By Reade Pickert and Olivia Rockeman
Omicron. Revisions. Big seasonal factors. Friday’s U.S. jobs report is poised to be a bit of a doozy.
Nonfarm payrolls forecasts range from a 400,000 monthly decline in January to a 250,000 advance, and the confluence of crosscurrents will likely make analyzing the report more challenging than normal. So much so that White House officials have already warned the report could be confusing or even misleading.
“This is going to be a particularly tricky one,” said Nick Bunker, director of economic research at Indeed Inc. “It is reflective of a really disruptive period during this pandemic.”
A report earlier Wednesday showed U.S. companies shed 301,000 employees from payrolls in January, the most since April 2020, as the omicron variant registered a swift blow to the nation’s labor market, according to ADP Research Institute.
Here’s a guide of what to look for — and what to look through:
The monthly jobs report is a mid-month snapshot of the labor market, and from a Covid-19 perspective, mid-January was ugly. The omicron variant drove coronavirus infections to record highs. Some businesses temporarily closed their doors and many workers called in sick.
“The wheels of the entire hiring process slowed down in January,” said Sarah House, senior economist at Wells Fargo & Co.
The highly infectious variant is guaranteed to have some impact on Friday’s numbers, but it will affect each of the report’s two surveys differently.
The establishment survey, which relies on responses from businesses and government agencies, is the basis for the monthly payrolls figure. The survey of households is what informs data points like the unemployment rate.
To be counted in the headline payrolls figure, employees must have worked or received pay for at least part of the reference pay period — which includes Jan. 12, in this case. If an employee with no sick pay didn’t work for any portion of that period, the data would show that the worker had lost their job.
In the household survey, if a worker has a job, but is not at work due to illness, they are still counted as employed. That could lead to a wide gap of the employment pictures painted by each survey.
“There’s definitely going to be a lot of noise,” said Brett Ryan, senior U.S. economist at Deutsche Bank AG. “The way to sift through that noise is the household survey, and specifically the people who reported not being at work due to illness.”
High frequency data point to a meaningful impact from omicron. Nearly 8.8 million Americans said they were not working because they were sick or caring for someone with coronavirus symptoms, according to a Census Bureau survey conducted Dec. 29-Jan. 10. And applications for unemployment benefits in the week ended Jan. 15 — which corresponds with the reference period for the establishment survey — jumped by the most in nearly 10 months.
What Bloomberg Economics Says…
Worker absenteeism and temporary business closures were prevalent in that period. Initial jobless claims increased to a local high, and Bloomberg Economics’ daily GDP tracker indicated a contraction of around 11% during that week.
–Anna Wong, Yelena Shulyatyeva, Andrew Husby and Eliza Winger, economists
The good news is most economists expect the January report to represent more of a hiring blip than a meaningful shift in the health of the labor market.
Federal Reserve Chair Jerome Powell described the labor market as “strong” last week and said the central bank expects the omicron-related softness in the economy to be “temporary.” Aneta Markowska, chief financial economist at Jefferies LLC, agrees.
“I don’t see any reason why we shouldn’t see a good snapback in the pace of hiring,” Markowska said. “We’re not going to go back to a million-type job growth, but I think we could go back to 500,000-per-month-type job growth for a while.”
That said, sustaining that level could be challenging due to a low unemployment rate and a smaller workforce today than before the pandemic.
“We’re also at a point in the cycle where you expect job growth to slow,” Ryan said. “There’s just not that labor slack left.”
One line in the report to watch is jobs related to holiday seasonal hiring. Normally most of these people depart in January and the BLS adjusts for the big swings. For example, in January of last year, nonfarm payrolls increased 233,000 from a month earlier when adjusted for seasonal factors. However, on an unadjusted basis, payrolls fell by more than 2.6 million.
But this time employers could have decided to hold on to more workers in a tight labor market, especially in sectors such as retail and warehousing.
The jobs report will also include the BLS annual benchmark revision, which aligns data from the establishment survey with state unemployment insurance tax records. The revisions will impact payrolls, hours and earnings figures. Up to five years’ worth of seasonally adjusted data are subject to revision.
Additionally, January figures derived from the household survey will reflect updated population estimates to incorporate data from the 2020 Census and other sources. Because of that, the BLS discourages direct comparisons between December and January figures, though in practice there’s no alternative.
Average hourly earnings, a measure likely to be closely watched by market participants due to its potential inflationary impacts, are expected to rise 0.5% from a month earlier and 5.2% from a year ago.
While wage growth has been exceptionally strong in recent months — as evidenced by other compensation metrics — the figure is likely to be distorted higher by compositional factors. It also still lags the increase in consumer prices, which have eroded Americans’ paychecks.
Like the payrolls figure, average hourly earnings are derived from the establishment survey. The surge in Covid-19 cases — and related absences — likely disproportionately affected high-contact, lower-paid service jobs. If those wages aren’t included in the calculation, higher-wage jobs will make up a greater share of earners and inflate the wage data.
More stories like this are available on bloomberg.com.