We thank the Economic and Social Research Council (ESRC) which funded this research with grant RES-060-25-0037 which spanned September 2008 to March 2014.
DATA
Multinational Time Use Study - (MTUS) was (and remains) a central activity of the time use group and is our longest running project. expanded the collection of national studies, harmonized for the purposes of comparative research, to 65 studies from 20 countries, as well as developing both a harmonised episode file, and a new simple file.
American Heritage Time Use Study - (AHTUS) contains harmonised time use datasets from the United States from 1965-66 through 2013, and facilitates the study of national accounts as well as monitoring of changes in time use in this country. CTUR improved documentation, and added both the 1998-01 historical survey and new years of American Time Use Survey data released during this project.
Diaries from Young People - These files follow the MTUS format and cover the diaries of people aged 17 and younger. The project harmonised more children's diaries and improved documentation.
Supplementary MTUS files for the United Kingdom - These files offer additional variables for use with the MTUS files from the United Kingdom, and were expanded as older UK surveys were added and upgraded during this project.
This project also involved building the case for a new UK Time Use Survey in 2013-14 or 2014-15.
METHODOLOGY
- Innovations in Time Use Analysis
- Sequential activity analysis
- Sequence modelling
- Multiple activities from "heavy" diaries
The CTUR team and its associates innovated in the development of the analysis of time use data, including sequences and synchronization, modelling of sequences, multiple activities.
The CTUR team and its associates innovate in the development of the analysis of time use data, including sequences and synchronization, modelling of sequences, multiple activities. This work is a part of the ESRC-funded Developing the Centre for Time Use Research (grant RES-060-25-0037).
Activity sequences constitute among other things the chains of production through which skills are combined to produce useful goods and services. The objective of this project is to analyse historical changes in the times that different sorts of people engage in various activities, and how these are connected to changes in public regulations and services (e.g. maximum work hours, level of childcare provision) and to “Zeitgebers” more generally (societally determined fixed points in the day, such as pub closing hours or timings of popular television programmes around which activities are structured). An innovative means of visualizing these changes as 'day-profile' charts, showing the time devoted to different activities at different times of the day, was developed by the ISER/CTUR team. Using such charts, Fisher et al show from AHTUS data that while the daily sequence of activities for US men in 1965 and 2003 look generally similar (though with a remarkable loss of defined meal times), the charts of women for these years look quite different, with women dramatically reducing their unpaid work time in the middle of the day by 2003 (Fisher et al 2007).
Bittman has used the MTUS Australian time use data series to assess the effect of the increase in unsocial working hours, in particular the effect of the implementation of obligatory Sunday working (for those occupational groups affected), on 'family togetherness' over a week of activities. Those who worked on Sundays enjoyed substantially less time for recreational activities, family conviviality and civic association on a Sunday, and that this time was not made up during the week, so that Sunday workers missed out overall on social participation and had less time to balance the demands of work and family (Bittman 2005).
The representation of the day in diaries is similar to that of events in life-history studies. Seven-day diaries in particular, yielding on average around 150 sequential events of varying durations for each respondent, lend themselves particularly to the formal modelling of daily activity sequences. Lesnard introduced new techniques for modelling 'collective rhythms'—the synchronization of joint activity sequences—based on Optimal Matching Analysis. The methodology, presented in the Bulletin de Methodologie Sociologique(Lesnard and de Saint Pol 2006), is applied in substantive research to modelling changes over time in the joint work-schedules of dual-earner husbands and wives in France between 1985 and 1998. The modelling process identifies twelve underlying work arrangements for such couples, differentiated the degree of synchronicity of work schedules (Lesnard 2004). Within these, a significant proportion overall (20%) of the spouses who both worked a full time schedule experienced a high degree (> 50%) of desynchronization of schedules (the relation of this to the 2007 Voorpostel et al result above is currently under investigation). In addition, desynchronization dramatically increased between 1985 and 1998, both because more spouses worked more desynchronized days over the week but also because there was a general increase in desynchronization on most of the days of the week. The development of schedules other than the traditional 9-5 day (long working days, shifted schedules and fragmented hours) was particularly marked among disadvantaged social groups (Lesnard 2006).
The focus of this research is the redirection of analysts’ attention beyond an exclusive focus on 'primary activity' diary information. The objective was to make use of the complex range of information contained in the 'heavy' diary designs which register multiple simultaneous activities, co-presence during activities, and often also collect simultaneous diaries from multiple household members—materials which have seldom previously been analysed together. The full potential of these opportunities will in the future be easier to realize with the World 6.0 upgrade of the MTUS, which improves the recording of both primary and secondary activities (where collected) and adds Harmonized European Time Use Study (HETUS) variables including time spent in the presence of partner. This will allow for the first time proper cross-national comparisons taking account of the full analytical potential of the heavy diary design.
This research has so far focussed on cross-time comparable materials from the USA and the UK. Voorpostel, Gershuny and van de Lippe (2007) used AHTUS data to contribute to the debate about changing intimacies in personal relationships. Analysing changes in the joint activities of couples over 4 decades, they found that by 2003 partners indeed spent a larger part of their leisure time together (controlling for different compositions of the groups over the years), congruent with the idea that marriage has become more intimate. Gimenez and Sevilla-Sanz extended this idea by investigating how leisure inequalities have changed over time according to educational group, also from AHTUS data. The main conclusion was that highly-educated individuals now have less though higher quality leisure, as measured by spending more leisure time with their spouse and other adults, and enjoying a higher percentage of uncontaminated leisure, i.e., leisure that is not done simultaneously with non-leisure activities (Gimenez and Sevilla-Sanz 2008).
Central to progress in this area are methods for conceptualising multiple activity in real time, while maintaining the 24 hour/day constraint. A presentational innovation developed for the CTUR work in this area, is the base-proportional histogram in which the horizontal axis represents the 24 hours of primary activities, the vertical representing the proportions spent solely in that activity and with other secondary activities and combinations of activities. We see from this 'propogram' of the time of UK mothers of young children in 2001, that around two thirds of their leisure time is spent co-present with (and hence perhaps responsible for) children, and about 10% involves caring for children as a simultaneous activity. CTUR is contributing to various ongoing international projects developing this important area of analytic practice. In the childcare area, Gershuny contributed a short chapter to the Bianchi, Robinson and Milkie (2006) book. Similar issues arise with the use of diaries which include time specific measures of affect: Gershuny prepared a paper on activity and enjoyment measures for a meeting convened by Diener, Kahneman and Krueger (arguably the world leaders in this work) in Princeton in October 2008.
Development of the Time Use Research Field
We compiled an extensive library of time use resources and are developing lists of time use publications. This project both documented methodological developments and innovations in the field over the last 150s years and also improved the CTUR on-line resources.
This project had resource and publication components. The resource component built up two extensive CTUR web site resources, a list of time use publications and a table of details of time use studies (both are presently the most extensive lists in the field as they stand). This project updated the table and add to the publications list, while also developing underlying searchable databases which will ease the use of this information. In addition to expanding and improving these resources, we wrote a series of articles profiling the development of time use survey collection, as well as developments in publications using this data. We addressed questions including:
- Who funds the research and for what ends?
- How has data collection changed and what effect has this had on the data?
- Who publishes using this data, and what topics most interest different gender, regional, discipline groups?
In addition, we produced a collection of key time use publications published over the last 125 years published by Oxford Bibliographies in Sociology.
SUBSTANTIVE TOPICS
Cross-National Policy Differences and the Relationship to Time Use
This project investigated the policy correlates of national differences in historical changes in 'time-budgets'.
There has been considerable interest over recent years in the effects of distinct regime types, which might be expected to have common time use patterns, and therefore to produce consistent national groupings by time use. Preliminary work suggested that, on the contrary, consideration of various different aspects of daily life (eg employment, unpaid work, leisure consumption) in fact produce somewhat different groupings of countries, sometimes organised by geographical regions, sometimes reflecting systematic policy differences. A cross-national examination of differences in the time parents spend caring for children aged under five in Norway and Sweden based on HETUS data showed that even between countries customarily grouped together in regime-based analyses, variation in child care and parental leave policies can produce quite significant differences in the time both fathers and mothers spend in child care at different ages of children (Sullivan et al, under review).
However, contrasting Scandinavian countries with countries generally classified as representing the most different policy regime type—the UK and the USA—reveals how the absence of child care related policies has an effect on female employment for women with children below school age, but also and more strikingly that, in a cross-sectional comparison of these countries, the time parents spend in child care is greater where policies are less 'family friendly'! That social norms and externalities, rather than more obvious indicators of policy, may be of paramount importance is a conclusion supported in a series of papers by Sevilla-Sanz and others, which investigate the seeming paradox of the national-level relationship at the end of the 20th century observed in OECD countries between traditional household ideologies and 'low low' fertility (Delaat and Sevilla-Sanz, under review; Gimenez et al 2007).
These findings have led to the development of a new initiative. Recognising that an adequate analysis of the MTUS requires an understanding of changes in public policy at the national level, we have started the construction of a cross-national policy-change data base (building on the more static materials developed for the Luxembourg Income Study), initially including measures of parental and child care policies intended to be used alongside time use data in multilevel models of child care and related activities.
Gender and Time Allocation
This was a node of the ESRC Gender Network led by Dr Jacquie Scott of Cambridge University. It uses time diary materials alongside panel survey materials to investigate the relationship between division of domestic labour within households and the gender wage gap.
- The gender division of domestic labour
- Women's proportion of unpaid work
The MTUS provides the richest information for the analysis of national differences and historical changes in the domestic division of labour available from any source. The regularity of historical change and cross-national similarities and differences in time use would simply not be known if the MTUS project had not reconstructed and harmonised “heritage” time-diary studies, some almost 50 years old.
The 12 countries, with a total of 43 national surveys, that currently constitute the MTUS’ diachronic (two or more surveys per country) sub-sample, provides a promising basis for pseudo-panel (or birth cohort-following) analysis. But we have opted for the moment to focus first on the analysis of our real time diary panel materials. Using a combination of analysis of the Home on Line (HoL) panel, and the fused HoL/BHPS data (see Social Change and Daily Life methodological project on data calibration), Gershuny and Kan are currently working on a challenge to Brines’ “doing gender” hypothesis. Kan (2008), for example, found that, in the UK, women generally report their housework hours more accurately than men. Therefore, when the means of stylized estimates of men’s and women’s housework hours are compared, the gender gap in housework participation will be underestimated. In addition, men holding traditional gender-role attitudes tend in general to report more housework time in surveys than in diaries, but the tendency is reversed when men do more housework. This finding casts doubts on earlier influential research that suggested economically dependent men tend to undertake less housework in order to avoid further loss in their masculine identity. An analysis of the gender differences traceable to previous time allocation patterns (Kan and Gershuny 2009) is due to be published shortly in a book edited by Ermisch and Brynin of the ISER.
Household Accounts
Money and socio-economic time accounts. This project involves the macro-level extension of household accounts, with the objective of providing a combined monetary and time based system of national accounts.
The theoretical framework is provided in Gershuny (2008; under review). Traditional household extension accounts have relied on input methods: time use diaries are used to identify time devoted by households to the various categories of unpaid work, which are then attributed a value. Output methods, in which household consumption events are counted and valued by market equivalents are technically preferable and enable the non-circular estimation of, for example, domestic productivity (the rate of domestic output per hour of unpaid labour). Measurement of such outputs has been an interest of the Office of National Statistics (ONS), and an alternative to estimates of the money value of labour or consumption focuses on the value of the time spent in that labour or consumption activity.
An empirical example using the UK 2000 HETUS survey draws on previous research carried out by the ONS to place a monetary value on the time spent in both formally organized and informal voluntary work. Modelling adjusting for time spent in other activities, it is shown that while median wage rates for formal voluntary work are greater than those for informal helping, the latter is greater in frequency and duration and therefore more economically valuable from the societal perspective. The finding has implications for gender and citizenship issues since it is the highly educated who spend more time on formally organised voluntary work while the less qualified, particularly women, spend more time on extra-household unpaid helping activities (Egerton and Mullen 2008).
Extending this use of time diary data into a comprehensive comparable system of time-based socio-economic accounts (Gershuny 2008) in one of the aims of this research. Time, unlike money, is finite. Since at the societal level the total of work time (including that embodied in exports) must be sufficient to produce (or import) the goods and services required for the society’s consumption time, and since the distribution of occupations in the economy must mirror the pattern of consumption, time use indicators can provide not only measures of labour supply, but also (through consumption) the demand for labour. Thus the society’s time-budget can be used both to summarise all economic activity, and to connect the various sorts of economic activity to unpaid work and to consumption activities. Gershuny 2008 calculates societal- level time-budgets (the 'Great Day') from the UK time use data series from 1961 to 2001, providing evidence of 40 years of change in paid and unpaid work and its relationship to the fulfilment of consumption 'wants'. It is shown that over the period the modality of the 'Great Day' in the UK shifted decisively away from the fulfillment of basic needs to the satisfaction of luxury wants, associated with a growth in high-skilled employment and an overall substitution of unpaid for paid labour.
An assessment of the AHTUS data for the construction of extended National Accounts for the USA also formed part of the development process of the AHTUS data by CTUR (Egerton et al 2005).