Joann Vanek, United Nations: The UN is involved with a number of activities to look at unpaid work and labor force statistics that are presently poorly measured in developing countries. We are also trying to improve the comparability of data in the UN — working on an international classification of time-use activities.

Countries generally collect time-use data to measure unpaid work, validate national accounts, measure labor force participation, and plan types of training activities. The key basis of the UN's work was to not go off and do something new that was not supported by data from countries. We stretched the existing data. The ILO had a problem with it because they wanted to measure job-related activities. There are problems with the same activity being recorded in two different places (e.g., cooking). The experts meeting decided it was premature to develop job-related classifications,. They tried to avoid duplication and came to an agreement of classification that they are now fleshing out. Hopefully within a year they will have tested it in developing countries and have something they can publish. We will be putting papers from the meeting on the UN web site.

She concluded by looking over the past couple of days. It has been a very educational meeting in terms of what we have learned and from a personal perspective. Her perspective is that what has come out of this meeting is richer than what has gone on in the past. The academic community is not so much a part of the user community in other countries as it is here in the US. It is good to see us talking about data collection from both private foundation and government points of view. It has provided a very strong basis for the future, better than ever before.

Stephen Jenkens, University of Essex:

1. He wants to make general observations about getting new time-use data. He sees that as the goal of this agenda. He is an applied economist on the periphery of time use thus far.

2. He wants to say something about the distribution of income within households. You can observe it empirically, but time-use studies can reveal it. He is using a British survey that uses time-use questions similar to the PSID.

3. One main observation: There is no single time-use data set that will meet everyone's research needs, contrary to the tone of the discussion in the last two days. There are hard choices that need to be made. We all are coming from different disciplines with different applications for the data. Thus, there is a great diversity of time-use data demands. Hard choices between these will mean winners and losers, but the choices have to be made. Determinants of time use, heterogeneity and context of time use are important, but they may be best gathered as a piggyback on another survey. Psychologists and macroeconomists may have different demands.

4. In terms of building a successful survey, we need to build constituencies of users, which this conference is a step toward, across disciplines and government agencies.

5. In terms of educating people of the importance of time-use data, we need to spread the word through publications. People have to be prepared so they can understand why existing data is not good enough, since the age of the data is not a sufficient criticism.

6. The British Economic and Social Research Council (ESRC) is calling for applications for research on informal economic activities. He gave the web site address. http://www.irc.essex.ac.uk and/or http://www.escr.ac.uk/ieaspec.htm. The specification covers many of the ideas we have covered in the last few days. Cross-national collaborative research could be fruitful.

7. Finally, thanks for a stimulating meeting.

P. Lindsay Chase-Lansdale, University of Chicago: The old proverb, "If you'd like to get something done, ask a busy woman" is very true here.

She disagrees with Tom Juster's comment in Session 7 about the time-use value for the parent of the child playing next door. The BLS should decide on certain types of time-use that can be captured well. People will tell you about very difficult events in their family lives. They might not seem that difficult to them. We should not be as conservative as Juster wants us to be.

We need to talk more about routine and ritual in family life. It cannot be caught by immediate time-use surveys which we are looking at right now, but maybe it can in supplementary studies. Also, we need to look more at time management, from the gatekeeping parents to the running of the household and the family. These interpretations of the meaning of time should not be lost.

Harriet Harper, The Women's Bureau, US Department of Labor

1. She thanked the BLS and MacArthur for putting on the meeting.

2. The primary mission of the Women's Bureau is to promote the wage earning of women. This meeting has discussed women as unwaged workers. We have difficulty meeting their needs. This meeting has been interesting and challenging.

3. When people talk about constituents, we have to get to the real ones — the women of the US who are at home full-time, part-time, or working full-time, who feel their work has not been valued in the economic scheme of things. Women are determined to get credit for the value their work adds to society. This meeting is an outgrowth to the Fourth World Conference on Women and the Clinton administration's efforts toward women. The main thing to remember is that women want credit for what they do.

Horrigan: Would your office fund a time-use study in the US? (Laughter.)

Group Discussion and Comments

Horrigan: When we all leave, please pick up the cups. Don't forget to give Joanne Spitz your transparencies. We need to leave at 4:45 p.m.

As a result of this meeting, the commissioner will want to know where we go next. The biggest question right now is, even if the BLS did have the money, what sort of surveys should they mount?

1. What would different types of surveys give you in both a managerial sense and a scientific sense?
2. One size does not fit all. His concern is in knowing which size fits which need.
3. Do you want context and occasionally get time use? Where do you attach time-use modules, to which surveys?
4. Right now, they have minimal time-use modules that could easily be expanded.
5. First question he has to answer is, what is the purpose? What is the mission? What are the secondary benefits to the user community (both academic and government)?
6. His short-term next step is to find out about personal-visit methodology and figure out, fundamentally, what the unit of observation is (family, part of a family, individual within a family).
7. He is thinking along the lines of two separate sorts of enterprises rather than a one-size-fits-all approach.

Timothy Smeeding, Syracuse University: Thanks to Michael, Nancy and Joanne for all their hard work.

Jennifer Ward-Batts, University of Washington: Some of the issues that we can address with time-use data deal with how families make decisions about how to allocate their full income across goods, services, and activities [less equal] by income, meaning time and monetary income. There are clearly substitutions and complements between some goods purchased in the market, and some activities "purchased" with time. Tax policy affects time spent producing something in the home vs. in the market. In order to fully and correctly examine issues surrounding the allocation of time or the allocation of money income, we must examine both together. This means we need time-use data collected in conjunction with consumption or expenditure data. She recognizes the constraints on collecting all these data. However, she wants to raise the issue of whether we want to add a time-use survey to an existing expenditure survey or add this expenditure component to a time-use survey.

Klas Rydenstam, Statistics Sweden: We could combine a light time-use survey with a light expenditure survey, especially when respondents are young and see these issues as salient. That might be possible.

Duncan Ironmonger, University of Melbourne: If you read his work carefully, you will see we put together how people spend their time with how they spend their money. We can add data together from separate modules on the same respondent group. We get input-output data on all sorts of households. That is what input-output tables do, given big enough samples. That is the great thing about these tables.

Robert A. Pollak, Washington University: Is how people spend time related to how people spend money? Because theoretically, you could make the case that they would be.

Ironmonger: Depending upon the model you use and the size of your sample, you could get modeling of these joint allocations.

Endre Sik, Budapest University of Economic Sciences: He spent time to learn what the British survey already collected. The Hungary survey contains both time use and expenditure data, and is a longitudinal survey, as is the British survey. The Hungarian version is a public good, so it is free.

Smeeding: We aren't going to get far studying time use without studying context. He gave examples of how findings need context.

Chase-Lansdale: She would like to ask Horrigan about what he sees as feasible in terms of implementing time-use surveys.

Horrigan: From his point of view, one result of this meeting is that the commission will want to know where we should go next. One thought in the short-term, if there was legislation passed to mandate the BLS to move forward, he wants to know what methodology and budgetary costs would be and how valuable the data would be. Even knowing that we cannot give you everything you want, it is still good to know what the needs and interests are. If you want context, is it more important to get that and settle for a module attached to the CPS or the NLS. This would be relatively easy. There are lots of avenues that could take advantage of existing surveys. But in the question of a new time-use survey, we need to know what the mission and the secondary benefits of implementing and paying for this are. What will be the policy benefits?

If you put a gun to his head, he would asked more about methodology. What is the unit of observation — the family, a family member, or the individual? This one choice will affect the contribution we make to the research literature. We should think along the line of these separate enterprises rather than a one size fits all.

Smeeding thanked Michael Horrigan, Nancy Folbre, and Robert Pollak!

Paula England, University of Arizona: and Joanne Spitz!

Horrigan, "I thank you."

Session and conference concluded.

back to top