Related BLS programs | Related articles
July, 2000, Vol. 123, No.7
On the decline in average weekly hours worked
How many hours per week do workers in the United States spend at their paying jobs? The answer can be found by examining two principal BLS surveys used to track the number of hours that Americans work per week. The Current Population Survey (CPS) shows that there has been little change in average weekly hours worked; from 1964 to 1999, there was a decline of 0.5 percent in the average weekly hours at work in nonagricultural industries. This statistic contrasts information on the average workweek from the Current Employment Statistics (CES) program, otherwise known as the "establishment survey" or the "payroll survey." Here, data show a long-term downward trend in the average length of the workweek. From 1964 to 1999, average weekly hours fell by a substantial 11 percent, from 38.7 to 34.5 hours, based on annual averages of monthly data.
Considering that most people do not differentiate between paid and unpaid work, it becomes clearer why these two labor economics surveys from the BLS report contradictory data on the workweek. The most apparent reason is that the two surveys use different sources of information, resulting in a variation in the type of data gathered.
The CPS survey is a household survey; the CES survey is an establishment survey. The CPS hours data is based on workers’ reports on the hours they actually worked and includes all jobs they held during the survey reference period.1 The CES survey represents employers’ reports on the employees’ paid hours of work. If a person works for more than one employer, the hours are reported separately for each. For example, in the CES a person working two part-time jobs of 20 hours a week is counted as having two 20-hour jobs, but in the CPS , the same individual is counted as one worker working 40 hours. In May 2000, 5.7 percent of all employed persons 16 years and older were multiple jobholders. Thus, the CPS is the appropriate survey to use to examine trends in a person’s average workweek, while the CES is used to examine trends in the average number of hours people spend at each job.
Further, the scope of workers covered by the two measures differs. The CPS presents data for the total civilian noninstitutional population, while the CES hours data are limited to the private sector. The CES reports hours data for production workers in the goods-producing sector and nonsupervisory workers in the service-producing sector. The data sources and scopes of the two surveys explain why they could have different trends in the length of the workweek, but why does the CES measure show declining hours?
This excerpt is from an article published in the July 2000 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 (40K)
1 The CPS provides data on several different concepts of hours worked. Each month, all survey respondents are asked about the total hours worked at all jobs during the survey reference week; a quarter of the sample respondents each month are asked about the usual hours worked per week on the primary job; and, each year in a supplement to the CPS conducted in March, all survey respondents are asked about the usual hours worked per week during the last year. For further information, see "Hours of Work," Report on the American Workforce (Bureau of Labor Statistics, 1999), Chapter 3.
Related BLS programs
Labor Force Statistics from the Current Population Survey
Nonfarm Payroll Statistics from the Current Employment Statistics (National)
Related Monthly Labor Review articles
Trends in hours of work since the mid-1970s.—Apr. 1997.
Within Monthly Labor Review Online:
Welcome | Current Issue | Index | Subscribe | Archives
Exit Monthly Labor Review Online:
BLS Home | Publications & Research Papers