Comparing childcare measures in the ATUS and earlier time-diary studies

Monthly Labor Review

Volumn 130, Issue 5, 2007, Pages 27-36

  Allard, Mary Dorinda ^a   Bianchi, Suzanne ^b   Stewart, Jay ^a   Wight, Vanessa R ^b  

^a BUREAU OF LABOR STATISTICS   (United States)

^b UNIVERSITY OF MARYLAND   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 34548187218     PISSN: 00981818     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Article

References (25)

1. Liana C. Sayer, Suzanne M. Bianchi, and John P. Robinson, "Are Parents Investing Less in Children? Trends in Mothers' and Fathers' Time with Children," American Journal of Sociology, July 2004, pp. 1-43;
2. Suzanne M. Bianchi, "Maternal Employment and Time with Children: Dramatic Change or Surprising Continuity?" Demography, November 2000, pp. 139-54.
3. Time-diary data, which describe a person's activities on a given day, are considered to be more accurate for activities such as household work and childcare than are data gathered using stylized questions, which ask respondents to report about time spent on an activity over time, such as for a week ("About how much time do you spend taking care of children per week?"). See John P. Robinson, "The Validity and Reliability of Diaries versus Alternative Time Use Measures," in F. Thomas Juster and Frank P. Stafford, eds., Time, Goods, and Well-Being (Ann Arbor, University of Michigan Survey Research Center, Institute for Social Research, 1985).
4. W. K. Bryant, "A Comparison of the Household Work of Married Females: The Mid-1920s and the Late 1960s," Family and Consumer Sciences Research Journal, 1996, vol. 24, pp. 358-84.
5. The 1965 and 1975 time-diary studies were conducted by the Survey Research Center at the University of Michigan, and the 1985 and 1995 studies were conducted by the Survey Research Center at the University of Maryland. See Suzanne M. Bianchi, John P. Robinson, and Melissa A. Milkie, Changing Rhythms of American Family Life (New York, Russell Sage, 2006);
6. John P. Robinson and Geoffrey Godbey, Time for Life: The Surprising Ways Americans Spend Their Time (University Park, PA, Pennsylvania State University Press, 1999).
7. Common secondary activities include listening to the radio, watching TV, eating, or communicating, but in principle they could be anything that could be done as a primary activity (except travel).
8. In the ATUS, individuals are considered to be "with" the respondent if they were in the same room or if they accompanied the respondent at locations away from home. Earlier time-diary studies did not have explicit definitions.
9. Muriel Egerton, Kimberly Fisher, Jonathan I. Gershuny, and others, "American time use 1965-2003: The Construction of a Historical Comparative File, and Consideration of its Usefulness in the Construction of Extended National Accounts for the USA," ISER Working Paper 2005-28 (Colchester, University of Essex, December 2005);
10. Suzanne M. Bianchi, Vanessa R. Wight, and Sara B. Raley, "Maternal Employment and Family Caregiving: Rethinking Time with Children in the ATUS," paper presented at the ATUS Early Results Conference, Bethesda, MD, Dec. 9, 2005.
11. Nancy Folbre, Jayoung Yoon, Kade Finnoff, and Allison Sidle Fuligni, "By What Measure? Family Time Devoted to Children in the United States," Demography, May 2005, pp. 373-90.
12. The main criterion for the ATUS concept of secondary childcare is that the respondent must be able to provide assistance to the child if necessary. This implies that the respondent is in the general vicinity of the child and has a general idea what the child is doing. However, the respondent need not be in the same room as the child to be providing secondary childcare. Thus, the respondent may not have been "with" a household child under age 13 when providing secondary childcare at home - the child could have been in another part of the house or in the respondent's yard. It is also possible, although not common, for the respondent to have been "with" a child under age 13 and not have provided secondary childcare. One way this could occur would be when one or more adults were present, and the respondent did not consider himself or herself to be looking after the child.
13. Separate questions are asked for the respondent's children living in the household, respondent's children not living in the household, other children living in the household, and other children not living in the household. For all but the last, the children's names are filled in from the household roster.
14. Previous U.S. time-diary studies used a variety of different codes, and the actual activity descriptions may have varied. However, previous studies would have identified the nature of secondary activities.
15. Floyd J. Fowler, Survey Research Methods, Applied Social Research Methods Series, vol. 1 (Newbury Park, CA, Sage Publications, Inc., 1993).
16. Cognitive testing of the secondary childcare questions revealed that "in your care" best conveyed the passive childcare concept to respondents. See Lisa K. Schwartz, "The American Time Use Survey: cognitive pretesting," Monthly Labor Review, February 2002, pp. 34-44.
17. "Own" children are either biological children, stepchildren, or adopted children. Other relatives under age 18, such as grandchildren, would not be considered "own" children.
18. The response rate in the NSP was 64.0 percent.
19. This adjustment would tend to increase time with children relative to the ATUS, because all primary and secondary childcare time would be counted as having a child present, even though it appears that this is not always the case in the ATUS. As discussed in footnote 9, the respondent need not be "with" a child when providing secondary childcare in the ATUS. It is also possible, in both surveys, for a respondent to provide primary childcare without a child present. For example, the respondent may be driving to school to pick up a child. It is likely that the NSP misses some time with children for nonchildcare episodes, although we do not believe this effect is very large.
20. The response rate in the ATUS was 57.6 percent for 2003-04.
21. The last two restrictions were made because cognitive testing of the ATUS questions revealed some inconsistencies across respondents in how they answered the secondary childcare questions (some respondents included times when they or all household children under age 13 were asleep while others did not). For this reason, official estimates of secondary childcare exclude times when the respondent or all household children under age 13 were asleep. To determine when the household children were asleep, the respondents were asked when the first child under age 13 woke up and when the last child under age 13 went to sleep (naps are ignored).
22. Our conclusion that primary childcare is comparable between the ATUS and earlier time-diary studies is at odds with Egerton et al., who compared the 2003 ATUS to the earlier U.S. time-diary studies conducted in 1965, 1975, 1985, and 1992-94 (see Egerton and others, "American Tme Use 1965-2003"). They noted that primary childcare "steeply increases" between the 1992-94 study and the 2003 ATUS, and concluded that while sample composition may explain some of the increase," ...it also seems likely that there is a strong instrument effect." Our estimates using the same data combined with data from the 1995 University of Maryland time-diary study and the 2000 NSP lead us to believe that it is the 1992-94 data that are anomalous. Primary childcare time fell by 1 hour per week between the 1985 and 1992-94 studies, but increased by about 2 hours per week between the 1992-94 and 1995 studies, by 3 hours between the 1995 and 2000 studies, and by 1.5 hours between the 2000 NSP and the 2003 ATUS. Thus, we agree that time spent in primary childcare did increase between 1985 and 2003, but it seems more likely that there was a gradual increase between 1985 and 1995, rather than a decrease between 1985 and 1992-94 and a sharp increase between 1992-94 and 1995.
23. One difference between the surveys that we have not discussed is the difference in the procedures used to contact respondents. The NSP called respondents every day until the respondent was reached, while the ATUS used a designated-day approach. It has been shown that the NSP approach tended to oversample days when the respondent was away from home (see Jay Stewart, "Assessing the Bias Associated with Alternative Contact Strategies in Telephone Time-Use Surveys," Survey Methodology, December 2002, pp. 157-68). This could bias estimates of childcare upward if childcare tends to be done away from home and downward if childcare tends to be done at home. Taking a quick look at the data, it appears that the two datasets do not differ much with respect to where primary childcare activities occurred. About 59 percent of primary childcare episodes (77 percent of time) were at home in the NSP, compared with about 57 percent (73 percent of time) in the ATUS. This suggests that the difference in contact procedures did not have a large effect on the childcare measures.
24. Our inability to restrict time with children to children under age 13 for nonhousehold children in ATUS likely made very little difference. Very little secondary childcare was done for nonhousehold children, and ATUS estimates generated using only data on household children were virtually identical.
25. We noted earlier that probes for the "who-with" questions were inconsistently applied in the NSP. However, we do not believe this effect to be large. For example, the percentage in the last column of table 3 for the NSP would be at most 1 percentage point higher if we were to assume that a child was present during all episodes of secondary childcare.