Although much is known about how Americans budget their financial resources, very little is known about how Americans budget their time resources. How Americans spend their time is one of the most important and least understood characteristics of the population. For example, remarkably little is known about how retired people spend their time, where young children spend their days, or even whether the amount of time Americans are working is increasing or decreasing.
Data on how Americans spend their time, on the activities in which they participate, on how many hours they participate in each activity and on who they were with at the time are not collected on a regular and on-going basis in the United States. Yet such information on time use could be effectively used to better understand the well-being of the population, social and economic behavior, and the implications of public policies. For example, data on time use can be utilized to improve the coverage of national income and product accounts by measuring the time inputs and outputs in nonmarket production. Such measures are important for achieving more complete production accounts and for understanding the effects of public policies on the labor market. Time-use data are also important for making international comparisons. Improved coverage in national income and product accounts that include measures of nonmarket production will enhance the nation’s ability to compare the output and income of the United States with those of other high-income and developing economies. Time-use data can also be used to help understand cultural and social differences across countries.
One measure of a society’s well-being is the amount of leisure time people
have. Many Americans believe they are experiencing a “time crunch,” that they do not have enough time for everything they have to do or want to do. Time-use data could be used to help understand why many Americans feel so short of time by tracking trends in time spent in work and leisure activities.
Data on time allocation can be used to further understanding of individuals’ decisions to work or not work for pay and, more generally, decisions on how to allocate time to different activities. These data are also important for understanding the allocation of time and goods among members of households. In addition, better measures of how workers spend their time while working for pay can help improve productivity statistics.
An important social, demographic, and economic trend is that Americans are living longer and spending more time in retirement. Little is known about how retirees spend their time. Such information could be important for understanding the contributions of retirees to economic output, both paid and unpaid or in volunteer activities, in considering the care needs of the elderly and the care the elderly provide for others (spouses or grandchildren), and in understanding the health care and other service needs of the elderly.
One of the most substantial policy changes in the past decade was the elimination of the main social welfare program for poor families, Aid to Families with Dependent Children, ending the entitlement to cash benefits and replacing it with a policy emphasizing work. A question relevant for understanding the consequences of this policy change is how the time allocation among work and family care activities of poor families has changed.
Unlike many other countries, the United States has no regular national surveys of time use, so questions about American “time budgets” are largely based on incomplete data and speculation. Australia and Canada both have regular and comprehensive surveys for collecting time use data on a national basis. A harmonized European time-use survey that will be carried out in almost 20 countries is also moving forward through Eurostat.
President Clinton’s proposed budget for fiscal 2001 includes funds for the Bureau of Labor Statistics (BLS) to develop a survey to measure how Americans spend their time (U.S. Department of Labor, 2000). BLS has already explored the feasibility of such a survey. In 1997, a pilot study that collected time-use data for a sample of Americans was conducted, and the results of that study were presented at a 1997 conference sponsored by BLS and the MacArthur Network on the Family and the Economy. Using knowledge gained from the pilot study and the conference, BLS published a report on the feasibility of a national time-use survey and developed a proposal to conduct the survey (Horrigan et al., 1999).
With the release of the BLS feasibility report and proposal and with renewed interest in time-use data and research, the Committee on National Statistics of the National Research Council held a workshop to consider data and methodological issues in measuring time use. The workshop brought
together experts in the fields of survey methodology, demography, economics, psychology, sociology, and statistics. Representatives from BLS presented their report on the feasibility of a national time-use survey. Other papers presented at the workshop covered four broad topics: (1) theories of time allocation and public policy considerations, (2) applications of time-use data, (3) accounting for nonmarket household production in national accounts, and (4) approaches to measuring time use. The workshop also included two roundtable discussions—one on conceptual issues in measuring time use and one on future research priorities. See Appendix A for the workshop agenda; summaries of the papers presented at the workshop are in Appendix B.
This document summarizes the workshop, drawing on the presented papers and the formal discussions of them and on general discussions at the workshop. The next chapter discusses why time-use data are needed, highlighting many of the policy and behavioral applications of time-use data introduced above. Next, the report summarizes conceptual issues covered during the workshop. This chapter includes discussion of a framework for how individuals and households allocate their time and a commentary on some conceptual issues in measuring time use—specifically, in defining work and leisure activities, measuring time spent simultaneously in multiple activities, and valuing time. Chapter 4 summarizes time-use studies that have been carried out in the United States, as well as the time-use studies that other countries have conducted. Chapter 5 summarizes the discussion of methods for collecting time-use data, sampling issues in using diaries to collect time-use data, and the importance of ensuring that the data collected match the uses for which the data are intended. Chapter 6 covers features of the proposed BLS time-use survey and a summary of discussants’ comments on the proposal. Chapter 7 summarizes the common themes that emerged at the workshop.
This workshop summary is not intended to provide a complete account of all the behavioral and policy issues that can be better informed with data on time allocation. There are important topical areas for which time-use data could be used that were not covered in the papers and discussion of the workshop, and so are not covered in this summary. For example, there are many potential uses for time-use data in the private sector. Marketers often want to know when and how often Americans use different forms of media so that marketing campaigns can be more effectively directed. Retailers may want to know how the Internet is shaping patterns in time spent shopping. Employers certainly have an interest in better understanding how employees use their time at work and how employees spend time working for pay when they are at home or off-site.
Time-use data can also be used for public policy issues that were not discussed at the workshop. For example, time-use data can improve understanding of the use of and needs for, publicly provided goods, such as parks,
recreational facilities, roads, and mass transit systems. Patterns of use of such public goods can help governments and regional planners design future transportation systems, zoning, and recreational facilities.
Within the topics of methodological and statistical issues for measuring time use and the proposed BLS time-use survey, the workshop was not intended to provide a definitive review of all the methodological and statistical issues in measuring time use, nor to give the BLS specific guidance in planning its time-use survey. Moreover, workshop participants did not agree on all methodological considerations. For example, there was substantial disagreement between participants about recall error and how long survey respondents can accurately remember what they did on a previous day. Yet there were broad areas of agreement with regard to the importance for time-use data collection and with regard to the need to consider certain methodological approaches in future time-use surveys; the final chapter summarizes these common themes.