Government intervention at the onset of the coronavirus pandemic helped to prop up U.S. wages even as demand for labor plummeted by nearly 30%, according to an intensive review of job advertisement data.

Researchers affiliated with Chmura Economics & Analytics found in a paper published this month that the pandemic and the government response had surprising impacts on the labor market, diverging somewhat from their expectations.

Contrary to economic theories that say millions of job losses will put a downward pressure on wages, most occupations saw wages increase or remain relatively constant, which Chmura researchers attributed to employers paying more to entice people to work during a pandemic and government programs such as the Payment Protection Program.

The analysis was conducted using proprietary job advertisement data from Real-Time Intelligence, Chmura’s job posting database that is updated daily from over 15,000 sources, rather than official government statistics, which are released monthly or quarterly depending on the dataset.

“This pandemic, I think, is going to prompt a lot of economists to start using commercial data,” Xiaobing Shuai, senior economist and director of research at Chmura, said in an interview. “Because in the past, we always trusted the government-issued data, but there is a time lag in the time of COVID-19 — it becomes less relevant.”

In order to ensure the credibility of the data and corroborate their findings, researchers implemented quality control measures and compared the job listing data with official sources when they became available.

Job listings in the first half of the year were down 12.5% compared to the first half of 2019, but in January and February — prior to the spread of COVID-19 in the U.S. — job postings were 7.9% higher. Chmura’s research covers job listings through June, leaving the question of whether the trend continued later into the pandemic unanswered.

Occupations that suffered the steepest drops in number of job ads included those in food preparation and service, as well as those in arts, design, entertainment and media. Occupations in transportation and material moving saw a milder decline as many consumers shifted to online shopping, as did production workers.

Although wages mostly held firm, transportation and material moving occupations were exceptions, seeing a $5,000 decline in average wages in March through June when compared with January and February.

“It’s never been like this before in such a scale,” Shuai said. “In the past, there might be a little outbreak somewhere, then there’s a localized shutdown, but not on the scale that the whole country or the globe, many countries are shutting down.”

Shuai and his fellow researchers had expected infection rates would hurt labor demand, but that turned out not to be the case. Instead, death rates, which represents the magnitude of severely ill patients, were found to be a positive driver of labor demand.

“Even right now, because we have a surge, you will start reading reports that hospitals, they are low on workers, they need more help,” Shuai said. “Right now, a lot of hospitals have to add additional beds, stuff like that. So they would have a demand for the health care workers.”

Government interventions may have exacerbated the decline in overall labor demand, with expanded unemployment benefits offering a potential disincentive to return to the workforce if benefits were higher than wages, the researchers said. Paycheck Protection Program loans may have also dampened labor demand if businesses were incentivized not to hire more workers.

Data from the U.S. Bureau of Labor Statistics and Gallup suggests that roughly 80-85% of coronavirus-related layoffs could be categorized as temporary, the Chmura researchers wrote, but other researchers have estimated that between 32% and 43% of job losses during COVID-19 could be permanent. Further tracking of job ad data may be required to determine the full impact.

“People guess that certain jobs will never come back, but we don’t really know,” Shuai said.

The study “COVID‑19, labor demand, and government responses: evidence from job posting data,” published Dec. 7, 2020, in Business Economics, was authored by Xiaobing Shuai, Christine Chmura and James Stinchcomb of Chmura Economics & Analytics.