Labor input is measured as either the number of workers employed or the number of hours they worked during a given time period, say a year.
Most countries collect data on the number of workers, and can calculate labor productivity as output per worker.
Hours worked account for differences between workers (full-time vs. part-time, vacations, strikes, etc.) and is more precise.
To measure labor productivity we prefer to compare the number of hours worked to the output produced during that time.
Some countries, including the United States, collect data on hours worked, making it possible to measure output per hour worked.
U.S. hours worked are calculated from data collected in monthly BLS surveys of businesses and households.
The Current Employment Statistics (CES) survey collects the number of employees and the number of hours they were paid to work from nonfarm businesses.
The Current Population Survey (CPS) collects the number of workers of all types—employees, self-employed, and unpaid family workers—and the hours they actually worked. It is a survey of households, including farms.
To measure labor productivity we want to compare the number of hours actually worked to the output produced during that time. Hours worked is different than paid hours, because some workers have paid sick leave and vacations.
Paid hours from the CES are adjusted to hours worked using data from the National Compensation Survey (NCS), also conducted by BLS.
Hours worked by employees of nonfarm businesses account for about 90% of hours worked, so the CES is the primary source of labor input for productivity measurement.
CPS hours worked by self-employed workers and unpaid family workers in the nonfarm sector, as well as hours worked by all farm workers, are added to hours worked by employees based on the CES to get total hours worked in the U.S. business sector.
Like the measure of output, labor input does not include hours worked in government, private homes, or nonprofit institutions.
Hours worked used for labor productivity measurement are not adjusted for differences in workers, such as education, experience, skills, etc.—every hour worked is counted the same.
Labor hours for multifactor productivity measurement relies on the same data sources and methods for employment and hours worked as described for labor productivity measurement.
However, labor input for some MFP measures is adjusted to take account of differences in worker characteristics, such as age and education.
Not every work hour contributes the same amount to output. The productiveness of an hour of labor can differ for college-educated professionals, teenagers in their first jobs, and experienced precision craftspeople.
It would be difficult to define all the “types” of workers, and further to measure each type’s work hours. Instead, the labor input contributing to some multifactor productivity estimates is adjusted to account for demographic changes to the workforce over time—changes in the education levels and age distribution of workers.