'any research that might be undertaken should be so controlled as to secure that it never developed from a servant into a master, to the detriment of the essential qualities of good broadcasting ? a responsible but sensitive outlook and a readiness to experiment' (G.A.C. 23, memorandum presented by Sir Stephen Tallents to the BBC General Advisory Council, January 1936.)
'Audience measurement, properly used, can be a good servant; but it is a bad master? the fate of the battleship Potemkin shows what happens when the ratings take over.' (Robert Silvey, head of BBC audience research 1936-1960)1
The British Broadcasting Company, later British Broadcasting Corporation (BBC), first took to the air in November 1922. It was soon an indispensable part of British life and, incorporated by Royal Charter in 1927, a model of public service broadcasting for the rest of the world. Yet it was not until 1936 that the BBC began systematic research into its own audience, establishing a Listener Research Section in the Home Intelligence Department of its Public Relations Division. On the outbreak of war in 1939, the Listener Research Section was expanded and renamed the BBC Listener Research Department (LRD); in 1950 the LRD was rechristened the BBC Audience Research Department in deference to the growing importance of television. During these first fourteen years, Listener Research provided BBC policy executives and programme makers with a remarkable record of the audiences for radio (and early television) broadcasting in Britain, as well as a window onto the views, tastes and listening habits of the British public. Available to historical researchers on a limited basis since the opening up of the BBC written archives in the 1970s, the records of the BBC Listener Research Department 1937-c.1950 have now for the first time been published in microfilm and electronic form, as the first part in a collection of BBC Audience Research Reports in the BBC Written Archives Collection series. This introduction to the first part in the collection of BBC Audience Research Reports seeks to place the origins of BBC Listener Research in context, to outline the history of Listener Research from its early days through the war years, and to provide an overview of the wealth of material that this collection offers to researchers.
The Origins of BBC Listener Research
It is a remarkable feature of the early BBC that its broadcasters worked in almost complete ignorance of their audiences. None of those involved in making programmes for the British public had any systematic knowledge of the audiences for whom they were making these programmes nor the reactions of those audiences to the programmes they had made. As the BBC's first Listener Research Director, Robert Silvey, himself pointed out, unlike other inter-war mass media, the cinema or the newspaper press, broadcasting had 'no box office, no sales figures'; the 'link between microphone and listener ? carried only a one-way traffic'.2 Yet in fact the BBC was not uniquely slow to adopt techniques of market research. Although social surveys had been conducted in Britain since the late nineteenth century (notably by Booth and Rowntree), and continued throughout the inter-war years, their primary focus was urban poverty. By contrast, the first professional readership survey of the British press took place as late as 1931, conducted by the advertising agency London Press Exchange in association with the News Chronicle newspaper; the British Institute of Public Opinion (BIPO), the first British public opinion agency (an offshoot of the American Gallup organisation), was not founded until 1937; and the first by-election poll was carried out for the Labour Party in West Fulham by Tom Harrisson and Charles Madge's equally new Mass-Observation organisation in 1938. But while American radio in this period, for instance, had a powerful commercial interest in learning about its audience, the BBC's Director General, Sir John Reith, famously had an equally powerful philosophical aversion to the very concept of audience research. As public service broadcasters, the BBC's obligation, as he saw it, was to provide something for everyone through the broadcasting schedule as a whole, and not simply to satisfy the majority at any specific time. Market research on radio audiences threatened to drive broadcasters into the pursuit of ratings at the expense of quality and diversity; a BBC in thrall to audience research would surely end up broadcasting what people already liked or wanted, and not ? what Reith believed to be one of the BBC's greatest educative functions ? what they might come to like, or might even discover for the first time. In this he had the support of most if not all of his colleagues.3
There were, of course, various channels of audience information available to the BBC: its own Advisory Committees, public meetings, personal contacts ? and of course listener correspondence, which was considerable (over 50,000 items a year by 1927) and heartfelt, if not necessarily representative.4 Certainly, the BBC took its listener correspondence seriously: letters were almost always replied to, and summaries of correspondents' opinions were passed on to Programme departments on both a daily and weekly basis. The BBC also solicited listener opinion on specific issues, either through microphone appeals or via notices in the Radio Times; it periodically employed paid advisors to act as 'programme consultants'; and on at least one occasion it conducted a house-to-house audience survey. Newspapers had begun printing listener 'preference polls' as early as 1927, and carried out their own listening surveys (for instance the Daily Express during the General Election of 1935); they also ran regular and opinionated columns by radio critics who professed to represent the voice of the listeners, and who were often unsparing in their criticism of the BBC.5 Listeners' associations and the Radio Manufacturers' Association also collected listener opinions and forwarded them to the BBC. However, the BBC's apparent refusal to engage formally with listener opinion contributed to its growing reputation as aloof, self-satisfied, and lacking the common touch.6
Despite Reith's misgivings, there was considerable internal debate within the BBC about the value of listener research. In May 1930, Charles Siepmann of the Talks Department and Val Gielgud (brother of Sir John), Director of Features and Drama, both of whom had long expressed frustration at the lack of substantive information about the audiences for their programmes, strongly argued for a more systematic approach to audience research. In response to those who feared the tyranny of crude audience figures, Siepmann argued that what was required was not a 'statistical' but a 'sociological' approach.7 This point was subsequently reinforced by Hilda Matheson, former BBC Director of Talks (1927-31), who, writing in the Sociological Review in 1935, argued that information about the modes of presentation that audiences preferred were just as important as information about the types of programme they most listened to; further, that 'more attention was being paid to social change in primitive societies than to the social effects of radio'. One of the BBC's principal programme consultants, Filson Young, underlined this point, arguing that what was needed above all was an understanding of the role that listening itself played in people's lives.8
In January 1936 the BBC's General Advisory Council finally decided to institute a (deliberately small-scale) programme of specialised listener research. Their decision was prompted in large part by Sir Stephen Tallents, recently appointed BBC Controller of Public Relations, who in his previous posts as Secretary of the Empire Marketing Board and then Public Relations Officer of the General Post Office, had transformed both their publicity operations through a pioneering use of advertising and, in particular, documentary film. The new 'Listener Research Section' was placed under his supervision. With no specified terms of reference beyond a general injunction that its findings would not be used to determine future policy, the BBC Listener Research Section was inevitably going to be very much the creation of its first Listener Research Officer. Tallents' choice, Robert J. Silvey, was an inspired one, and it is to Silvey that the entire BBC audience research operation owes both its character and its success.9
Despite his relative youth (he was just 31) and lack of experience in broadcasting,10 Silvey's qualifications were about as strong as they could be for such a new venture. He had previously been employed in the statistical department at the London Press Exchange (LPE), where he had authored the News Chronicle's path-breaking 1931 Reader Interest Survey of the British Press and had also first come into contact with Tallents' deputy at the BBC, A.P. (Patrick) Ryan, then Publicity Manager of the Gas Light and Coke Company, one of the LPE's major clients. Silvey also had experience of studying radio audiences, having previously compiled a survey of listening to the continental radio stations, Luxembourg and Normandie, English-language commercial broadcasters whose advertising was directed at British listeners and who during the 1930s ? particularly on Sundays ? regularly attracted British audiences in the tens of millions. He took up the post of BBC Listener Research Officer on 1 October 1936, with the broad remit, as he later described it, to 'gather such information about the public as was at once relevant to the needs of the BBC and susceptible of study by the methods of social research and to present this as succinctly and clearly as possible to those whom it concerned.'11 Which programmes should, or should not, be investigated was largely up to the Listener Research Section (i.e. Silvey) itself.12 It was also left up to Silvey to devise an appropriate research methodology (of which more below). However, it was 'emphatically not part of audience research's role to apply [its] findings ? or even to decide whether they should be applied or not'. This explicit separation of audience research from programme-making was crucial to its initial acceptance within the BBC; as Silvey later put it, 'Map-making and navigating were quite different functions. Ours was map-making'.13 Listener research findings were to be made freely available to BBC staff, but were considered to be confidential to the BBC and not to be published or otherwise disseminated in any way except in certain narrowly defined circumstances, usually publicity-related.14
Silvey's first excursions into audience research were cautious but characteristic. After an initial foray into the field with a report on the embryonic BBC television service,15 it was decided to institute a major survey of Features and Drama, a popular and high-profile genre of radio programming, whose head of department, Val Gielgud, had been one of the most persistent advocates of audience research.16 Unable for reasons of cost to run a full interview survey using orthodox sampling methods, it was decided instead to recruit a Drama Panel, made up of ordinary listeners with a declared taste for plays and features, who over a three-month period would complete questionnaires about each production of the Department of Features and Drama to which they had listened (a possible total of forty-seven productions). The aim of the questionnaires at this stage was an essentially qualitative one: to discover listeners' opinions of what they listened to. 350 panellists were recruited from volunteers, from people who had previously written to the BBC on drama-related matters, and from personal contacts. The questionnaires offered a series of responses to specific questions broadly based around each programme's theme, production and performance, plus space to comment further. Silvey was well aware that the function of the questionnaire as a 'prompt' to listen was a key methodological weakness of this survey approach, and respondents were strongly requested to listen only to those broadcasts they would have listened to even if they had not been members of the panel (the injuction 'No Duty-Listening please' became a recurrent phrase in Listener Research appeals). Despite reading in some places as a catalogue of criticisms, the final report confirmed many of Gielgud and his staff's own thoughts about the current state of radio drama17 and was overall deemed so successful that Reith himself congratulated the Listener Research Section on its work. A series of listener panels for other types of programmes (notably Talks) was swiftly established, to be run on the same basis. This type of research became known as the Listening Thermometer (i.e., it registered the 'temperature' of the audience response to the programme'). The most revolutionary part of this report, however, was one of its final conclusions: that in some circumstances, the principal need was for some way of measuring not simply audience opinions about a programme but the actual size of that audience. That this conclusion was accepted without demur demonstrates how far LR had gone in such a short time.
Silvey's second experiment, therefore, was to set up an appropriate quantitative audience research methodology. The rapidly developing American research methods into radio audiences (the 'simultaneous telephone calls' method, the 'recall' method' and the Neilson 'meter' method) he rejected as too one-dimensional and unreliable.18 However, sampling, and extrapolating figures from the sample, was still considered by many within the BBC to be a highly dubious research methodology. Silvey's compromise was to draw up what he termed the Listening Barometer (i.e., it measured 'pressure rather than heat'), setting out to measure not a monolithic 'radio public' per se but the variety of 'radio publics'.19 The first such 'barometer', the Variety Listening Barometer, was established in the autumn of 1937, and comprised 2000 listeners randomly recruited from an astonishing 47,000 volunteers after a microphone appeal by Director of Variety, John Watt. These volunteers completed weekly logs specifically listing the Variety programmes they had listened to over the previous week (again, 'No Duty-Listening, please!'). After interviewing a parallel control group it was discovered that while the members of the Variety panel were, unsurprisingly, atypical of the average listener in the extent of their listening to Variety programmes, they were in fact quite representative in their relative choices of programme.20 The panel findings could therefore be used to register listener choice by estimating relative if not absolute audience size. On this basis, LR was permitted to extend the experiment and establish a General Listening Barometer of each day's listening, based on weekly returns from 4,000 volunteer listeners.
Silvey's third Listener Research experiment was perhaps the most far-reaching, focussing as it did not on specific instances of listening but on listening habits more broadly defined. Over the winter of 1937/8 LR sent short (single-sheet) questionnaires to a random sample of 3,000 households from the GPO's list of wireless licence-holders, requesting at what times of day respondents listened to the radio, and information about what kinds of programme they liked at particular times of day and night. Questionnaires were returned from 44 per cent of households, the first national survey of listening habits ? and by extension, living habits ? of its kind.21 The findings demonstrated a pattern of daytime listening that peaked between 1 and 2pm, then again between 5 and 6pm. The overwhelming majority of respondents preferred light music in the middle of the day, with variety, plays and talks in the later afternoon, and the audience remained steady in the evenings till around 10pm (later on Saturday nights) after which time it rapidly faded. In many respects, again, the findings were primarily useful in simply confirming what many programme-makers had long suspected; however, members of the BBC senior management who had long argued that 'nobody dines before 8' were shocked to learn that in fact most people had finished their evening meal before 7pm. A second survey,22 conducted the following summer to compare seasonal listening habits, went further, including for the first time the question, 'What do you like?' and listing twenty-one types of programme to choose from. This, the first census of broad 'tastes' in listening ever carried out by the BBC, confirmed that overwhelmingly the most popular output was Variety (93 per cent), followed broadly by different categories of light music; plays, talks and discussions; sports commentaries; and classical music (with chamber music last at 8 per cent). Age, class and family size were sharp discriminators of taste: broadly, the younger or the more working-class the respondent, and the larger his/her family size, the more likely they were to prefer Variety and light music and the less talks, discussions and classical music. However, regional and rural/urban differences in tastes appeared to be mostly negligible, as ? to the report's obvious surprise ? was gender (with the exception of cricket commentaries, liked by 66 per cent of men but only 30 per cent of women). These kinds of background information survey became the third key area of BBC audience research.
Silvey's work in the pre-war Listener Research Section achieved a huge amount in a relatively short space of time. Against considerable scepticism from some parts of the BBC, he evolved a research methodology that broadly satisfied everyone, principally by downplaying any emphasis on crude audience figures and opting instead for a more impressionistic exploration of listeners' opinions, relative preferences and listening habits. He had also demonstrated that gathering information about listeners' habits, tastes and preferences did not and would not necessarily lead the BBC to surrender to majority opinion. This highlighting of the 'sociological' rather than the 'statistical' ? what Siepmann had urged in 1930 and Matheson had reiterated in 1935 ? put the BBC Listener Research Section very much in the mainstream of social research in Britain in the late 1930s. The parallels with Tom Harrisson and Charles Madge's Mass-Observation movement, founded in 1937, are particularly compelling. Harrisson and Madge championed a quasi-anthropological research methodology that sought to explore British attitudes and behaviour by volunteer-based 'observation' in much the same way that Silvey sought to build up a qualitative picture of listeners' tastes and patterns of listening. Hilda Jennings and Winifred Gill's impressionistic study of the impact of broadcasting on a working class area of Bristol, Broadcasting in Everyday Life (1939), commissioned by Silvey, has distinct parallels of tone with Harrisson and Madge's Penguin Special Britain, by Mass-Observation, published almost contemporaneously.23 It took the Second World War however fully to vindicate Silvey's approach, and to give him scope both to consolidate his research and to extend it into previously forbidden territory.
The LRD in Wartime
BBC Listener Research came into maturity during the war years. The war was of course both an unprecedented challenge and a unique opportunity for the BBC.24 First, with normal modes of entertainment curtailed (by blackout, closures, travel restrictions, etc.), the radio was the principal source of escape from the travails of daily life. Second, in a war in which government needed above all to communicate with people, the BBC was the most direct ? and versatile ? means of providing news, advice and information. Third, with the continental English-language stations closed down, the BBC's principal broadcasting rivals now were the enemy stations broadcasting light music and pro-German propaganda from the continent to Britain. In all three cases, the BBC needed Listener Research to help it carry out its most fundamental Reithian remit to best advantage: to tell it how best to 'inform, educate and entertain' the British people at war ? and in doing so, how best to counter and ultimately defeat the enemy. And just as Mass-Observation, for instance, gained official legitimacy from its wartime association with the government's Home Intelligence operations, so BBC Listener Research was transformed into an indispensable tool of war.
According to the BBC's initial emergency wartime plans, the Listener Research Section was designated of secondary importance ('Category B') and effectively put on standby until further notice. However, within a matter of weeks Silvey recalled his staff, convinced that audience research had an urgent role to play. Silvey used the (less than encouraging) findings of a special listener report on the wartime Home Service conducted in late September to argue for a substantial extension of the remit and personnel (from seven to ten) of the Listener Research Section.25 The practical usefulness of LR was further underlined that same month by a special survey of 3,800 urban householders and their lunchtime habits. The point of particular concern was whether the new midday news bulletin should be moved from 12pm (where it had been scheduled since the start of the war) to 1pm in order to gain the maximum possible lunchtime audience.26 Contradicting both press opinion (which denied the utility of lunchtime news at all) and well-meaning but mostly ill-informed advice from both BBC senior management and government officials (several of whom cited their own domestic staff as a representative sample of the working man and woman), the survey established conclusively that, of those working people who went home for their midday meal, 57 per cent were home at 1pm but only 16 per cent at home at noon; further, that while there was an avid audience for news at 1pm, most housewives were too busy preparing dinner at noon to listen to the radio.27 The lunchtime bulletin was therefore moved to 1pm, in which time-slot (as LR later confirmed) it would reach an audience of up to 40 per cent, well over 13 million listeners.28 The Listener Research Section itself was subsequently transferred from the BBC Public Relations Division to Programme Division and elevated to the status of Department. Silvey became Listener Research Director (though not Director of Listener Research, in the BBC's arcane administrative code a far more prestigious title).29 In the circumstances, no one pointed out that this concern to maximise potential audience, not to mention the specific use of Listener Research to change the schedule, represented a startling breach with BBC tradition.
The wartime importance of LR was again highlighted when the Listener Research Department was asked by the Ministry of Information (MOI) (to which the wartime BBC was answerable) to conduct a special report into British listening to the enemy station, Radio Hamburg. In the early months of the war, when BBC news was subject to stringent government censorship, it had soon become evident that a substantial element of the British population was tuning for war news to any available source, including enemy stations. Premier among these were Radio Hamburg's broadcasts in English, in particular William Joyce's 'Views on the News', broadcast on Sunday nights immediately after the BBC's own 9 o'clock news. Anecdotal evidence suggested ? and the MOI was convinced ? that Joyce ('Lord Haw-Haw' as the British press famously dubbed him) was exerting a terrible influence on his listeners, fomenting rumour and defeatism among the most susceptible members of society. LR's report into the Hamburg broadcasts, compiled over the winter of 1939/40 and published in March 1940, concluded instead that although Joyce's broadcasts had at one point attracted a startlingly large British audience of up to ten million, most listened out of curiosity, out of a perhaps misplaced search for wireless 'entertainment', or out of sheer frustration with the 'official' news. The most assiduous listeners tended to be well educated and politically interested ? not the credulous mass of the MOI's imagination. And, most reassuringly, the novelty had worn off: listening to Haw-Haw had already peaked and was now in decline.30 This report like no other underlined the sheer scope and potential of BBC listener research: in providing substantive listening figures, in setting those figures in context, and in drawing important social and political conclusions from its findings. It established the BBC as an authority on public behaviour and public opinion. It challenged government policy, with its implicit critique of the MOI's policy on the censorship and release of news. And it also, again, had a direct influence on programming: the BBC's celebrated Sunday Postscripts to the News, in which the playwright J.B. Priestley found wartime fame, would begin life as a 'spoiler', deliberately scheduled against Joyce to draw away what remained of his audience.31
The progressive emboldening of the Listener Research Department can be seen in how it redefined both its role and the scope and methods of its research throughout the war. As early as December 1939 Silvey had taken the decision to reconfigure the weekly General Listening Barometer as a continuous daily survey of listening, with a specific remit to measure audience figures for all programmes across the schedule ? thus marking both a decisive move into 'statistical' audience research and a confirmation of the validity of sampling as a research technique. This new General Listening Barometer (which appears as the Survey of Listening: Listening Barometer from February 1944) was compiled from face-to-face interviews with a random sample of the population (a different sample every day) asked to itemise their previous day's listening. The daily sample was 800 interviews, weighted by region, urban/rural area, gender, age group, whether 'occupied' or 'unoccupied', and by social class, and was designed to represent the entire adult civilian population (over sixteen years of age) of Great Britain. Audience figures were usually represented as percentages, with every 1% deemed as representing approximately 330,000 listeners.32 Interviewing was initially subcontracted to the British Institute of Public Opinion, but was taken in-house in 1943, with the LRD recruiting its own pool of part-time interviewers.33 In addition, interviewers were often given extra questions (EQs) that might relate to any current issue of broadcasting concern, from the appropriateness of Variety programmes at times of bad news to vulgarity in BBC comedy. They also asked a general question whether the respondent was dissatisfied or satisfied with BBC programmes. This 'satisfaction index' became known as the BBC Thermometer.
In addition, in order to assess public opinion about broadcasting more broadly, Silvey was already recruiting a network of around 2000 'Local Correspondents'. These unpaid correspondents submitted monthly questionnaires on a range of issues to which the BBC sought information, from programme content and scheduling to more general questions. Using broadly the same methodology pioneered by Mass-Observation, correspondents were asked to report, not just their own thoughts, but the comments and attitudes of the people with whom they came into daily contact, in order to build up an impression of the prevailing state of opinion around them. During the war the Local Correspondents were used to report on a range of issues, including opinions about specific programmes or types of programme, styles of presentation, listening habits and lifestyles, as well as opinion about wider public issues more generally, sometimes only tangentially related to the war (see below). A separate panel of Forces Correspondents was also established, specifically to report on servicemen and servicewomen's opinions of the BBC Forces Programme.
The third wartime means of assessing BBC audiences was the network of Listener Panels, set up in 1941. These were groups of 500 volunteers each, who professed a strong preference for a particular genre of programming (thus the Music Panel, the Plays Panel, etc.). They filled in weekly questionnaires detailing their responses to broadcasts in their preferred programme genre, and also awarded marks out of ten for each programme concerned (the average tally, expressed as a percentage, became known as the Appreciation Index). All three kinds of finding were then collated and circulated internally, as combined Listener Research Weekly Reports (also known as Listener Research Bulletins), and also in the form of a monthly News Letter.
Clearly there were methodological problems with all of these approaches. For instance, the daily Listening Barometer relied on memory (albeit only of the previous day's listening) and the interviewer's prompting. It also might not include emergency or special broadcasts scheduled too late to appear on the interviewers' log sheets. The most famous ? and regretted ? omission (this time, for security reasons) was the audience figures for the first week of War Report, the nightly news magazine that broadcast front-line coverage of the Allied invasion of Europe almost without a break from D-Day to VE-Day.34 However, where possible interviewers were notified of last-minute changes by telegram, which is presumably how the Listening Barometers managed to record listening figures for Winston Churchill's broadcasts as Prime Minister, many of which were last minute additions to the schedule.35
Wartime conditions threw up other problems. On one infamous occasion in August 1940, the Listening Barometer recorded a Forces Programme audience one Friday night at 10pm of 11.7 per cent listening to Horace Finch at the Cinema Organ, with 5.5 per cent for the programme which followed it at 10.40pm, a gramophone Cabaret entitled Lights Out; unfortunately, the entire service had been shut down at 10.06pm because of an air raid and no one could possibly have been listening to either programme after that time. This does, though, appear to have been a rare error, generated by highly unusual circumstances, and sufficiently unusual to have prompted embarrassed comment at the time.36 Silvey would always maintain that the barometer's margin of error was generally no more than +/- 0.8 per cent, and that it remained 'a sufficiently reliable guide for the practical purposes of informing those responsible for directing programme policy'.37 LR's measure of general listener satisfaction with the BBC, the Thermometer, was less reliable, Silvey discovering that it fluctuated above all with the state of the news of the day.
Moving beyond the Listening Barometer, Local Correspondents' Reports had their own in-built biases, in particular the relatively narrow base of their social milieu (a problem shared by Mass-Observation, although the BBC's Local Correspondents probably represented a wider political base), and similarly relied for their accuracy on the truthfulness of respondents. In the case of the Hamburg broadcasts, for instance, it has been plausibly argued that actual listening to Haw-Haw was less widespread than reported by LR in the early months of the war (i.e., when it was 'the thing' to say one listened to him), but far more widespread later on (when to admit to listening became regarded as almost treasonous).38 As for Listening Panels, Silvey acknowledged the problem of volunteer bias in this mode of sampling, but argued that as the keenest listeners to these kinds of programmes the Panellists provided a possibly exaggerated but not otherwise distorted picture of audience opinion. Despite these drawbacks, when taken together these various modes of audience assessment provided programme makers with a uniquely detailed and nuanced picture of wartime life and listening. In this way Silvey ensured that audience research maintained a key role as an integral part of departmental and programme self-assessment and planning.
Wartime Listener Research Findings
The breadth and range of Silvey's wartime audience research interests ensured that LR's wartime findings provide not merely a record of wartime radio listening, but also a vivid and unique insight into home front life in wartime Britain. This can be seen in a variety of subject areas.
First, of course, the listening figures themselves. Through the Listening Barometers and the Listener Research Bulletins, one can trace both daily and weekly patterns of listening and the programmes that received the highest (as well as the lowest audiences) throughout the war. The importance of news bulletins, the extraordinary audiences for such iconic wartime programmes as ITMA (40 per cent of the population by 1944), The Shadow of the Swastika (over a third, some twelve million, during the winter of 1939/40) or The Kitchen Front (five to seven million listeners every morning), can all be identified here, as can what was probably the highest recorded audience during the war, a statement from the King at 9 o'clock on D-Day, 6 June 1944, just before the main news of the day, heard by a staggering 80 per cent of the adult population.39
Second, listener opinions about BBC programmes. Local correspondents' reports, panel reports and special investigations all provide insight into what listeners thought about the programmes they were given: what it was about certain programmes, or presenters, that made them popular; why, for instance, the 'sports' commentary of a dogfight over the straits of Dover by Charles Gardner in July 1940 struck such a chord in listeners, or why the austere socialist, Sir Stafford Cripps, was deemed the highest-rated political speaker (aside from Churchill) of the whole war. These reports indicate just what made so many iconic wartime programmes so hugely popular with listeners ? The Brains Trust, for instance, with its unique combination of high- and low-brow entertainment, its frequent dabbling with controversy and its potentially explosive political mix, the subject of regular annual special reports (though strangely ITMA, about whose remarkable and incontrovertible popularity everybody had a view, was never investigated by LR). But Listener Research Reports also record (sometimes in pungent detail) the unpopularity of programmes: witness, for instance, the devastating listener response to the BBC's highly propagandist programmes for factory workers From Factory to Front Line and Award for Industry (later retitled Worker of the Week):
When the chaps in my works get a rotten job or feel 'browned off' they often mimic these programmes, 'Ho yes, I love my work', 'Ho no, I never get fed up', 'Ho yes, I should like to work longer hours'.40
With their sustained criticism of these programmes as smug, artificial and condescending, these reports should be read as a necessary corrective to anyone who thinks either that the factory front was typically characterised by enthusiasm and high morale or that the BBC was invariably a subtle and effective propagandist for the war effort.
Third, listening tastes more broadly conceived. LR special reports cover topics ranging from listener attitudes to repeats or preferred programme lengths to the most appropriate presentation styles and ? hugely contentious at the time ? the BBC's introduction of the first newsreader with a distinctly regional accent (northerner Wilfred Pickles) in 1941. These investigations reflect both general but also specifically war-related attitudes, for instance reporting listener attitudes to censorship in broadcasting, or investigating the impact of air raids on listening.
Finally, the BBC's wartime listener research provided programme makers and policy makers ? and provides historians today ? with some startling evidence of public opinion during the war. Thus for instance, a February 1941 special investigation on listeners' attitudes to the progress of the war reported that although it was 'the thing' to say 'We can take it', this superficial cheerfulness masked deeper dissatisfactions, with listeners deploring any attempt by official agencies to downplay the seriousness of air raids.41 In mid-1943 an investigation into listener opinions about the British Empire found a general ambivalence about the origins of the Empire allied to support for a more equal partnership in the future; however, listener interest in the Empire paled beside the extraordinary interest shown towards the wartime Soviet Union and the gradual worsening of attitudes towards the USA after its entry into the war.42 Meanwhile, the extraordinary listener response to the eight-part talks series Jobs for All in December 1944, explored in minute detail in a particularly extensive LR special report, demonstrates an interest in and commitment to the principles of post-war reconstruction that directly counters some recent claims about the outlook of the British people in the run-up to the 1945 General Election.43
LR and Government
Despite (or perhaps because of) its wartime activities, BBC LR attracted a degree of hostility from government agencies. Frank Pick, Director-General of the Ministry of Information from August to December 1940, for instance, refused to acknowledge that LR had any wartime value to the BBC and three times attempted to close the department down ? though the fact that he wished to transfer its staff to the MOI's own opinion gathering operation, the Home Intelligence Division, suggests an ulterior motive.44 The MOI's own foray into social research, the Wartime Social Survey, established in July 1940 with the remit to monitor wartime morale and assess the efficacy of government policy, had attracted such intensely hostile newspaper coverage in its first months that its opinion polling activities were discontinued (though its qualitative social surveys survived).45 Other officials, comparing its methods and results favourably with those of MOI Home Intelligence, saw LR as a particularly useful tool in the planning of propaganda, a potentially worrying development given the BBC's concern to retain at least the principle of independence in wartime. In early 1941 LR did carry out, at the MOI's instigation, an enquiry into listeners' opinions of the war (see above), primarily to test how far it could stray from broadcasting matters in its surveys; though considered a success in itself (and the model for several other similar investigations, for instance three 1942 reports on the state of public opinion regarding, respectively, the USA, the USSR and post-war Reconstruction), the BBC Home Board was concerned enough to note that such experiments should not be repeated without special sanction.46 It was perhaps fortunate that the BBC found an ally within the MOI in Patrick Ryan, Tallents' former deputy. Appointed the Ministry's Home Controller of the BBC in April 1940, he made a point of defending the BBC's autonomy and, as BBC Controller of News in the latter part of the war, was largely behind the BBC developing a genuinely popular as well as respected news organisation.
And perhaps surprisingly, given its manifest wartime achievements, LR's position within the BBC still often seemed less than secure. By the end of 1940, LR evidence had become a regular item on the BBC Home Board's agenda. Yet it was constantly stressed that listening figures would not be used to support any 'undue popularisation' of programmes. Silvey himself often regretted that departmental heads were not more proactive in initiating LR reports themselves. In January 1943, under orders to cut costs by 10 per cent, the BBC's Programme Division went so far as to discuss abandoning the whole department for the remainder of the war. At this point, however, departmental heads strenuously opposed the suggestion, arguing (in a resonant metaphor for the times) that the role of LR within the BBC was that of an aircraft spotting for the guns.47 But as late as July 1944, the BBC Director General, Sir William Haley, felt it necessary to warn that LR should be 'confined strictly to programme matters and aspects of listening concerned with actual output'.48
The LRD also proved useful to the BBC in another way. It provided the ammunition with which the BBC could counter the unreasonable demands of government, thereby preventing ministerial attempts to influence, for example, the selection of particular speakers, the choice of presentation style or the scheduling of particular programmes in which they had an interest. Tellingly, when in January 1943 Silvey offered to reduce LR's costs by discontinuing the Forces Survey, senior management rejected the offer on the grounds that this would leave the BBC with no means of countering unwanted interference from the Service ministries.49 (However, Sir Stephen Tallents, now BBC Controller (Overseas), warned that LR should not be used to make 'disagreeable comparisons' regarding the respective popularity of government ministers.)
LR and BBC Policy
Throughout all this, how far did LR maintain its official position that its findings were not to be used to shape future programmes, merely to report on current ones? Certainly there was never any formal attempt to use LR to guide or shape BBC policy during the war, nor any evidence that LR was used to influence the essential character of the BBC schedules or the range and types of programme commissioned; and Silvey kept the LRD firmly separate from the administrative machinery of policy or programming. But equally certainly, those programmes that were scheduled were put together in the light of ongoing listener research about their intended audience. In most cases the influence of LR was indirect. In one famous case, that of the factory music programme, Music While You Work, it was quite specific, the programme being deliberately designed and redesigned around listener research in order to discover exactly the most appropriate dance music to play to war workers on long shifts in factories.50
However, the spotlight on policy is and was in many respects a red herring. In peacetime the BBC saw its role as to provide one of a range of available entertainment and information options. In wartime the BBC had a completely new task: to keep its listeners, and to keep them happy. The wartime BBC therefore, as a matter of policy, paid far more attention than ever before to appealing to, and even maximising, its audiences. In this, LR was essential. However, there is no evidence that BBC policy was 'ratings driven'. Gathering information with a view to accommodating listeners' habits, tastes and preferences was not necessarily the same as pandering to them, and maximising the potential audience for any particular kind of programme ? by scheduling, by adapting style or presentation, etc. ? could still be seen as a Reithian objective. The BBC saw itself as catering not to one monolithic audience, but to a range of (overlapping) audiences. Silvey's multilayered and nuanced approach to listener research (his identification with broadcasting 'publics' rather than a public) made this possible.
After the War
From needing to justify its existence, the post-war Listener Research Department was at last recognised as a key part of the Corporation. Its first peacetime tasks included refining its sampling methods, in particular with the institution of regional samples rather than a single national one to reflect the diversity of BBC audiences and the new regional shape of the post-war BBC Home Service. And of course, following the reintroduction of television in 1946, LR began the challenge of developing an entire new area of audience research. In 1948 listener research into audience tastes formed a key part of the evidence submitted by the BBC to the Beveridge Committee on Broadcasting. On 1 June 1950, in a highly symbolic change prefiguring the ascendance of television over radio, the BBC Listener Research Department became the BBC Audience Research Department.
Robert Silvey himself handed in his notice to the BBC in 1948, intending to return (this time as a director) to the London Press Exchange. Just weeks later an eye infection, serious enough to threaten his sight, put the move in jeopardy. In a mark of the regard in which the BBC held him, he was invited by Haley to withdraw his resignation and return to his old post at a higher salary. In fact his sight recovered, and Silvey continued for a further twenty years to lead and develop audience research for the BBC, in which field he was now recognised as an international authority. Awarded an OBE in 1960 for services to British broadcasting, he retired from the BBC in 1968.
Final Note on the LRD Collection
For the user, therefore, the Listener Research Department 1937-1950 collection offers a comprehensive picture of audience research in the era before television was a mass medium, before commercial broadcasting was instituted in Britain, and before mass media research itself became a subject of widespread academic study and practice. The files included in the collection are working files preserved in the form in which they were originally compiled, representing the day-to-day collection and collation of LR material across the pre-war, wartime and immediate post-war BBC. They are grouped into three principal categories: the Listener Research Bulletins, which provide summaries of listener research findings for internal BBC information from July 1940 to June 1953 (R9/1/1-14) along with Listener Research Bulletin Supplements for December 1945 to October 1951(R9/3/1-4); the Special Reports, which provide qualitative analysis of particular programmes, programme themes, or related issues from 1937 to 1950 (R9/9/1-14); and the Listening Barometers, which provide a daily record of audience figures programme by programme throughout the day from December 1938 to July 1952 (R9/11/1-12 and R9/12/1-7).
As users will see, the collection's cut-off date of 1950 is loosely applied in some instances in order to retain the integrity of these individual working files. The form in which they are presented is in itself a historical marker, and users will note, for instance, changes in terminology in the file headings over time, as well as the occasional duplication of some material across the files collected here. There is also one regrettable absence: the volume of collected Listening Barometers for May to November 1939 inclusive evidently went missing long before the material was catalogued for public use. Users who also visit the BBC Written Archives Centre at Caversham will find much of this material copied and filed across the range of BBC programme and policy files to which it was forwarded at the time for information (they will also find a number of LRD correspondence files which, for copyright reasons are beyond the remit of this collection, but which are well worth exploring further).
When LR was launched, both audience research as a concept and sampling as a research methodology were in their relative infancy. Within the BBC, the latter was mistrusted, and the former was feared. Robert Silvey's triumph was to demonstrate the validity of a research method that uniquely combined the 'statistical' and the 'sociological', and to prove that audience research was not a potential tyrant but, in his own words, 'a good servant' ? indeed, the indispensable servant of public service broadcasting.
Notes & References:
3 For Reith's attitude to listener research, see Briggs, The Golden Age of Wireless (Oxford,  1995), pp.247-50. For the history of public opinion research in Britain, see Laura Dumont Beers, 'Whose opinion? Changing attitudes towards opinion polling in British politics, 1937-1964', Twentieth Century British History, 17 (2) (2006), pp.177-205. The best introduction to Mass-Observation is Tom Jeffery, Mass-Observation ? a Short History (Brighton,  1998); see too Nick Hubble, Mass-Observation and Everyday Life: Culture, History, Theory (Basingstoke, 2006).
4 Silvey was always highly sceptical about the utility of the 'post-bag' as a signifier of wider audience opinion, pointing out that those who felt strongly enough about a programme to write to the BBC about it were prima facie likely to be unrepresentative of the generality of listeners, since they represented extremes of feeling only. Robert J. Silvey, Who's Listening? The Story of BBC Audience Research (London, 1974), pp.28-31.
8 Hilda Matheson, 'Listener research in broadcasting,' Sociological Review, 1935, pp.408-22; Briggs, Golden Age of Wireless, pp.243, 70. Tom Harrisson, co-founder of Mass-Observation, was inspired to study working class life in Britain after returning from an expedition to study the cannibal tribes of the New Hebrides (documented in his bestselling Savage Civilisation, 1936); Mass-Observation's mission statement was likewise to establish 'a scientific study of human social behaviour beginning at home'.
9 The best account of the origins and development of BBC audience research remains Silvey's own memoir, Who's Listening? See also, Briggs, Golden Age of Wireless; Paddy Scannell and David Cardiff, A Social History of British Broadcasting Vol I: 1922-1939 Serving the Nation (Oxford, 1991), pp.375-80; Mark Pegg, Broadcasting and Society 1918-39 (London, 1983), Chapter 5. For a cogent overview of BBC audience research see Erin O'Neill, Series level description for the Audience Research papers BBC WAC R9.
10 Silvey was not exceptional in this: Reith himself had been appointed General Manager of the British Broadcasting Company in 1922 aged just 33 and with no knowledge or experience of broadcasting at all.
12 The only part of the BBC where listener research was not permitted to intrude was Broadcasting for Schools, which was the responsibility of the Schools Broadcasting Council. See Silvey, Who's Listening? p.41.
16 See Briggs, Golden Age of Wireless, pp.244-45. In March 1934 Gielgud had even taken matters into his own hands by making a broadcast appeal for listener opinion on radio plays. He received a staggering 12,000 replies.
18 In the simultaneous telephone call method, homes were randomly telephoned and asked what station the radio was tuned to at that particular moment. Problems with this method included the exclusion of homes without a telephone; the inability to assess audience fluctuations during radio shows; and the hours at which such calls would be tolerated. The recall method depended on face-to-face or telephone interviews, but relied on the accuracy of the respondent's memory of previous days' listening. The Neilson Audiometer, installed in the homes of volunteers, provided on tape a record of the stations to which the radio had been tuned in the previous week; however this method could only record household and not individual listening, and took no account of whether anyone was actually listening to the radio when it was on. See Silvey, Who's Listening? pp.73-75.
24 During the 1930s the BBC had operated in effect two parallel radio stations, the National Programme (available across the country) and the Regional Programme (which varied regionally within a broad scheduling framework), plus a very limited television service in London and the South-East. On the outbreak of war, BBC television closed down altogether and the BBC launched a single national radio station, the Home Service. This was joined in early 1940 by the BBC Forces Programme, which, though primarily aimed at the armed forces based around Britain, became the light entertainment station of choice for the entire civilian population during the war. In 1944, to coincide with the invasion of Europe, the Forces Programme was redesignated the General Forces Programme and much of its light entertainment output was moved to the Home Service. In 1946 the entire BBC broadcasting framework was recast, with the Home Service joined by the Light Programme and the Third Programme, and with the relaunch of BBC television.
25 LR/81, The Home Service, 11 October 1939; LR/83, Proposals for Listener Research in Wartime, 9 November 1939, R9/9/3. The report warned that the new Home Service had 'not captured the public's imagination', but was reassured to find that 'it has not, as some press critics would have us believe, aroused universal fury'.
26 Before the war, the BBC notoriously broadcast no news before 7pm. This practice was the result of an agreement with the newspaper industry dating originally from 1922, and had been suspended only twice previously, during the General Strike in May 1926 and the Munich Crisis in October 1938. However on 25 August 1939, with war looming, the BBC began broadcasting daily morning and lunchtime news bulletins, with afternoon bulletins added shortly after. The BBC defended the decision against considerable press opposition by citing the terms of the agreement, which permitted news broadcasts earlier than 7pm in the case of national emergency. At the end of the war the BBC declared the old agreement in permanent abeyance. See Siân Nicholas, '"All the news that's fit to broadcast": the press versus the BBC, 1922-45', in Peter Catterall, Colin Seymour-Ure and Adrian Smith, Northcliffe's Century: Aspects of the Popular Press 1896-1996 (Basingstoke, 2001), p.121-148.
31 See Siân Nicholas, The Echo of War: Home Front Propaganda and the Wartime BBC 1939-45 (Manchester, 1996), pp.53-54, 57-62. Not all historians have agreed with LR's own interpretation of its findings in this case [see below, page 10].
32 The projected maximum total of 33 million adult listeners was an extrapolation from the estimated radio licence-holding population of Great Britain rather than the population of Great Britain as a whole. Given the extent of licence evasion, it is likely therefore that the Listening Barometer figures in general slightly but not significantly underestimate total listening.
33 Most of the interviewers were housewives. The best interviewers, Silvey told the Parliamentary Labour Party's Committee of Broadcasting, were 'the kind of people one found in WEA classes'. (Silvey, Who's Listening? pp.94-5). A Handbook for Interviewers (1943) can be found in R9/13/3/3.
34 See for instance, Cockburn to Public Relations Executive, 17 April 140, R9/13/3/2. Although the approximate timing of D-Day was widely predicted, the precise date was a closely-guarded military secret. Silvey had rejected the idea of listing the programme as a contingency item every day in early June 1944 on the unarguable grounds that it was likely to foment rumour and anxiety. Silvey, Who's Listening?, p.112. The first confirmed audience figures for War Report appear on the Listening Barometer for Wednesday 14 June 1944, R9/11/10.
35 The only special report conducted into Churchill's broadcasts is LR/713, The Prime Minister's Broadcast, 3 March 1942, R9/9/6, investigating his broadcast of Sunday 15 February 1942 concerning the fall of Singapore. The audience figure of 65.4 per cent was typical; the Satisfaction Rating was unusually low, probably due to the depressing nature of the subject matter. Listening Barometer figures for Churchill's other broadcasts confirm anecdotal evidence that audiences for Churchill's broadcasts were among the highest of the war.
38 For Mass-Observation's methodological limitations see Penny Summerfield, 'Mass-Observation: social research or social movement?' Journal of Contemporary History 20 (1985), pp.439-52. For the Hamburg broadcasts, see Martin Doherty, Nazi Wireless Propaganda: Lord Haw-Haw and British Public Opinion in the Second World War (Edinburgh, 2000), especially Chapter 4.
43 LR/3164, 'Jobs for All', January 1945, R9/9/9 (summarised in LR/3163, 'Jobs for All'). See, for instance, Steven Fielding, 'What did "the people" want? The meaning of the 1945 General Election', Historical Journal, 35 (1992), pp.623-39.
To cite this resource:
Nicholas, Siân (2006) The good servant : the origins and development of BBC Listener Research 1936-1950, https://boa.microform.digital/collections/6/view. Last updated: 27 February 2008.