|contact info | more about the|
Session VII: Time Use and Public Policy
Caren Grown, MacArthur Foundation
Robert Michael, University of Chicago
Tim Smeeding: Time and Public Policy: Why do we care and What instruments are needed?
I. He will hit points that were not hard-hit today or yesterday. He began with two basic reasons we should collect data:
B. To determine the determinants of time use (need modules). He thinks we learn more from looking at distributions than averages (a stratification focus).
C. The theme is heterogeneity.
II. Economic measurement
A. He wants to look at macro issues such as growth, CPI, productivity.
B. He wants to look at micro issues such as the distribution of real income (RI), where RI = MI (money income) - CW (costs of earnings) + NMW (non-market work)
C. Example: the hours-worked controversy:
1. Define "work." There is a difference between asking how many hours you worked vs. a time diary.
2. There are at least two methodologies same population survey and reconciliation of points mentioned earlier.
3. Variance issues: DINKS (Dual Income-No Kids families) vs. stay-at-home mothers; single parents; early retirees.
D. Data needs: MI, CW and NMW
III. Broader issues in social policy arena (He will talk about care-giving, human capital and volunteerism.)
1. There is a disagreement in research over what the real estimate is of middle-age families providing elder care, 25%, 12% or none.
2. Single parents and child inputs, what gives?
3. Need estimate of child care, care-giving extent, burden, social security reform, etc.
B. Human capital: need estimates of time invested and real full returns
C. Volunteerism: Who is providing services for whom and to what effect, where, and why?
1. He posits that there is a big difference in volunteer work between helping a needy child to eat vs. planning a trip for a senior citizens group.
2. If neighborhoods sensibly build and use social and community capital, they can prevent crime and negative neighborhood outcome. We need to understand volunteerism.
3. Volunteerism and Youth.
4. Policy issue need extend and determinants.
D. Intra-household allocations and decisions: Who does what and who gets what? Research has looked at and needs to revisit:
1. Bargaining models
2. Income distributions within households
3. Violence, care-giving, stress, child-rearing, location, child support/contact
4. Data needs: unit based surveys.
IV. Infrastructure Time
A. He thinks this is most important gap in research and data.
B. Now is a good time for fitting it into the national budget since monies are becoming available.
C. We need to know what sells vs. what is needed.
1. Need more than one instrument: The National Science Foundation (NSF) and the National Academy of Sciences are both looking and could collaborate with the BLS and other funding/research groups. The BLS/BEA needs CPS, CEX, other.
2. Social Policy needs modeling, longitudinal, saturation by area, topic, and group (e.g., race or gender subgroups).
D. How can we do this?
1. Develop user-driven surveys by bringing together government, academics in many disciplines and other stakeholders for a long-term planning process.
2. Understand what is desirable (i.e., wish list) vs. what is feasible. The result is an available survey with a whole lot of user interest; even economists would use it!
E. International data process: Whatever is collected in the US needs to be available internationally. A data highway.
F. It is time for a social science infrastructure relief act.
Althea Huston: Implications for Child Policy:
Longitudinal studies of child time with TV:
The Displacement Hypothesis: Something must decrease when TV watching increases. They found it depends on the type of programming being watched, whether intended for children or for a general audience. Play and watching TV often occur simultaneously and have a positive association. However, there is a trade-off between reading and watching general-audience TV programs.
Eighty minutes per day of secondary viewing of adult programs is done by 2-year-olds. This declines as age increases, since time with parents declines with age. This serves to remind us that time spent with parents is not always necessarily good time.
There are gender differences in time use. For example, at age 4, girls spend more time watching educational TV, reading, doing personal care, and chores. Boys spend more time playing video games (a computer-related activity that may give them an edge in technology).
When considering maternal employment, more than just whether or not the mother is present affects TV watching. Children in child-care centers watch less TV than those in home-based care.
By age 5, school-readiness was higher for children who watched more educational TV, but other negative outcomes increased with more general-audience TV viewing. It was found in one study that early exposure to educational TV predicts high school grades and some other outcomes, controlling for parents education and the birth order of the child.
In order to talk about child outcomes, we need childrens, not just adults, time-use data. Much of childrens time is not spent with their parents. We need more information on parents work schedules. In addition to the number of hours they work, we should know if they work at non-standard times. We should use non-linear models because of diminishing marginal returns. We found children in self-care after school were okay except if they lived in a dangerous area. Therefore, we must consider the context.
Cathleen Zick: Family Policy, Parent-Child Time, and the Well-Being of Children. (She will be somewhat repetitive.)
Why focus on this topic?
What do we know about the correlates of parent-child time?
What do we know about parent-child time other than primary child care?
What do we know about parent-child time and child outcomes?
What is missing from the picture?
Where should we go in terms of new data collection?
Where should we go in terms of assessing the impact of family policies?
Examples of public policies that may directly affect parental time use and child outcomes:
Examples of public policies that may indirectly affect parental time use and child outcomes:
Robert Michael: This conference has been a good first step, with international and interdisciplinary perspectives.
His own reflections: Lets drop work leisure. The dichotomy it suggests is more misleading than helpful. It is unarguable that non-market time is productive. Considering the households command over resources, it is inappropriate to ignore how time affects households in poverty.
Non-market work has been out of sight and out of mind. Money is easier to measure and observe. But by not looking at time, we ignore half the cost of raising children. We only look at marginal returns to education in the labor market, but we ignore non-market returns. At the time of Gary Beckers "Theory of the Allocation of Time," there was no national data on time use, and there is still none.
Group Discussion and Comments
Frank Stafford, University of Michigan: Question for Zick: I am trying to look at data for Scandinavia. There is a trend toward more investment in time. How does your data compare?
Zick: I have taken data from 1920s and 1930s and compared it with Michigan data for 1985 in the US and found that the amount of time spent by women on child care has not changed, despite the decline in fertility and rise in womens employment.
Robert A. Pollak, Washington University: Why the notion that all time is productive? The issue I want to raise is: Are we trying to build a production-based index, which is the conceptual framework for national product accounts? The other side is a utility of welfare approach. People care how they spend their time. Time in various activities are direct arguments in ones utility function. Are we moving over to a utility-based approach as we think about these indices?
The Becker paper on time allocation was very casual about the measurement of output. The case in that paper was the easy case, constant returns to scale and no joint production. Maybe we have made less progress because Becker did not tell us about the problems we have encountered.
What about volunteering? What about elder care? Someone is doing it, yet, in a large cross-section, we record very little of it less than one hour per week.
John Robinson, University of Maryland: One reason for not including elder care in Robinsons time-use survey is because the amount of time spent on this activity is relatively small. They got more general information on volunteerism. One problem is what level of activity people do report and how they subsume other activities into the one they reported.
Smeeding: But we really need to know what is going on here. It is important how much time is spent caring for elders. There is disagreement in different sources of data.
Robinson: You are just going to find that data in a cross-sectional study. You are running into a data collection problem.
Smeeding: I do not trust it. It should be a national policy to know whats going on, whos doing it, are women over-represented in elder care?
Robinson: I just think this is a kind of activity that is not reported.
Duncan Ironmonger, University of Melbourne: On the issue of child care, we need to make the unit of observation the child. As we can see from the different methodologies we have looked at, you get three to four times as much child care when you ask about secondary activities. So child-care reports are much higher in Australia. Different methodologies give us a very wide range. The methodology that we use has to be reliable so that we know what we are getting. If we asked elders about their time and who is caring for them, we might get a better measure.
Zick: To respond on child care in 11 state surveys, in all cases except Connecticut, respondents had to proactively bring up secondary activities, else they were not recorded. In Connecticut, all time was coded as child care (secondary) as long as a child was at home.
Huston: This is not the issue. The issue is how the child reports the time they spend with adults whether that is with parent or paid care-giver. To understand inputs into children, you have to look at it from the childs perspective. Is some adult doing something with them actively? It does not have to be the parents.
Robin Douthitt, University of Wisconsin-Madison: When parents check on children from work via the telephone, this is not picked up by time-use data. The time parents spend making health-care appointments for children or for elders, or doing other primary-care activities, does not get counted as care time. This is a methodological issue of importance. These short events may be missed in data collection. When women are employed, they have a lot of hidden time in the paid-work domain to do things and be available for the child, contrary to reports that child care declines when mothers go to work. Children who live in single-parent households do more household chores, sometimes with the parent, and this can be "quality" time (e.g., going to the Laundromat).
Finally, an old-fashioned idea, time management which was looked at in old surveys found that in families there is a person who is the time manager for the family, and that is usually the woman. The stress of this management is high. Stress and quality of time do matter.
Michael Bittman, University of New South Wales: Comment on elder care he agrees with Robinson that volunteering constitutes a tiny amount in time-use diaries. A marker you can use is the amount of time spent in personal care and grooming where spouses do this care for their partners to keep them out of institutional care. So these may be an intra- rather than inter-household exchange.
Thomas Juster, University of Michigan: If a woman is home and her child is next door playing, she cannot leave, but this is not really time use. This will never be picked up as child care by a time-use survey, and we would not want it to be. What you need to ask is, does this person have responsibility for children? However, this is not a use of time that respondents will see. This is not a time use in the conventional sense of the term.
William Nordhaus, Yale University: Do we want to think about valuing different units of time differently? Can we put a price tag or something like it on time use? He read a Swedish study that found Swedish women are not better off than American women in terms of time spent on housework and care-giving. However, Swedish women reported liking what they do more often than American women.
Michael Horrigan, BLS: When we are thinking about Robinsons remarks, a lot of discussion centers around what you get out of time use and what you need in terms of context variables. This has important implications for designing surveys. Also, how often do we need to do time-use surveys? How often do you think the time-use distribution changes? What about adding modules to existing longitudinal studies and doing those in waves rather than annually? We need to think about where to put our resources.
Smeeding: He referenced Barbara Wolfes paper on the value of human capital in non-earnings related domains, which is similar to his paper. People are increasingly responsible for investment and health decisions, for example, and if you just look at income, you miss this return to human capital.
Huston: We could do assessments for children in strategic ways. After-school hours are an important time for children, especially primary and secondary school-aged children. Telephone calls at 6 p.m. about after-school time use would reveal a lot. Adolescent pregnancy (i.e., conception) and delinquency are most likely to occur between 3 and 7 p.m. In terms of policy, we need to look at those hours. Do children have other activity options during these times? Do they have time constraints, such as having to care for an elder? In terms of elder care, this may determine certain opportunities and constraints for the care-giver that may not show up in time use studies.
Michael: Nordhaus should have been here yesterday to get the answer to his question. Responding to Pollak: we have an output, but we do not know how to measure input. He disagrees with Ironmonger that we need to study productivity. We can study inputs instead.
Stephen Jenkens, University of Essex, provided the web address for the Economic and Social Research Council (ESRC)-Informal Economic Activities web site: www.esrc.ac.uk/ieaspec.htm
Session ended at scheduled time.