The monthly jobs numbers are based on two data sources, which use subtly different definitions of employment. For January, more than for most months, those differences could result in some mixed messages.
The number that usually gets the most attention, the count of jobs gained or lost, is based on a government survey that asks thousands of employers how many employees they have on their payrolls in a given pay period. People who miss work – because they are out sick, are quarantining because of coronavirus exposure or are caring for children because their day care arrangements have been upended – might not be counted even though they have not lost their jobs.
The payroll figure is meant to include anyone who worked even a single hour in a pay period, so people who miss only a few days of work will still be counted. Employees taking paid time off, too, are counted as working. Still, the sheer scale of the Omicron wave means that absences almost certainly took a toll.
The jobs report also includes data from a survey of households, which uses a different measure: It considers people employed if they report having a job, even if they are out sick or absent for other reasons.
Economists typically pay more attention to the survey of businesses, which is larger and seen as more reliable. But some said they would be paying closer attention than usual this time to the data from the survey of households because it will do a better job of distinguishing between temporary absences and more-lasting effects from Omicron, such as layoffs or postponed expansions.