Chapter 1Fifty Years of Change in UK Daily Life at a Glance
—
Jonathan Gershuny
Oriel Sullivan
We start with the most intuitive representations of changes in the sequential pattern of our daily lives. When we write an ordinary diary or daily ‘journal’, we are typically recording a sequence of activities, events or feelings – often all of these together. This daily sequence, the way we progress from activity to activity and place to place, meeting different people at different times – at work, for lunch, in the evening – is also the way in which respondents to time-use diary surveys record their daily activities. However, this sequential progression of our daily lives tends to get lost in the aggregate statistics of time use. Aggregate statistics typically average the amount of time that we spend in different activities, in different locations, and with different people (an average 8 hours per day at work, and 2 hours watching TV with family, for example). They have an important place in research into changes in time use, as we will show in Chapter 2, but they do not have the same intuitive appeal as the visualizations of the sequences of our daily activities that we present here. Strangely, this more intuitive, sequence-based, method for analysing how people record their daily lives in diaries is a relatively recent addition to time-use diary methodology. A rapidly growing body of research now moves beyond the analysis of aggregate time-use statistics to examine how time is structured sequentially.1 These newer sequential approaches study not just what people spend their time on, but also the order and timing of their activities. The order and timing of people’s daily lives matter for important social processes such as coordinating household divisions of labour, productive activity, and the scheduling of social contacts. Putting together these sequential accounts of daily life gives us an insight into how individuals’ activity sequences are aligned throughout the day.
Analysing time-use diaries as sequences has led to new breakthroughs in the measurement and modelling of everyday time. An example, enabling the comparison of daily sequences for different individuals or groups, is Andrew Abbott’s classical optimal matching approach, where we can envisage sequences of activities as strings of events that can be compared in the same way that strings of DNA or protein sequences are compared to each other.2 This allows the computation of similarities and dissimilarities between pairs, or clusters, of activity sequences. Using developments of this method we can compare differences, for instance, in the sequence of activities between those who have standard and non-standard work shifts, or between men’s and women’s work/family schedules. In addition, since we can compare when and to what extent individuals’ activity sequences intersect, we are able to analyse activities that take place at particular times as part of the larger networks of social action that connect individuals through space and time.
We use our historical sequence of UK time-use diary surveys here to compare activity sequences for the whole population from the 1960s through to the second decade of the 21st century, giving us a unique perspective on changes in the sequential patterns of activities of the UK population over a 50-year period. This technique provides the most visually intuitive, detailed and comprehensive picture of broad changes over the past half-century in the national population’s patterns of daily activity sequences.3
Tempograms: the way through the UK day
The sequential activity ‘tempograms’ we illustrate here show the proportions of the population engaged in various activities through the 10-minute episodes of the day from the early 1960s to the second decade of the 21st century.
Figure 1.1 illustrates the temporal sequence of primary activities for the whole UK for the population aged 16–64 (working-age population), from the earliest (1961) to the latest (2015) time-use surveys. This provides us with a picture of the activities that diarists are doing in successive 10-minute time slots throughout the day. The tempogram shows the time of day charted on the horizontal axis, and the percentage of the population engaging in a range of activities on the vertical axis. We have simplified the activities of the day into nine distinct sorts of activity: 1) sleep, personal care and eating at home; 2) paid work and education – the latter for those identified as being full-time in school, further or higher education; 3) unpaid cooking, cleaning or DIY; 4) shopping; 5) childcare; 6) out of home leisure and recreation; 7) exercise; 8) watching TV, video, audio and reading; and 9) other leisure at home.4 The differently shaded areas are proportional to the average time devoted to each of these activity categories through the day.
Figure 1.1 Activity tempograms: women and men aged 16-64, UK (1961-2015)
TV, video, audio and print media
Other leisure at home
Exercise
Leisure out
Childcare
Shopping
Cooking, cleaning, DIY
Paid work, education
Sleep, personal care, eating
By simply eyeballing the areas representing the different activity groups across the five graphs we get a preliminary overall impression of historical change in how time is used across the 24 hours of the day. The graphs as a group show substantial changes over the period from 1961 to 2015. Perhaps the most striking change is the gradual, but historically consistent, fading away of the division of the waking day by the three peaks which marked mealtimes (shown by the lightly shaded areas at the base of the graphs). In 1961 we see two prominent peaks, at 1pm and at 5.45pm. Averaging across the whole population aged 16–64, 30 per cent or more of men and women were eating at these times, and 80–90 per cent were involved in some eating within an hour of each of these peak mealtimes. The peaks determine the overall shape of the 1965 graph. By 2015 it is evident that they have substantially diminished in size, and the graph has a distinctly ‘flatter’ (less ‘peaky’) profile. But, as we will see in subsequent chapters (in particular Chapter 8), this does not mean that eating itself is disappearing. Rather, it points to a combination of the desynchronization of mealtimes, and the spread of between-meals ‘snacking’.
On the other hand, there are also some remarkable continuities evident in the historical succession of tempograms. We are frequently assured that the emergence of the ‘24/7 society’ has led to a substantial reduction in sleep. However, little evidence for this is apparent (lightly shaded areas at the beginning and end of the day). Indeed, Chapter 2, which looks at aggregate statistics for sleep time, shows a small increase of time spent in bed, most clearly through the more recent period, which is precisely when the supposed decrease is supposed to have occurred. The disappearance of work as envisaged by the sociological prophets of ‘the leisure society’ is likewise not evident. While the area of the graphs relating to (traditionally feminine-defined) tasks like cooking and cleaning (see horizontal stripes) has reduced in size somewhat over the period since 1961, the overall area devoted to other categories of unpaid work – shopping, childcare – (horizontal stripes and the area immediately above) has increased. Similarly, a casual impression of the area devoted to leisure (both indoor and outdoor) suggests that there has not been that much change in leisure time over successive decades.
Admittedly, when we dig down further according to gender and across the days of the week, which we do below, we do see some more substantial changes. However, the main point here is that when we use this unique body of sequential evidence to examine many of the more dramatic prognostications of change in the broader patterns of the day – work, leisure, sleep – they simply vanish away.
Delving deeper into some of the differences we find if we look separately at particular subgroups of the population, we distinguish in the next sets of figures between tempograms for women and men, and for weekdays, Saturdays and Sundays. Weekdays usually resemble each other quite strongly in terms of time use due to the structured sequencing imposed by traditional weekday patterns of work, while Saturdays and Sundays are usually rather different, both from each other and from weekdays.5 Here we restrict the figures to show only tempograms from the first (1961), middle (1985) and last (2105) surveys of the UK time-use diary sequence.
The first group of tempograms (Figure 1.2) shows women’s and men’s weekdays. Most strikingly, and emerging throughout this and the next chapter, we see a convergence over time between the weekday tempograms of women and men. In 1961 women’s weekday pattern was really very different to that of men, with much of the main part of their weekdays – up to 5pm or thereabout – being largely devoted to the unpaid work activities of cooking, cleaning (in horizontal stripes), shopping, and childcare, with the corresponding portion for men being almost exclusively devoted to paid work. Focusing in on these changes in paid and unpaid work by gender, in the three graphs for men we see a quite sizeable reduction in those parts of the weekday spent in paid work across the period 1961 to 1985. But this reduction ceases when we compare men’s paid work from 1985 to 2015. If anything, over this second period we have a picture of a slight increase in the part of the day occupied by men’s paid work (with less time devoted to meal breaks). And in parallel, over both the 1961–85 and 1985–2015 periods, and pretty much throughout the weekday from morning to evening, we see small but growing increases in the areas representing men’s cooking and cleaning, shopping and childcare.
The series for women shows some complementary changes. Over the two time periods between 1961–85 and 1985–2015, we see significant reductions in those areas of the weekday tempogram devoted to cooking and cleaning activities (horizontal stripes). The area indicating women’s paid work time appears, as expected, to grow between 1985 and 2015. But, as is the case for men, this area actually somewhat decreased in size over the earlier period between 1961 and 1985. How does this apparent discontinuity square with what is known from other sources to be a continuous growth in the rates of women’s participation in the labour force through the 55 years covered by these graphs? The answer is that it reflects both the effects of the general reduction in hours spent in all full-time paid jobs between 1961 and 1985 (shown in the tempograms for men), but also the predominance of part-time work in the paid employment taken up by women in this earlier period.
Figure 1.2 Weekday activity tempograms: women and men aged 16-64, UK (1961-2015)
TV, video, audio and print media
Other leisure at home
Exercise
Leisure out
Childcare
Shopping
Cooking, cleaning, DIY
Paid work, education
Sleep, personal care, eating
The Saturday tempograms (Figure 1.3) remind us of a mostly forgotten achievement of the mid-20th century: the invention of the two-day weekend. If we follow the historical sequence for men, we see a dramatic reduction in the area of the tempogram devoted to paid work on Saturdays. This was, it seems, a post-Second World War development in the UK. Until the 1930s, in the UK paid work for the employed usually occupied six of the seven days of the week. Our UK time-use data series starts about halfway through the process that transformed Saturday into a day more resembling a day of leisure, shopping and unpaid work.
By 1985 we see, apparently compensating for the end of the Saturday morning ‘at work’, a considerable increase in the amount of unpaid work (cooking and cleaning, and growing most rapidly, shopping) done by men. And between 1985 and 2015, the area representing men’s cooking and cleaning continues to increase, while the equivalent area diminishes somewhat for women: suggesting perhaps an actual gendered transfer of work, with men seemingly taking some unpaid work in a category previously done almost entirely by women. In relation to leisure time on Saturday, in 1961 it appears that men had more out-of-home leisure, particularly during Saturday afternoon, than did women. Correspondingly, women had much more at home leisure during the daytime than men did. But by 2015, the gender specialization in daytime Saturday leisure had disappeared. In fact, the disappearance of men’s paid work on Saturday mornings seems to lead to an even more marked historical gender convergence than we saw in the case of weekdays. The Saturday tempograms in 2015 have a distinctly similar appearance for men and women (although men still do somewhat more paid work and women more unpaid work).
Figure 1.3
Saturday tempograms: women and men aged 16-64, UK (1961-2015)
TV, video, audio and print media
Other leisure at home
Exercise
Leisure out
Childcare
Shopping
Cooking, cleaning, DIY
Paid work, education
Sleep, personal care, eating
Finally, linking the tempograms for Saturday and Sunday, the area of the tempogram representing women’s and (more particularly) men’s Saturday shopping increased remarkably, particularly during the afternoon, over the first part of the 55-year period, but remained hardly changed – or even slightly reduced – from 1985 to 2015. Why did this happen?
Figure 1.4 Sunday tempograms: women and men aged 16-64, UK (1961-2015)
TV, video, audio and print media
Other leisure at home
Exercise
Leisure out
Childcare
Shopping
Cooking, cleaning, DIY
Paid work, education
Sleep, personal care, eating
The answer to this question is found in the Sunday tempograms (Figure 1.4). Weekday and Saturday shopping have shifted to Sunday. Gradual changes in legislation governing the opening of shops on Sunday (‘Sunday trading’) from the 1970s onwards introduced a whole new category of activity to Sundays in the UK. In the 1960s only a very few specialist shops (off-licence shops selling alcohol, some do-it-yourself equipment shops, and shops catering for ethnic or religious minorities with restrictions on shopping on other days of the week) were permitted to open during the day on Sunday. So in the 1961 tempogram we can detect only a tiny sliver of shopping activity through the morning. By the mid-1980s, the number of exceptions to the Sunday trading restrictions had increased, showing up in a little more shopping activity throughout the day. But major changes to the Sunday trading laws came only at around the start of the present century. As a consequence, we see a dramatic growth in Sunday shopping between the 1985 and the 2015 tempograms. Just 4 per cent of our sample went to church on Sunday in 2015, while 25 per cent went shopping. In the UK, clearly, shopping rather than church now emerges as the major act of Sunday worship.
Other quite dramatic changes specific to Sundays emerge from Figure 1.4. In 1961 substantial amounts of exercise, particularly for men, took place on Sunday mornings and afternoons: this largely involved sports participation in team games. By 2015 the dominance of team sports on Sundays was gone, and exercise had become more evenly distributed both across the day, and between men and women. At pretty much the same times on Sunday mornings in 1961, while the men were out playing sports, the women were doing housework, the weekly wash, and making Sunday lunch. In 1961 something approaching half the adult population were having their Sunday lunch at the same time of the day (represented by the peaks of light shading).
With respect to at-home leisure activities, in 1961 on Sundays there were small surges of media participation before and after Sunday lunch – this was one of the high points of weekly radio listening in 1961: ‘family favourite’ musical performances and comedy programmes on the BBC’s Light Programme. And then, substantial time devoted to home leisure on Sunday afternoons – digesting the unusually heavy midday meal, and then starting mid-late afternoon television – which was in 1961 on the cusp of its diffusion across British households: by 1967, 88 per cent of British households had television licences, while only 48 per cent had them in 1957.
In summary, over the three successive day graphs spanning the 55 years of Sundays, we can identify some really striking historical changes. The prominence of mealtimes, which in 1961 were even more strongly delineated on Sundays than on weekdays or Saturdays, is substantially eroded. Media activities and exercise have spread more evenly across the day, though with the peak of concentration – now overwhelmingly television viewing – in the evening. Whole new categories of activity have come to prominence, in addition to the previously mentioned Sunday shopping. For example, childcare, which hardly signified in 1961, but by 2015 was reported as a significant Sunday activity by both men and women. And, most remarkably, the gender convergence in activities. The essential characteristic of traditional gender relations, the dependence of men on women’s immediate supply of timely domestic services through the day – caring for the youngest children, cleaning and preparing the lunches while the men are out at work or playing – can be seen from the 1961 Sunday tempograms. The equivalent pair of 2015 tempograms still show some gendered disparity in work time, with men doing somewhat more paid work than women, and women a bit more unpaid work. But, 55 years on, it seems that all our days are now much less strongly differentiated by gender.