INTRODUCTIONThe Timeliness of Time Use

—

Jonathan Gershuny

Oriel Sullivan

Some of us may remember the Flintstones and the Jetsons; cartoon nuclear families, respectively stone-age and space-age, from 1960s and 1970s television. The stone-age Flintstones with their all-modern conveniences and fully fitted cave residence lived, from the viewpoint of the less affluent half of the British population, in a still-aspired-to future of American suburban luxury. They, and the Jetson mother and father in their orbiting satellite home, had just the same homemaker-breadwinner arrangements as our parents and grandparents did. The employed husbands left the cave or space station after eating a breakfast cooked by their wives, travelling to work and back in their foot- or jet-propelled private vehicles, and at night watched the same small TV screens placed in the corner of the cave or leisure-deck. Both families had the same reassuringly unchanging patterns of daily activity. The lack of change was the core running joke, of course, at the heart of both series. And, as the evidence presented in this book shows us – at least over the more than half a century of change captured by our time-use data, and despite the enormous progress of technology over that period – there is indeed a certain sameness about the daily round. The pattern through the day, from morning to night, changes somewhat, but in reality quite slowly.

Through much of the 20th century, people had a sense that ‘in the future’ daily life was going to be substantially transformed: less work, more leisure, and more equality in the distribution of these by gender, age and social stratum. The economist John Maynard Keynes, in a lecture delivered in 1924, forecast a future for his audience’s grandchildren with a work-week of no more than a dozen hours or so. In predicting this he was following an idea central to John Stuart Mill’s Principles of Political Economy (1848): that eventually societies would evolve into a ‘steady state’ in which all human needs could be met completely, through the operation of the economy accompanied by an ever-diminishing work-week. Technological change ‘operating like compound interest’ would increase productivity exponentially so as to satisfy all conceivable human wants.¹

Nineteenth- and 20th-century socialists, liberals and conservatives all – if for a variety of different reasons – saw the reduction of working hours as the natural and proper concomitant of economic progress. Sociologists in the 1960s and 1970s interpreted the recent economic and social history of the developed world as progress ‘towards a society of leisure’. Twenty years later, observers, noticing that work time in the USA was if anything increasing, assumed that the apparent end of progress towards the leisure society was a symptom of errors in the management of the US economy.²

But not everyone had taken this optimistic view. Keynes himself (in the second half of the same lecture) identified what others in the 1920s and 1930s characterized as ‘The Leisure Problem’. If work time diminished to this extent, what would a population not educated for leisure actually do with its unoccupied time? Worse, the vacant time left by a substantial reduction of work might actually be damaging. An influential group of European researchers studying the central Austrian village of Marienthal, where almost all the male population became unemployed as a result of the Great Crash in 1928, came to a clear conclusion: work provides a range of experiences (time structure, exercise, a framework for sociability and a sense of wider social purpose) whose loss produces real physiological and psychological damage. In some circumstances, winning freedom from the necessities of work is not necessarily a good social outcome.³

So, do we actually see the emergence of a leisure society in the UK over the past half-century? Or – a theme addressed in several chapters of this book – are we in fact getting busier? The UK population structure has changed quite markedly over the period from the 1960s to the first decade and a half of the 21st century: fewer children are born per family, marriage occurs later, and there have been significant increases in life expectancy. In addition, human needs at the various stages of the life and family cycles may be constant, but the ways they are met are transformed by technological innovations and new sorts of institutional arrangements. And over this period, important new technologies for meeting wants have diffused into UK households. But how has day-by-day life actually changed?

Time-use diary data

Time use is difficult to measure in conventional surveys. Like fish in water, we live in time. While we are aware of the passage of time, we generally do not remember how much of it we devote to each of our activities. For this reason, the sorts of direct questions about time use that are sometimes asked in conventional questionnaire surveys – for example, ‘How many hours did you spend in your job last week?’ – produce rather unreliable estimates. Unreliable in part because of a failure of recall. We do not, as part of our normal routines, undertake the various mental processes that would be necessary to answer this sort of question accurately: firstly, classifying activities into groups – deciding, for example, what belongs in the category of ‘paid work’ – then adding the duration of each successive episode to the accumulated total for that category. And finally, deciding what constitutes ‘the week’, and starting to accumulate a new total of time as a result. Not knowing the answer, we make a guess.

And guessing, we may perhaps be influenced to represent ourselves in particular ways, as, for example, the sort of person who works long hours at her job or cooks, or does not cook, who plays a sport or goes to the theatre, by exaggerating or reducing our estimates of the time we devote to particular activities. These two sources of error, respectively, measurement error and normative bias, are known to distort the responses that people give to these sorts of questions.

The approach taken in time-use diaries is different. We collect a form of information that does not require guesses about totals of time over a day or a week. We simply ask people to record the sequence of what they did over a specific day, together with approximate clock-times. This is a much simpler sort of exercise, which requires respondents to recall a continuous stream of activities, together with the times during which they did those activities. In addition, we ask them to record where they were when they did them, who they were with, and how much they enjoyed themselves. We are used to this sort of record-keeping in our daily lives. We have schedules of activities, often with fixed time points, that require us to keep track of when we are doing things. And we are used to answering questions about our schedules – from friends, parents and partners – about what we’ve done today, with whom, or how much we enjoyed what we did yesterday. It turns out that most people can record reasonably detailed and specific accounts of this sort. And the time-use researchers who collect these sequential narratives can classify the activities that people record into larger groups of activities (e.g. paid work, leisure and unpaid work), and calculate the totals of time spent in each – producing, strangely, information not immediately accessible to their informants!

A brief history of time-use diary collection in the UK

Details of the conduct of everyday life can be found (for example) in the diaries of remarkable mid-17th century individuals such as Samuel Pepys and John Evelyn, whose images of London daily life still capture the imagination. But the daily lives recorded in their diaries are certainly not in any sense representative of the society they describe – if only because the very act of keeping a diary renders the diarists themselves remarkable. In addition, the earliest diaries were simple free-form records of sequences of activities, often without detailed reference to timings. This makes it impossible to accurately calculate the aggregate time spent in different activities. The application of diaries to systematic social research depends on collecting carefully harmonized daily activity reports in large samples selected randomly from a population. Sociologists, economists and demographers can use the results of such time-use studies to describe and explain the factors that influence the chains of behaviours which comprise daily life.

Media-initiated surveys formed the first national-sample time-use surveys in many countries, including the UK, and their collection for these purposes spread quickly across the world. The first surviving large-scale UK sample of time-use diary-based data of this kind was the approximately 2,500 weeks of diary records collected by the BBC Audience Research Department in 1961, used to assist in planning the output of radio and television programmes. From the late 1930s, radio (or ‘wireless’) and early television providers needed to find a way to identify who listened to or watched their programmes. The BBC – and broadcasting companies worldwide – also wanted to know with whom listeners or viewers shared their programmes, as well as what else these people did at other times. This evidence could be used to inform future programming, reach out to new audiences, attract sponsorship for programmes and, in many countries (though not at that point in the UK), sell advertising by apprising potential sponsors about the demographic make-up of audiences at different points in the day or week.

The BBC drew the 1961 seven-day diary sample from electoral address registers in BBC regions. The first survey asked one person aged 15 or older in sampled households to keep a diary for one week in April 1961. The second surviving survey, in 1974–5, also sampled addresses with registered voters in BBC regions, but this time asked all people aged 5 and older who lived in the selected households to complete a diary. This whole-household diary design was quickly recognized as important for a variety of other research purposes. For example, in 1970 the British sociologists Michael Young and Peter Willmott collected a sample of time diaries from marital partners aged 25 to 45 in London as the basis for an early and influential study of the division of domestic labour (The Symmetrical Family⁴). Subsequent research from the 1980s and 1990s⁵ further demonstrated the value of obtaining both partners’ – and indeed whole households’ – accounts of their daily routines on the same days. This enables complex analyses of how the activities of different members of the same household interplay and interact across the diary day, looking, for example, at the interaction of the activities of partners, and also how parents and children interact with each other in the home.⁶

In the early 1980s, the UK Economic and Social Research Council (ESRC) funded, for the first time, a British nationally representative time-use survey designed by academic researchers. It drew a stratified random national sample of addresses, asking all members of participating households aged 14 and older to complete a one-week diary. This survey collected 10,360 days of diaries from 1,601 people during the autumn and winter months of 1983–4. Subsequently, as part of the 1985–8 Social Change and Economic Life Initiative (SCELI) study, the ESRC also funded a time-use diary sample which, like the 1983–4 study, asked all people aged 14 and older in the selected households to complete a one-week diary. Just over 1,700 people from 912 households completed 11,332 day diaries during the spring and summer of 1987. Taken together, these two mid-1980s surveys offer a picture of life across most months of the year, and form the 1980s component of the historical sequence of UK time-use diary surveys we feature in this book.

In the mid-1990s, the UK Office for National Statistics (ONS) took up an interest in time-use diary surveys, expanding the field in the UK from the academic and commercial arenas to that of official statistics. At the same time, the European Statistical Office, Eurostat, began the development of the Harmonized European Time Use Survey (HETUS) programme, designed to provide comparable daily-activity statistics across the countries of the European Union. These comparable, nationally representative time-use surveys are designed to be run every ten years or so. They are intended to contribute to European policymaking in areas as diverse as gender equality, extended national accounts,⁷ measuring hours of paid work,⁸ and capturing the impact of work patterns on family life and overall national health.⁹ The HETUS design incorporated the collection of diaries from all members of the household on the same days. Another important characteristic is that two diaries per person are collected, so that there is information from every respondent about how they spend one of their weekdays, and how they spend one of their weekend days, enabling, for example, the direct comparison of individual weekly and weekend working patterns. In addition, in order to provide a parallel measure to the European Labour Force Survey estimate of weekly hours of paid work, but based on something closer to a time-diary method, Eurostat also proposed a ‘weekly work schedule’ grid for inclusion with the HETUS surveys. This grid includes seven rows representing the days of the week, and 96 columns for quarter-hours through the day. Respondents mark this grid according to the times at which they undertake paid work across the selected week.¹⁰ The ONS 2000–2001 national time-use diary survey formed the UK’s first contribution to the HETUS programme, incorporating its guidelines. Based on a nationally representative stratified sample, 11,854 individuals in 6,414 households contributed 20,991 interviews and/or diaries.

The UK 2014–2015 Time Use Survey

The most recent survey in the UK historical sequence of time-use diary surveys is the UK 2014–2015 Time Use Survey (TUS), a collection of more than 16,000 diary days from 8,000 people in 4,000 randomly selected households. It was designed and organized by the authors of this book (the Centre for Time Use Research) during 2014–15,¹¹ and was funded by the ESRC. Like the previous 2000–2001 survey, the 2014–15 survey formed the UK’s contribution to the HETUS programme and followed its guidelines (see above).

The diary instrument used in our UK study – shown in Figure 1 – has rows representing successive 10-minute periods, and separate columns in which respondents record their answers to: ‘What were you doing?’ (main or primary activity); ‘Were you doing anything else at the same time?’ (secondary activity); ‘Where were you?’ (location); and ‘Where you alone or with somebody you know?’ (co-presence). An innovation of this particular survey was that we added two columns recording whether the respondent was using an electronic device (screen, tablet or smartphone), and how much she/he enjoyed each 10-minute period. The two activity columns are each coded into over 250 distinct activity categories (the ‘own words’ format means that they can be recoded into more detailed classifications if future research demands it). And since respondents often enter more than one distinct activity into these columns (e.g. ‘made breakfast listening to the radio’), we can record anything up to four simultaneous activities when we process the handwritten responses into electronic form for analysis.

Figure 1
The UK 2014–2015 use diary

Each of these simple or complex activity time-slots shown in the diary describes part of a day. The essence of time-use analysis is using one, some, or conceivably all, of the diary fields to produce a classification of each of these intervals of the day which together sum to the 24 hours (1,440 minutes) of the individual’s day. These individual classifications can then be averaged across the sample as a whole, to produce the 1,440 minutes of the population’s ‘Great Day’ (described in Chapter 2).

A large, nationally representative sample of randomly selected diary days collected from whole households provides a uniquely broad and varied analytic potential.¹² The same diary study might be used at one point to produce a classification that relates to issues of paid-work time done during unsocial hours (Chapter 4) or the domestic division of labour (Chapter 5), at another to estimate the amount of time that family members spend together in each other’s company (Chapter 7), at a fourth to measure the extent of ICT usage during the different activities of the day (Chapters 10 and 11), at a fifth to assess the wellbeing or ‘rushedness’ associated with the use of time or to estimate the total enjoyment (or utility) generated across the whole day (Chapters 13 and 14), at a sixth to assess life balance among the elderly (Chapter 12), at a seventh to produce a classification of the activities of the day into different levels of physical exertion (as in Chapter 9), at an eighth to measure the social and spatial context of eating (Chapter 8), and at a ninth to estimate the total value to the economy of unpaid work (Chapter 6). The range of possibilities is limited only by the creativity of the analyst.