Related BLS programs | Related articles
December 2004, Vol. 127, No.12
What can time-use data tell us about hours of work?
Harley Frazis and Jay Stewart
The number of hours people work for pay is an important economic measure. In addition to being a measure of labor utilization, it is a component of other economic statistics. For example, productivity measures are computed by dividing total output by total hours worked, and hourly wages are often computed by dividing usual weekly earnings by usual weekly hours worked.1 There are two major sources of hours data for the United States—the BLS Current Population Survey (CPS) and the BLS Current Employment Statistics survey (CES)—and estimates of weekly hours from these two surveys behave differently for a variety of reasons. The goal of this article is to use data from the new American Time Use Survey (ATUS) to shed light on the accuracy of hours-worked reports in the CPS. Because the purpose of this study is to determine whether respondents report hours correctly in CPS, it does not examine other factors that could result in differences in estimates of hours worked from CPS and ATUS. In addition to differences in the reporting of hours, differences in estimates can be due to differences in sample composition and differences in the reporting of other variables.2 We control for these other factors, but do not analyze their effects on differences in estimates. We examine the effect of these other factors on comparisons of weekly hours from CPS and ATUS in a forthcoming publication.3
Previous studies that assess the accuracy of hours data from establishment surveys either compare hours data for the same industries across surveys, or evaluate accuracy using cognitive methods such as focus groups and interviews with respondents.4 The former approach allows researchers to document differences between surveys (after accounting for differences in concepts), while the latter provides information on how respondents compile their data. Studies that are directed at verifying hours measures from household surveys such as the CPS typically take one of two approaches: they compare weekly hours reports from a CPS-like question either to (1) records from the individual’s employer or (2) data collected from the individual using a time diary.
This excerpt is from an article published in the December 2004 issue of the Monthly Labor Review. The full text of the article is available in Adobe Acrobat's Portable Document Format (PDF). See How to view a PDF file for more information.
Read abstract Download full article in PDF (89K)
1 For a discussion of the importance of hours data for measuring real hourly wages, see Katharine G. Abraham, James R. Spletzer, and Jay C. Stewart, "Divergent Trends in Alternative Wage Series," in John Haltiwanger, Marilyn E. Manser, and Robert Topel, eds., Labor Statistics Measurement Issues, NBER Studies in Income and Wealth, Vol. 60, (Chicago, University of Chicago Press, 1998) pp. 293–324; Katharine G. Abraham, James R. Spletzer, and Jay C. Stewart, "Why Do Different Wage Series Tell Different Stories?" American Economic Review Papers and Proceedings, Vol. 89, No. 2, 1999, pp. 34–39; and Lucy P. Eldridge, Marilyn E. Manser, and Phyllis Flohr Otto, "Hours Data and Their Impact on Measures of Productivity Change." Paper presented to the NBER Productivity Program meeting, Boston, March 2004.
2 For example, the ATUS has a higher multiple jobholding rate than does CPS, which would tend to result in ATUS hours exceeding CPS hours.
3 Harley Frazis and Jay Stewart, "Where Does the Time Go? Concepts and Measurement in the American Time-Use Survey," in Ernst Berndt and Charles Hulten, eds., Hard to Measure Goods and Services: Essays in Memory of Zvi Griliches, NBER Studies in Income and Wealth (Chicago, University of Chicago Press, forthcoming).
4 Sylvia Fisher, Karen Goldenberg, Eileen O’Brian, Clyde Tucker, and Diane Willimack, "Measuring Employee Hours in Government Surveys." Paper presented to the Federal Economic Statistics Advisory Council, Washington, DC, June 2001; and Karen L. Goldenberg and Jay Stewart, "Earnings Concepts and Data Availability for the Current Employment Statistics Survey: Findings from Cognitive Interviews," in Proceedings of the Section on Survey Research Methods, American Statistical Association, 1999.
American Time Use Survey
Related Monthly Labor Review articles
Within Monthly Labor Review Online:
Welcome | Current Issue | Index | Subscribe | Archives
Exit Monthly Labor Review Online:
BLS Home | Publications & Research Papers