it would confine them to the home and result in less opportunity to be away from the domestic sphere. In this sense, the survey captures what the BBC’s audience research did not: the reluctant female audience. The Mass Observation survey’s focus on the full and complex spectrum of experiences was in contrast to the reports produced within BBC Listener Research, which imposed order and consistency on audiences and worked to produce knowledge about and meaning from viewer experience.
By the 1950s, the notion that social categories such as age and gender influenced television viewing behaviour had become embedded in the BBC. Audience research continued to segment the viewing public by gender and to draw assumptions about viewing pattern based on this category. Viewing Panels (formed of a sample of households representative of the demographics of the British viewing public) were established and viewers issued with log books which asked them to identify their gender and age alongside their reactions to particular programmes. (BBC WAC R9/21 BBC Television Panel Log for Week 3 sample log for viewing week Sunday 15 January 1950) In his publication ‘Methods of Viewer Research Employed by the British Broadcasting Corporation’ Silvey maintained that social classifications such as gender and age were ‘all factors with which programme tastes are liable to be associated’. (BBC WAC R9/21 Methods of Viewer Research Employed by the British Broadcasting Corporation by Robert Silvey 13 December 1950) Although he conceded that there was ‘no invariable pattern’ and that ‘the tastes of men and women are frequently similar and frequently dissimilar,’ he nonetheless insisted upon the use of classification in determining some ‘basic facts’ about the audience. (BBC WAC R9/21 Methods of Viewer Research Employed by the British Broadcasting Corporation by Robert Silvey 13 December 1950 pp. 99-100) As the ‘female audience’ became a discursive object within the BBC and this was largely correlated with viewing time, attention to this audience shifted towards daytime viewing. Programme policy increasingly scheduled women’s programmes during the daytime rather than in the evening when more women— both in numbers and from different social backgrounds — watched. This was despite Silvey’s findings that ‘women do not want special women’s programmes every afternoon’. (BBC WAC R9/21 Silvey memo on report on women’s and children’s programmes 22 March 1950) Thus, while audience research was productive of institutional knowledge of the female audience, it continuously had to contend with anomalous and inconsistent behaviour in this viewing category. Research on weekday afternoon viewing, for example, noted that:
‘For the various women’s programmes an average of 15% of sets are in use but…there are wide variations. An occasional ‘Designed for Women’ has touched 29% while one ‘Health in the Home’ was as low as 5%. The ‘viewers per set-in-use’ figure is always lower for these programmes, for the obvious reason that they are directed at women who only constitute part of the public (albeit the major part in the afternoon). (BBC WAC R9/21 Memo from Silvey to Controller of Television Programmes 27 November 1950)
In other words, even when the use of the category ‘female audience’ did little to shed light on viewing patterns and viewer taste, the BBC Audience Research Department continued to deploy it as a meaningful category. This became significant in later years when, as Mary Irwin notes, women’s programmes were disappeared from the daytime schedule altogether which, in some ways, demonstrated an undervaluing of women by the BBC. (Irwin 2011)
Ultimately, the ‘female audience’ was not a social fact. Instead, it was a discursive object used within audience research to make sense of the viewing public. Indeed, during the early years of television, there was no ‘female audience’ but many female viewers. The creation of the ‘audience’ enabled the BBC to gain a sense of power in relation to what was once considered a mysterious viewing public. The creation of the ‘female audience’ allowed the organisation to map social inequalities onto programme policy with the result that the female viewer came to occupy as much a marginalised position in relation to the television service as she did in the social sphere. This was done first through the collapsing of vast numbers of culturally, geographically and socio-economically different individuals into one category of ‘woman’ and, secondly, by developing a programme and schedule strategy that segregated the ‘female audience,’ moving it to daytime schedules that would not interfere with the general audience. Ultimately, the use of sex classification within audience research resulted in the ghettoization of female viewers to specific time slots and specific genres. While female viewers, of course, were free to undertake whatever viewing they wished, the institutional production of a ‘female audience’ meant that women were addressed in gender-specific terms.
The production of the ‘female audience’ within early BBC audience research shows that it was not natural or inevitable that the viewing public would be defined and acted upon according to gender classifications. Neither was it inevitable that the gender of the viewing public would play so central a role in shaping the programmes and schedules of the television service. However, the methods deployed in measuring the audience resulted in this and an investigation of them reveals the mechanisms by which the viewing public came to be understood as an ‘audience.’ The documents within BBC Written Archives help shed light on the means by which this materialised and is a valuable source in tracing the ways in which gender materialised and materialises in screen industries. The history of BBC audience research offers a useful foundation and a valuable resource for understanding how, why and where gender is structured and shaped within contemporary screen experiences.
Archival materials were accessed at the BBC Written Archives in Caversham, UK
This research was funded by the International Federation of Television Archives.
Ang, Ien (1991), Desperately Seeking the Audience, London: Routledge.
Bowker, Geoffrey C. & Leigh Star, Susan (2000), Sorting Things Out: Classification and Its Consequences, Cambridge, MA: MIT Press.
Briggs, Asa (1995), The History of Broadcasting in the United Kingdom: Volume IV, Oxford: Open University Press.
Cronin, Anne (2012), Advertising Myths: The Strange Half-Lives of Images and Commodities, London: Routledge.
Demonstration by the Women’s League of Health and Beauty, (1936-1939) created by the BBC.
Elliot, B. Jane (1991), ‘Demographic trends in domestic life, 1945-87’, in David Clark (ed), Marriage, Domestic Life and Social Change: Writings for Jacqueline Burgoyne 1944-88, London: Routledge, pp. 71-93.
Foucault, Michel (1979/1994), The History of Sexuality, Volume 1: An Introduction, London: Allen Lane.
Gitelman, Lisa (2013), Raw Data is an Oxymoron, Cambridge, MA: MIT Press.
Harper, Richard (2006), Inside the Smart Home, New York: Springer.
Irwin, Mary (2011), ‘What Women Want on Television: Doreen Stephens and BBC Television Programmes for Women, 1953-64’, in Westminster Papers, Vol. 8, No. 3, pp. 99-122.
Meehan, Eileen (1984), ‘Ratings and the Institutional Approach: A Third Answer to the Commodity Question’, in Critical Studies in Mass Communication, Vol. 1, No. 2, pp. 216-225.
Napoli, Philip M. (2008), ‘The Rationalization of Audience Understanding’, The Donald McGannon Communication Research Centre. Fordham, New York: Fordham University.
Napoli, Philip M. (2014), ‘On Automation in Media Industries: Integrating Algorithmic Media Production in Media Industries Scholarship’, in Media Industries, Vol. 1, No. 1, pp. 33-38.
Schwarzkopf, Stefan (2014), ‘The Politics of Enjoyment: Competing Audience Measurement Systems in Britain, 1950-1980’, in Jerome Bourdon & Cecile Meadel (eds), Television Audiences Across the World: Deconstructing the Ratings Machine, Basingstoke: Palgrave, pp. 33-52.
Silvey, Robert (1974), Who’s Listening?: The Story of BBC Audience Research, London: George Allen & Unwin Ltd.
Stanton, Andrea L. (2012), ‘Television’, in Andrea L. Stanton, Edward Ramsamy, Peter. J Seybolt and Carolyn M. Elliott (eds), Cultural Sociology of the Middle East, Asia, and Africa: An Anthology, London: SAGE, pp. 362-365.
Summerfield, Penny (2010), ‘’It did me good in lots of ways’: British Women in Transition from War to Peace’, in Claire Duchen & Irene Bandhauer-Schoffmann (eds), When the War Was Over: Women, War, and Peace in Europe, 1940-1956, London: Leicester University Press, pp. 13-28.
Wood, Helen (2015), ‘Television- the Housewife’s Choice? The 1949 Mass Observation Television Directive, Reluctance and Revision’, in Media History, Vol. 21, No. 3, pp. 342-359.
Accidents in the Home. 1937 [TV]. BBC. 8 January. 3pm.
BBC News, (2005), ‘Pre-war Britons ‘were happier’’, BBC News Online, 1 September 2005, http://news.bbc.co.uk/2/hi/uk_news/wales/4203686.stm (last accessed 14 May 2018).
Quarter-of-an-Hour-Meals (1936) created by the BBC.
WHO SUPPORTS US
The team of MAI supporters and contributors is always expanding. We’re honoured to have a specialist collective of editors, whose enthusiasm & talent gave birth to MAI.
However, to turn our MAI dream into reality, we also relied on assistance from high-quality experts in web design, development and photography. Here we’d like to acknowledge their hard work and commitment to the feminist cause. Our feminist ‘thank you’ goes to:
Dots+Circles – a digital agency determined to make a difference, who’ve designed and built our MAI website. Their continuous support became a digital catalyst to our idealistic project.
Guy Martin – an award-winning and widely published British photographer who’s kindly agreed to share his images with our readers
Chandler Jernigan – a talented young American photographer whose portraits hugely enriched the visuals of MAI website
Matt Gillespie – a gifted professional British photographer who with no hesitation gave us permission to use some of his work
Julia Carbonell – an emerging Spanish photographer whose sharp outlook at contemporary women grasped our feminist attention
Ana Pedreira – a self-taught Portuguese photographer whose imagery from women protests beams with feminist aura
And other photographers whose images have been reproduced here: Cezanne Ali, Les Anderson, Mike Wilson, Annie Spratt, Cristian Newman, Peter Hershey