What to watch on jobs day: Omicron will weigh heavily on the labor market

While the official pandemic recession ended two months after it began, it is clear the pandemic is not behind us, as the Omicron variant has driven a huge uptick in COVID caseloads. On Friday, I’ll be looking for the fingerprints of Omicron on the jobs report, including top-line payroll jobs as well as labor force participation. I’ll also continue to track sector-level job shortfalls, notably the lack of public-sector job growth, and differences in the economic recovery by race and ethnicity. With the release of January data comes annual benchmarking procedures as well: the establishment survey is benchmarked to unemployment insurance tax records and the household survey incorporates new population controls.

The Centers for Disease Control and Prevention (CDC) COVID tracker shows that nearly seven times the number of cases were reported during the January reference week (January 9-15) compared with the December reference week. Average new caseloads exceeded 800,000 in the week ending January 15, the peak of Omicron in the United States. This is nearly five times the peak level during the Delta surge (164,000) and more than three times the peak last winter (250,000). The labor market experienced a slowdown in payroll employment growth during the Delta surge, and that is likely to happen again in January (or even a temporary decline).

The Census Bureau’s Household Pulse Survey also provides striking evidence of what to expect in the January jobs numbers. The number of people not working between the survey periods ending on December 13, 2021 and January 10, 2022 rose by 6.5 million. This dramatic rise is primarily due to a three-fold increase—5.8 million more people—reporting they did not work because they were caring for someone or sick themselves with coronavirus symptoms. Figure A illustrates the dramatic uptick in people not working because they were caring for themselves or someone else in the most recent survey.

Spike in people not working as Omicron surges in early January: People not working because they have or were caring for someone with Covid, in millions

Date Weighted population (millions)
2020-05-05 1.31
2020-05-12 1.81
2020-05-19 1.93
2020-05-26 1.64
2020-06-02 1.73
2020-06-09 1.86
2020-06-16 2.06
2020-06-23 2.35
2020-06-30 2.79
2020-07-07 3.06
2020-07-14 3.90
2020-07-21 3.81
2020-07-28 NA
2020-08-04 NA
2020-08-11 NA
2020-08-18 NA
2020-08-25 NA
2020-08-31 2.70
2020-09-01 NA
2020-09-08 NA
2020-09-14 2.20
2020-09-15 NA
2020-09-22 NA
2020-09-28 2.39
2020-09-29 NA
2020-10-06 NA
2020-10-12 2.54
2020-10-13 NA
2020-10-20 NA
2020-10-26 2.95
2020-10-27 NA
2020-11-03 NA
2020-11-09 4.21
2020-11-10 NA
2020-11-17 NA
2020-11-23 4.82
2020-11-24 NA
2020-12-01 NA
2020-12-07 5.88
2020-12-08 NA
2020-12-15 NA
2020-12-21 6.59
2020-12-22 NA
2020-12-29 NA
2021-01-05 NA
2021-01-12 NA
2021-01-18 6.65
2021-01-19 NA
2021-01-26 NA
2021-02-01 5.90
2021-02-02 NA
2021-02-09 NA
2021-02-15 4.30
2021-02-16 NA
2021-02-23 NA
2021-03-01 3.52
2021-03-02 NA
2021-03-09 NA
2021-03-15 3.12
2021-03-16 NA
2021-03-23 NA
2021-03-29 2.63
2021-03-30 NA
2021-04-06 NA
2021-04-13 NA
2021-04-20 NA
2021-04-26 2.49
2021-04-27 NA
2021-05-04 NA
2021-05-10 2.14
2021-05-11 NA
2021-05-18 NA
2021-05-24 1.89
2021-05-25 NA
2021-06-01 NA
2021-06-07 2.11
2021-06-08 NA
2021-06-15 NA
2021-06-21 1.75
2021-06-22 NA
2021-06-29 NA
2021-07-05 1.78
2021-07-06 NA
2021-07-13 NA
2021-07-20 NA
2021-07-27 NA
2021-08-02 2.02
2021-08-03 NA
2021-08-10 NA
2021-08-16 2.94
2021-08-17 NA
2021-08-24 NA
2021-08-30 3.95
2021-08-31 NA
2021-09-07 NA
2021-09-13 4.65
2021-09-14 NA
2021-09-21 NA
2021-09-27 4.20
2021-09-28 NA
2021-10-05 NA
2021-10-11 3.72
2021-10-12 NA
2021-10-19 NA
2021-10-26 NA
2021-11-02 NA
2021-11-09 NA
2021-11-16 NA
2021-11-23 NA
2021-11-30 NA
2021-12-07 NA
2021-12-13 2.96
2022-01-10 8.75
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Population sample for people 18 and over who have not worked in the last 7 days.

Source: Authors’ analysis of microdata from the Household Pulse Survey (HHPS) from the U.S. Census Bureau.

Copy the code below to embed this chart on your website.

The same data also report a striking increase of people not working because they did not want to be employed, were concerned about getting or spreading the virus, their employer temporarily closed due to the pandemic, or other reasons. Counterintuitively, there was a notable reduction in the number of those who said they were not working because they were retired, which may be driven by seasonal factors. But in the long term, there has been an uptick in retirements in the pandemic in part due to fewer retirees returning to work; as the virus recedes, I suspect that will turn around and more will return to the labor market.

To be clear, the Household Pulse Survey will not translate one-for-one into the jobs data we will see on Friday because it has a different survey methodology and doesn’t consider seasonal swings—which can be very dramatic between December and January. However, it paints a clear picture that the pandemic surge in January will likely matter for both the number of jobs created as well as potentially softening participation in the labor force.

What is fairly certain, though, is that the number of workers absent due to illness will reach historic levels. The relative scarcity of workers in recent months has been one of the factors contributing to an increase in worker bargaining power, particularly in leisure and hospitality. As the virus comes under control, I expect more workers to flood back into the labor force, likely removing the temporary leverage workers had to quit their jobs and take better ones that would secure faster wage growth.

On Friday, I will also continue to track the sectors that exhibit the largest job shortfalls since February 2020. A temporary dip in leisure and hospitality employment may occur because of the rise in Omicron and the tendency for jobs in that sector to be more sensitive to the ebbs and flows of the pandemic.

Another sector I’ve been watching closely is the government sector, particularly state and local public-sector jobs, which have been slow to rebound compared with the private sector. As shown in Figure B, private-sector employment fell further during the pandemic recession—losing 16.5% of employment—but rebounded faster than state and local government jobs—which initially lost 7.7% of employment. Recently, state and local employment has been floundering and hasn’t seen any improvement in six months.

Private-sector employment fell further but rebounded faster than state and local government jobs: Percent change in payrolls since February 2020, for all private and state and local government employment

Total private-sector employment State and local employment
Feb-2020 0.00% 0.00%
Mar-2020 -1.30% -0.40%
Apr-2020 -16.50% -5.20%
May-2020 -13.90% -7.70%
Jun-2020 -10.20% -7.50%
Jul-2020 -9.00% -6.60%
Aug-2020 -8.20% -5.30%
Sep-2020 -7.50% -6.20%
Oct-2020 -6.70% -6.80%
Nov-2020 -6.50% -6.90%
Dec-2020 -6.70% -7.00%
Jan-2021 -6.60% -6.40%
Feb-2021 -6.10% -6.90%
Mar-2021 -5.50% -6.60%
Apr-2021 -5.40% -6.40%
May-2021 -4.90% -6.10%
Jun-2021 -4.30% -5.30%
Jul-2021 -3.70% -3.90%
Aug-2021 -3.30% -4.00%
Sep-2021 -3.00% -4.20%
Oct-2021 -2.40% -4.60%
Nov-2021 -2.20% -4.70%
Dec-2021 -2.00% -4.70%
ChartData Download data

The data below can be saved or copied directly into Excel.

Source: Bureau of Labor Statistics' (BLS) Current Employment Statistics, Establishment Survey (CES) public data series.

Copy the code below to embed this chart on your website.

A significant share of the public-sector shortfall has been in state and local education jobs, a sector that has experienced shortages in a variety of occupations, including both teachers and support staff such as bus drivers and teaching assistants. Fortunately, states and localities can use American Rescue Plan funds to shore up employment and invest in the fragile infrastructure within education, public health, and other vital systems. It is imperative that they do so.

Other metrics to follow within the household survey include what’s happening by various demographic groups. Black unemployment ticked up again in December to 7.1%, while the overall unemployment rate fell to 3.9%. Although the survey exhibits a fair amount of volatility month-to-month, longer-term trends clearly show that white workers have fared better than Black or Hispanic workers at the height of the pandemic recession and in the recovery thus far.

On Friday, the Bureau of Labor Statistics (BLS) will also release their annual benchmark revisions to the payroll survey based on administrative data from state unemployment insurance tax records. There are two important things to note about this release. First, the preliminary benchmark announcement from August suggests a downward revision in nonfarm employment, by 166,000 in March 2021. This reflects a large downward revision in private-sector employment (-421,000) and a large upward revision in government employment (+255,000). Again, these numbers are preliminary and will be updated on Friday.

Second, this benchmarking process will not tell us anything about what has happened with employment changes since March 2021, since each month from March 2021–December 2021 will get adjusted by the same amount. (The total change will however be “wedged” in from April 2020 to March 2021.) It is worth noting that these benchmark revisions are unrelated to the “sample-based” revisions that may occur in the two months following each initial monthly data release.

The household data released on Friday will also reflect updated population estimates for data beginning in January 2022. These population controls change the weights in the household sample to better reflect changes in the population. Notably, these changes are not used to update prior month’s data so caution should be used when comparing December 2021 data to January 2022 statistics in the household survey. The BLS will provide comparisons of some relevant metrics for users to assess the impact of the population controls.