CTUR Technical Papers

CAPTURE24: Testing self-report time-use diaries against objective instruments in real time
Jonathan Gershuny, Teresa Harms, Aiden Doherty, Emma Thomas, Karen Milton, Paul Kelly, Charlie Foster
Published: 18 October 2017

This study provides a new test of time-use diary methodology, comparing diaries with a pair of
objective criterion measures: wearable cameras and accelerometers.  

A volunteer sample of respondents (n=148) completed conventional self-report paper time-use
diaries using the standard UK Harmonised European Time Use Study (HETUS) instrument (Eurostat 2009). On the diary day, respondents wore a camera that continuously recorded images of their activities during waking hours (approx. 1500-2000 images/day) and also an accelerometer that recorded their physical activity (PA) continuously throughout the 24-hour period covered by the diary. Of the initial 148 participants recruited, 131 returned usable diary and camera records, of whom 124 also provided a usable whole-day accelerometer record.  

The comparison of the diary data with the camera and accelerometer records strongly supports
using diary methodology at both the aggregate (sample) and individual levels and provides evidence that time-use data may be a preferable alternative to PA Questionnaires (PAQs) for providing population-level estimates of physical activity energy expenditure (PAEE). It implies new
opportunities for calibrating metabolic equivalent of task (MET) attributions to activities, using large scale time-use diary studies deployed for samples representative of national populations.

The Impact of Social Norms on Gender Differences in the Distribution of Total Work-Time in Latin-American Families
Juan Carlos Campaña, José Ignacio Giménez-Nadal, José Alberto Molina
Published: 17 March 2015

We analyse the impact of social norms on differences by gender in the time dedicated to total work (paid and unpaid) by families in Latin America. We use survey data on time use in Mexico (2009), Peru (2010), Ecuador (2012) and Colombia (2012), to estimate differential equations through OLS. Our results reveal differences between countries in terms of the gender distribution of total work (paid work plus unpaid work), with Colombia and Peru being more equitable. These two countries could be approaching a situation of "iso-work", or equality of work. When considering the social norms that explain gender differences in the time spent in total work, we use data from the World Values Survey (WVS) and we obtain that the more egalitarian countries exhibit the highest levels of equality in the distribution of work. It is important to know how men and women distribute their time in total work.

A ‘Queer’ Omission: What Time Use Surveys Might Gain From Asking About Sexuality
Kimberly Fisher, Yiu-Tung Suen
Published: 4 December 2014

By collecting accounts of daily behaviour, time diaries can make conditions of people whose interests have been ignored visible. Not surprisingly, time use surveys often have had an association with social justice, for poor working class families in industrialised countries, migrant communities, women, carers, people with disabilities, and those of advanced age. More recently, regional UN Economic Commissions have shown renewed interest in collecting time use surveys to track progress toward gender equality. Lesbian and gay communities, however, remain invisible, not just in time use research, but across the full range of official statistics. Very few national sample surveys ask any question about sexuality, and consequently the degree to which policies serve the needs of LGBTI people remain unknown. Time diary surveys offer a particularly fitting place for collecting information on LGBTI communities. Much of the lingering prejudice against LGBTI people manifests in daily routines, influencing where people go and when, and with whom they undertake which activities. Measuring the degree to which communities share use of civic spaces gives insight to the levels of social integration. We discuss methodological practicalities for including sexuality in time use surveys.

An Application of ‘The Timing Map’ for Working Time Research: Using MTUS Episode Data in the Netherlands and the USA
Takeshi Mizunoya
Published: 30 August 2014

This paper is to examine the relationship between the length and the timing of working time by applying ‘the timing map’ in working time research. There is a basic analysis for the timing, which is participation rates by time of day and their graph (referred to as ‘the timing graph’ in this paper). Although the graph is still useful to look at when a person does a particular activity, the relationship between the length and the timing of the activity time still cannot quite be seen. As examples, additional research questions could be: * When they work longer hours, to what extent do people work later in the evening or earlier in the morning? * At what time do they take work breaks, meals and sleep? In order to answer these questions, ‘the timing map’ is proposed in this paper, using episode data from the Multinational Time Use Study (MTUS) provided by the Centre for Time Use Research. By comparing the timing of full-time workers between the U.S.A. and the Netherlands, we find that American workers who work excessively long hours on weekdays may have to not only work later but also get up earlier and work earlier in the morning, shorten the work break around noon, and eat later evening meals. This paper also shows that ‘the timing map’ has the advantage of visually representing a way one can see directly how people work and how they spend time on other activities, without losing personal episode information.

Post-Industrious Society: Why Work Time Will Not Disappear for Our Grandchildren
Jonathan Gershuny, Kimberly Fisher
Published: 1 April 2014

We provide a comprehensive focussed discussion of the long-term evolution of time budgets in a range of European, North-American and Pacific democracies, summarising arguments about the changing balances between work and leisure as well as paid and unpaid work. We contrast economists’ assumptions about the purely instrumental nature of work, with sociological and social-psychological arguments as to why we might want or need work in and for itself. We use evidence from 16 countries drawn from the day-diaries included in the Multinational Time Use Study to describe trends in paid and unpaid work over five decades. We demonstrate: (1) the approximate historical constancy and cross-national similarity in the total of paid plus unpaid work time; (2) a gender convergence in work patterns and the emergence of the phenomenon of iso-work; and (3) a reversal in the human-capital-related work-leisure gradient, which we associate with a relative decline in “industriousness” in the paid work of early 21st century societies.

Health Status and the Allcation of Time: Cross-Country Evidence in Europe
José Ignacio Giménez-Nadal, José Alberto Molina
Published: 9 November 2013

This paper analyzes the relationship between health status and time allocation decisions in 6 European countries. Using the Multinational Time Use Study, we find that a better perception of own health is associated with less time devoted to sleep, personal care, and non-market work, and with less time in leisure for men, while it is associated with more time in market work. We also find that the relationship across the activities is very similar across countries, and that market work has a relationship of substitution with sleep, personal care, non-market work and leisure, with mixed evidence for the rest of relationships. This analysis of the relationship of health status and time allocation decisions represents a first step to understand cross-country differences in the relationship between health status and time devoted to activities different from market work, which has been shown to be important for well-being

full text (795.25 KB)
An Implementation of an Experimental Time Use Methodology: an ‘Australianised Counterfactual’ Estimate of National Subjective Wellbeing in the USA
Roger Patulny, Kimberly Fisher
Published: 21 December 2012

Kahneman and Krueger’s landmark Princeton Affect and Time Survey (PATS, 2006), which popularised collection of emotional responses to daily activities previously pioneered by Juster and Stafford (1985) and Robinson and Godbey (1997), reveals the emotional context of American time schedules. While only a handful of countries have collected national sample time use and affect data, those surveys available offer opportunities to measure the degree to which people in these countries find various activities 'pleasant' or 'unpleasant'. We can use time and affect data to construct national well-being accounts linked to lived daily experiences. Such accounts in turn create the opportunity to test how populations might feel should they adopt alternative lifestyles encouraged by national government policies. This paper outlines the technical procedures undertaken in producing a ‘counterfactual’ estimate of time and affect in America. We do this by comparing time use patterns in Australia and the USA and simulating what would happen to wellbeing if the latter country adopted the time patterns of the former. Even though we expect differences in unpleasant time to emerge between the two countries, a hypothetical examination of what would happen if Americans were to shift to Australian-style time schedules is revealing for policy in both countries.

An Overview of Time in Volunteering and Adult Care in the United Kingdom
Kimberly Fisher
Published: 9 June 2010

• Volunteering covers three main domains of activity – donated work for an organisation; informal work for other individuals or the community in general; and adult care.

• British men and women volunteered in roughly equal proportions in the 1970s, but women volunteer at a higher rate than men (though men tend to commit equal and possibly slightly higher time to volunteering than women when they do decide to volunteer).

• The British are less likely to volunteer at younger ages, and more likely to volunteer and to spend more time volunteering as they approach and in the years immediately after retirement.

• Voluntary activity in the United Kingdom falls in the middle range of volunteering in industrialised countries. People in many countries do less, but in some countries, including the USA and Turkey, people undertake considerably more voluntary activity.

• Voluntary behaviour in the UK has become more fragmented over the day with time.

• Over 40% of voluntary activity takes place at the same time as other activities. Diaries best collect these simultaneous activities that include volunteering.

• The Harmonised European Time Use Surveys design of diary, which enables people to complete activities in their own words, better collects data on the range of voluntary activities, particularly informal volunteering.

• Surveys need to collect large samples to collect diaries on sufficient numbers of days when people volunteer to assess the total volume and patterns of voluntary activity.

• Diaries show how people schedule voluntary activity into their other daily routines. Estimating the capacity of a society to take up more voluntary activity requires knowledge of people’s capacity to multi-task and the shift schedules. Diaries reveal what people do on days when they volunteer, but do not indicate the total number of people who might volunteer over the longer term. Additional survey questions or a one month volunteering schedule attached to a time diary survey could reveal this additional detail.

• While a wealth of time-use surveys permit tracking of behaviour trends in the UK from 1961 through 2005, the best historical survey informing voluntary activity was collected in 2000-01.

• Time diaries offer particular value for money. Diaries have higher administration costs than questionnaire surveys, but daily activity schedules inform a wide range of policy areas, including transport, physical activity, energy and resource use, total economic activity (paid and unpaid), work-life balance, parenting, eating and drinking behaviours, and quality of life. One time-use survey can address more areas than comparable funding on a series of questionnaire surveys.

Daily Routines in 22 Countries: Diary Evidence of Average Daily Time Spent in Thirty Activities
Kimberly Fisher, John P Robinson
Published: 5 February 2010

This technical paper provides the reader with a snapshot of daily behaviour in 22 countries for people of prime working age, between the ages of 18 and 641. The versions of this paper published in Social Indicators Research offer three tables showing average time per week in thirty activities for whole populations, then separately for men and for women. In contrast, this document offers six tables displaying time in these activities on an average day. The first three tables display the average time the national populations, then men and women separately, spend in these same 30 activities per day, showing average day as well as average weekday and weekend day figures. No person undertakes every activity they routinely perform in a month or year every single day. For this reason, Tables 1-3 show how long on any given day people across a society are doing the activity, but not how long someone does the activity when they chose to do it. The fourth through sixth tables cover this additional element of the daily activity picture, and show the mean time that people who performed each activity on their diary day spent in its undertaking. That is, these tables show how long the national population, and men and women separately, tend to do an activity if they decide to do that activity on any given day.

Gender Symmetry, Gender Convergence and Historical Work-time Invariance in 24 countries.
Jonathan Gershuny, Test person


We do everything in space and time, but, unaided, we have only imprecise knowledge of the distances we move and of the durations of our activities.  We are much more likely to know how much money we spend on various goods and services, than how much time we allocate to the different things we do.

National statistical agencies conduct questionnaire-based surveys to estimate population distributions of wages, and diary-type studies of households’ money budgets. Until relatively recently, policy makers remained incurious about the population’s allocation of time. But now various public issues (listed below) have led to an increase in the academic and policy salience of population time use data.  Since the 1960s growing numbers of countries have begun to collect occasional diary-based “time budget” surveys.  Bringing many of these together in a harmonised form, we are now able to consider, exhaustively, how (or if) populations’ time allocation changes, across much of the developed world, over an extended historical period, using evidence not available from any other sources.

This newly accumulating evidence reveals some unexpected, indeed puzzling, trends, particularly in relation to work.  What follows combines discussion of a range of theorising about changing time-use patterns, with findings from a collection of comparative statistics of historical change.  It deploys evidence from a sample of 81 time-use surveys, from 24 countries, covering the period 1961-2015, constructed from the 1.3 million randomly sampled whole-day diaries in the Multinational Time Use Study (Fisher et al 2014). It is mainly descriptive, tests no hypotheses, and rather than final conclusions, it provides theoretically-grounded speculations about the new historical puzzles that now emerge.


Looking into the Black Box of Employment: The 'Intensive Hour' Approach to Time Spent in Employment-Related Activities
Michael Bittman

This working paper describes a new software application for smartphones, designed to gather time-use data about the working day. Called random time sampling (RTS), the software was developed to overcome a void in the standard time-use survey data: namely, the lack of detail about activities undertaken and their social context during paid employment. The RTS system addresses the twin problems of (1) respondent burden, and (2) respondents providing potentially damaging information. Compared to a conventional time-diary which asks for an exhaustive recall of activities over 24 hours, the RTS only samples one hour of employment time per notification, with typically only a few notifications per working day, over a maximum of a few weeks. After becoming familiar with what is required, most respondents spend less than 90 seconds on each notification. Respondents are protected against ‘self-incrimination’ because the sampling aims to represent patterns typical of an occupation not of an individual. RTS collects insufficient information from any one individual to provide a useable measure of individual performance. The RTS system can be customised. It can be used the study the length of the (paid) workweek, the allocation of time to (99-999) subtasks, the timing of the tasks (by season, by day of the week and time of day), the social context and location of these activities, and the self-rated experience of doing these employment-related activities. Data from pilot studies, undertaken so far, illustrates how this done.