‘Crazy Mess’ of Jobs Data Has Economists Getting It Wrong Again

(Bloomberg) -- The economists got it wrong. Again.

After some estimates called for U.S. payrolls to decline by as much as 400,000, the labor market shockingly added that many jobs in January -- and then some. It’s not just last month: the median estimate has missed by an average of about 95,000 in the past year before revisions, versus less than 20,000 in the six months through March 2020.

Forecasting throughout the pandemic has proven to be a constant challenge -- economists have had to balance traditional indicators like job openings that are backward-looking, with high-frequency data like daily coronavirus cases that can be extremely volatile. But in January, economists underestimated how much businesses would retain workers and overestimated how many workers would drop off payrolls due to omicron.

“The January employment report is a crazy mess, but not in the ways that were expected,” Stephen Stanley, chief economist at Amherst Pierpont Securities, said in a note.

Most economists -- including those at the Federal Reserve and White House officials -- were bracing for a weak, or even negative payrolls print, due to the surge in Covid-19 cases. If the strong relationship between infections and employment held true, 2.3 million Americans would have lost their jobs in January, ZipRecruiter Chief Economist Julia Pollak said in a note.

Other indicators leading up to Friday also pointed to an impending doomsday. Jobless claims recorded during the employment report’s reference week surged the most in nearly 10 months, while ADP Research Institute data showed companies shed 301,000 employees in January, the most since April 2020.

Even if the negative projections came to fruition, most economists said the report would have done little to sway the Fed, which is expected to hike interest rates next month by 0.25 percentage point to curb the fastest inflation in nearly 40 years. The upside surprise, as well as a jump in hourly wages, only reinforced investors’ conviction further, and added to bets that the central bank would proceed with a half-point move.

One reason for the January payrolls advance was the Bureau of Labor Statistics’ seasonal adjustment, which helps account for employers dismissing workers hired for the holiday season. But in this case, especially in service sectors hit hard by omicron, businesses held onto employees due to the tight labor market.

“January is by far the biggest month for layoffs as seasonal work ends and schools close for winter break,” Wells Fargo (NYSE:WFC) & Co. economists Sarah House and Michael Pugliese said in a note. They had forecast a payrolls decline of 100,000. “However, with many businesses already understaffed, employers let go fewer workers than usual.”

Get The News You Want

Read market moving news with a personalized feed of stocks you care about.

Get The App

The job gains were broad-based, led by a 151,000 advance in leisure and hospitality. Transportation and warehousing, retail trade and professional and business services also posted solid increases.

What Bloomberg Economics Says...

“Neither the headline payrolls print nor the unemployment rate captured the full impact of the latest Covid wave. The household survey’s statistic, which counts workers who are employed but absent from work due to illness, did indicate a massive impact.”

--Anna Wong, Yelena Shulyatyeva, Andrew Husby and Eliza Winger, economists

Economists also famously got it wrong in April 2021 when they thought rising vaccinations would lead to more people returning to the workforce, and helping explain a forecast for a 1 million jump in payrolls. But the U.S. only added 266,000 jobs that month. Projections were overly optimistic about November too, overshooting payrolls by more than twofold.

Another quirk in January’s report was revisions to reflect updated population estimates used in the household survey. Had it not been for those controls, the number of employed Americans would have dropped by 272,000, according to the Labor Department. 

By the same token, the participation rate -- the share of the population that is working or looking for work -- would have been unchanged from December, rather than the registered 0.3 percentage point increase.

Despite the better-than-expected headline payrolls figure, the impact of omicron on the labor market in January was significant. There were 3.6 million employed Americans not at work due to illness, more than double that in December. At the same time, 6 million people were unable to work in the month because their employer closed or lost business due to the pandemic, roughly twice that in December.

“These absences do have a really big economic impact,” said Daniel Zhao, senior economist at Glassdoor. “It’s unclear how that will show up in the aggregate data, but is a trend to definitely keep an eye on.”

©2022 Bloomberg L.P.