An official website of the United States government
Hours data underlying labor productivity and cost measures for major sectors include hours worked by wage and salary workers, the self-employed, and unpaid family workers. To construct hours of wage and salary workers, which account for almost ninety percent of hours worked, the BLS Division of Major Sector Productivity (DMSP) relies primarily on data from the BLS Current Employment Statistics program (CES). DMSP uses CES monthly survey data on the number of jobs held by wage and salary workers, as well as the number of jobs and average weekly paid hours of nonsupervisory and production workers, in nonfarm establishments. DMSP makes adjustments to these data using supplementary information on employee hours. First, the paid hours of nonsupervisory and production workers are adjusted to a measure of hours at work by an application of hours-worked to hours-paid ratios. Then the hours worked by supervisory and nonproduction workers are estimated by applying ratios of supervisory (nonproduction) to nonsupervisory (production) worker average weekly hours ratios. (Both types of adjustment ratios are described in more detail below.)
Because the private sector series shown here excludes government hours, it is most closely related to the hours measures published in the Productivity and Costs news release. To obtain the employee hours component of the published measures this series will be further adjusted to remove hours of employees working in nonprofit establishments, and to add hours worked by employees of government enterprises, which are operated for profit.
Supplementary data are also required to provide coverage of self-employed and unpaid family workers in the nonfarm sector. The BLS Current Population Survey (CPS) provides data on employment and hours at work by job which DMSP uses to construct hours for these categories of workers, which are added to private nonfarm employee hours to yield hours of all persons in the nonfarm business sector. The CPS also provides hours for all categories of farm workers, which are added to nonfarm hours to obtain business sector hours of all persons. Note that because the CPS directly collects hours at work for each job, across all occupations, there is no need for adjustments from an hours-paid to an hours-worked concept, or to account for differences in average weekly hours between supervisory workers and production workers.
Since the CES collects data on the paid hours of nonsupervisory workers, they include the hours for which an employee is paid but is absent from a job. This includes factors such as holidays, sick leave and vacation time. Since this time was not spent in the production of a good or performance of a service, it should not be included when measuring productivity. To remove these non-work hours from the hours-paid data, the DMSP uses hours-worked to hours-paid ratios developed from information on employer leave practices in the BLS National Compensation Survey (NCS). Detailed NCS ratios are applied at the NAICS 3-digit industry level back to 1996. From 1990 to 1995, ratios are estimated using data from the discontinued BLS Hours at Work Survey and adjusted to be consistent with the ratios from NCS. NCS ratios are incorporated on an annual basis and estimated for intermediate quarters. Prior to 1990, BLS uses ratios provided by the BLS Hours at Work Survey. (See "A Note on the Incorporation of Hours-Worked Hours Paid-Ratios from the Employment Cost Index into Hours at Work Measures" PDF.)
DMSP estimates hours worked by supervisory and nonproduction workers using information from the BLS Current Population Survey (CPS), which collects the average weekly hours worked for a sample of all persons in the civilian noninstitutional population. Using data on the average weekly hours of wage and salary workers by industry and occupation, DMSP calculates ratios of CPS supervisory worker average weekly hours to CPS nonsupervisory worker average weekly hours at the NAICS supersector level of detail. (For information on how these ratios are constructed, see "Construction of Average Weekly Hours for Supervisory and Nonproduction Wage and Salary Workers in Private Nonfarm Establishments" (PDF).