The last national level survey of time use in the United States was fielded nearly 15 years ago, but the Bureau of Labor Statistics now has a well-developed plan to conduct a national-level survey of time use. During the workshop, representatives from BLS presented a report on the feasibility of the survey (Horrigan et al., 1999). In this section we review the history of the BLS efforts, describe the proposed survey, and summarize the workshop discussion of the proposal.
The Unremunerated Work Act of 1993 included a directive for the BLS to conduct a time-use survey for the purpose of counting unremunerated work performed in the United States and to calculate the monetary value of that work.1 Since the act was introduced, the BLS has developed and tested a pilot time-use survey, cosponsored a conference, established a working group for exploring the feasibility of a time-use survey, and developed the report presented at this workshop.
The pilot study was conducted during 1997 under a contract with Westat. The first phase of the study included 21 cognitive interviews designed to understand respondents’ difficulties in recalling activities from the past day. During the summer of 1997, a test random-digit dial telephone survey with
Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 51
Time-Use Measurement and Research: Report of a Workshop 6 The Proposed BLS Time-Use Survey The last national level survey of time use in the United States was fielded nearly 15 years ago, but the Bureau of Labor Statistics now has a well-developed plan to conduct a national-level survey of time use. During the workshop, representatives from BLS presented a report on the feasibility of the survey (Horrigan et al., 1999). In this section we review the history of the BLS efforts, describe the proposed survey, and summarize the workshop discussion of the proposal. HISTORY The Unremunerated Work Act of 1993 included a directive for the BLS to conduct a time-use survey for the purpose of counting unremunerated work performed in the United States and to calculate the monetary value of that work.1 Since the act was introduced, the BLS has developed and tested a pilot time-use survey, cosponsored a conference, established a working group for exploring the feasibility of a time-use survey, and developed the report presented at this workshop. The pilot study was conducted during 1997 under a contract with Westat. The first phase of the study included 21 cognitive interviews designed to understand respondents’ difficulties in recalling activities from the past day. During the summer of 1997, a test random-digit dial telephone survey with 1 This proposed law was not enacted.
OCR for page 51
Time-Use Measurement and Research: Report of a Workshop 1,000 interviews was fielded. Two types of interviews were conducted, with 500 interviews in each. The first asked respondents to recall what they were doing when, where they were doing it, and who was with them at the time. The second, aimed at measuring simultaneous activities, asked respondents what they were doing and whether they were doing anything else at the same time. Results of the test showed that some activities were underreported when respondents were not given the opportunity to distinguish simultaneous activities in the survey (primary and secondary activities). More time in nonmarket work activities was reported through the second type of interview, where respondents were cued to count time spent in simultaneous activities (Horrigan et al., 1999). In the fall of 1997, the BLS cosponsored a conference with the MacArthur Network on Family and the Economy called “Time Use, Nonmarket Work and Family Well-Being.” The conference brought together experts from a range of social science fields to talk about the economic aspects of time use, time use for children and families, childhood development and time use, public policy and time use, and methodologies for collecting time use data. Following the conference, the commissioner of BLS established a working group to examine the feasibility of collecting time-use data. The working group began with the following assumptions (Horrigan et al., 1999): The purpose of the survey would be to estimate the time individuals spend in various activities. The sample for the survey would be drawn from the outgoing rotation groups of the monthly Current Population Survey. A 24-hour day time diary would be used. The data collection protocol would be a computer assisted telephone interview (CATI). The report and recommendations of the working group were presented at the workshop by the group’s chair, Michael Horrigan. CURRENT PLANS The proposed time-use survey will draw a sample that is designed to be representative of the U.S. population 16 years of age and older. The survey will be designed to produce quarterly estimates of the proportion of time spent in different activities for this population and separately for a set of comparison groups. There are seven proposed sample stratification variables: gender; the presence of children (any under 6, any between 6-17, none under 18); education (less than high school, high school, some college, college graduate with no additional schooling, post-college study); age (16-24, 25-54, 55-64, 65 and older); employment (employed, unemployed, out of the labor
OCR for page 51
Time-Use Measurement and Research: Report of a Workshop force); family type (married couple families, families maintained by single adults, adults not in families); location (urban or rural); and race/ethnicity (Hispanic, non-Hispanic black, other). The survey will also be designed to generate annual estimates of a wide range of activities for an average week, weekday, and weekend day. Appendix C contains the classification codes of activities for which estimates will be produced. The proposed classification system is a modified version of the Australian system and is comparable to international coding systems. A subsample of the Current Population Survey (CPS, a monthly survey conducted by the Census Bureau) would be used for the survey. A subsample of persons aged 16 and over who were in responding households will be drawn from the outgoing panels of the CPS.2 The CPS interviews about 150,000 individuals in approximately 72,000 households each year. Given that one member of each CPS responding household is eligible for the proposed time-use survey and given nonresponse over the course of the CPS, the maximum possible sample size for the time-use survey is 72,000 per year. However, because the CPS oversamples small states and the goal of the time-use survey is not to be state representative but, rather, nationally representative, the maximum available sample size is closer to 54,000 per year. A subsample of these 54,000 will be drawn for the time-use survey. Currently, the proposal calls for at least 20,000 adults to be contacted annually. The proposal also calls for an additional 14,000 to be included in the sample to target smaller demographic populations (based on the stratification groups), for a total of 34,000 in the sample yearly. Assuming a response rate of 70 percent, the projected sample size will be about 24,000. The proposed strategy to gather information on an individual’s time use is to use the designated day approach: each household in the survey will be assigned a day of the week for which the respondent will report his or her activities. An attempt to interview the respondent will be made the day after the designated day. If the respondent cannot be reached the day after his or her designated day, the respondent will be reassigned the same day of the week for the next week. In other words, if a person with a designated day of Monday cannot be reached the Tuesday immediately following the designated day, that person will be reassigned to the next Monday and another attempt will be made to contact him or her following the next Monday. Up to four attempts to contact the individual will be made. Concern about recall error resulted in this approach. It was decided that relying on a recall period 2 The CPS uses a rotating panel design: panels of individuals are interviewed monthly for 4 months, are not interviewed for the next 8 months, and are then interviewed again for the next 4 months. Every month a new panel begins the 16-month rotation. Those finishing their eighth month of interviewing will be eligible for the time-use survey subsample.
OCR for page 51
Time-Use Measurement and Research: Report of a Workshop of more than 24 hours would diminish the quality of the data too much, but that nonresponse would be too high without reassignment. The assignment of individuals to days will be made so that half the days for which data are reported will be weekdays and half the days will be weekend days. Ten percent of the sample members will be assigned to each weekday (Monday through Friday), 25 percent will be assigned to Saturday and 25 percent will be assigned to Sunday. The feasibility report also recommends that data be collected on a monthly basis and reported on a quarterly basis, to cut across all seasons of the year. As currently proposed, data will not be collected for holidays (individuals will not be assigned to a designated day that is a holiday). The survey instrument will be composed mainly of an activity questionnaire (the time-use diary), which will document activities by the sample member over a 24-hour period. The survey will use CATI and respondents will be asked to recall the timing of their activities sequentially. Respondents will also be asked where they were during the activity, whom they were with, and whether they were doing anything else at the same time in order to record simultaneous activities. Respondents will also be asked what activities were done for pay to be able to better identify market and nonmarket activities. See Appendix D for the draft questionnaire presented at the workshop. The current proposal calls for time spent in simultaneous activities to be divided up according to the proportion of time members of the individual’s demographic group spend on the two activities in solo (see Chapter 3). In addition to the time-use component of the survey, other data on respondents will likely be collected. Those data include updated (from the CPS) household composition information, updated total family income, the respondent’s labor force status, the labor force status of his or her spouse or partner, updated earnings information for the respondent, and school enrollment. The projected length of interview is approximately 25 minutes; completing the diaries is estimated to take approximately 22 minutes of the total. The estimate is based on the pilot test results for the time-use component and on experience with the CPS for update information. In addition to the quarterly and annual estimates of time spent in various activities, the data could also be used to produce information on time spent in simultaneous activities and various estimates of time spent in activities around a theme, such as child care activities. These thematic estimates would add together time spent in the activity solely and simultaneously. The proposal calls for a public-use database to be made available for the research community. Because the sample members for the proposed time-use survey are also CPS sample members, the new data could be linked to the various CPS supplements. The BLS working group also considered several topical modules that could be attached to the time-use survey, such as: use of tools, child care,
OCR for page 51
Time-Use Measurement and Research: Report of a Workshop elder or adult care, working hours, division of labor within the household, household production, volunteer activities, subjective assessments of activities, and subjective questions about the experience of time. DISCUSSION Workshop participants were generally very enthusiastic about the proposed time-use survey, but there were some concerns about the topics covered in the survey and the methodologies proposed. One concern, raised by Nancy Folbre, regarded the measurement of time spent in child care activities. She was concerned that the core questions of the proposed survey are not refined enough to distinguish the time parents spend in direct interaction with children in contrast to the time spent in indirect care for children (such as being “on call”). The distinction is important for determining how labor force participation, wages, and public policies affect the allocation of time spent in direct child care activities and how qualitative aspects of time spent with children relate to outcomes. Jeanne Brooks-Gunn also expressed this concern. Other participants were concerned about accurately measuring time spent in the labor force and in educational activities because detailed information about what the respondent does during these times may be difficult to obtain with a time diary. Several participants argued it would also be useful to collect information about the flexibility of work hours from survey respondents. People with flexible schedules may use their time differently, for example, by commuting during non-rush hour times. The rigidity of one’s work schedule may also have implications for one’s availability to provide care for a child, elder relative, or a relative who is ill. No survey can include all the topics that interest researchers, policy makers, and the public because survey resources are limited. In making decisions about topics covered, workshop participants encouraged the BLS to set priorities for data collection. In doing so, the participants stressed the need for the data collection to be guided by how the data will be used. They did note that using outgoing CPS panels as the sample frame means that there are opportunities to link time-use data to previously collected CPS data and that it may be possible to add topical modules later as needed. Workshop participants were also enthusiastic about the survey because it will provide an excellent opportunity to test and develop alternative methods of collecting time-use data. Many participants emphasized the need to know how alternative approaches to collecting time-use data can be used together to gain a comprehensive picture of time use. For example, Thomas Juster suggested that a random hour technique could be used to collect data on time use in a work setting since this method is less burdensome on a respondent’s time than a time diary. Lorna Bailie explained that the Canadian time-use
OCR for page 51
Time-Use Measurement and Research: Report of a Workshop survey used stylized questions about the time use of the respondent’s spouse to substitute for not having the resources to interview more than one member of the household. Norman Bradburn urged BLS to conduct a systematic methodological program in conjunction with the main survey. Such a program could test different methodologies for collecting the data so that future data collection capacity could be built. Several participants urged that experts from a broad range of fields be consulted in the development of the tests. For example, Nancy Folbre suggested using ethnographers to get more detailed information on how people use time, and that studies using the experiential sampling method could further enhance knowledge of subjective measures of time use. Norman Bradburn suggested that cognitive psychologists could be consulted to develop stylized questions and to develop techniques for enhancing respondent recall for diaries. Several workshop participants suggested that a mechanism to give researchers and data users early input and advice for the BLS on questionnaire content and survey design be established. As currently planned, the BLS survey will collect diaries from only one household member. Many workshop participants stressed the importance and usefulness of collecting diaries for multiple household members. Daniel Hamermesh emphasized the need for time-use data for husbands and wives, highlighting many of the conceptual and public policy issues (summarized in Chapter 3) for which such data would be useful. Collecting time-use diaries from multiple household members will lengthen the survey, and sample members may be less willing to respond if the interview is too long. However, workshop participants suggested that a well-developed survey that is interesting to the respondents and properly trained interviewers will help ensure that high-quality data are collected with a longer survey. In terms of who in the household is interviewed, many participants urged the BLS to include those between the ages of 11 and 16 in the sample to provide valuable data on adolescents. Another methodological issue of concern was how days of the week would be sampled. As noted above, current plans call for each weekday to be sampled evenly (10 percent for each day) and for Saturday and Sunday to be sampled at 25 percent each. Daniel Hamermesh and Mihalyi Csikszentmihalyi both reported on research that showed that time-use patterns on Mondays and Fridays are different from time use patterns on Tuesdays, Wednesdays and Thursdays. Consequently, they argued that Mondays and Fridays should not be treated the same as other weekdays: consideration should be given to designating more than 10 percent of the sample to Monday and more than 10 percent to Friday. Oversampling these two days, in comparison with each day in the middle of the week (Tuesday-Thursday), would be beneficial to understanding what happens outside of the workplace because Mondays and Fridays tend to be atypical workdays. Several workshop participants also suggested that holidays should be included in the sample of designated days.
OCR for page 51
Time-Use Measurement and Research: Report of a Workshop Some workshop participants also expressed concern over the plan to reassign designated days to the same day the following week. The concern is that sample members who cannot be reached on the day after the designated day may always be unreachable on that day of the week. Also of concern is that data will not be collected on the days that respondents cannot be reached, which may be different days in terms of work schedules, travel schedules, etc. Yet it may be useful to document time spent in various activities during these unusual days. An alternative methodology suggested was to keep the same designated day, but allow for a longer recall period and try to reach the respondent later in the week. As discussed above, there are measurement error implications for using this methodology. Collecting multiple diaries from the same person at different times and over different times of the year was also stressed as important considerations for the proposed survey. Some workshop participants argued for longitudinal time-use data to better understand changes in behavior in response to either policy changes or changes in family circumstances that cross-sectional data cannot provide. Participants also argued that collecting longitudinal data on sample members could help capture fluctuations in time use over different times of the year, which may be important for studying specific populations, such as school-aged children and their families or employees with seasonal variations in work schedules, or for studying specific activities, such as vacation travel or outdoor recreation. Despite the methodological issues raised in the discussion of the BLS proposal, it was evident from the discussion that participants considered the BLS survey to have a sound beginning and that it should move forward. Overall, workshop participants were enthusiastic about the potential for a large national survey on time use. Though the methodological concerns provide challenges, participants suggested that these concerns are not serious enough to affect development of the survey. Robert Michael emphasized that statistics on time use have been missing from the federal statistical package for far too long and that data collected through the proposed BLS survey will be a significant step in furthering understanding of human behavior and social policy.