How economic data has been distorted during COVID-19
Share Now on:
How economic data has been distorted during COVID-19
Another 1.1 million Americans filed unemployment claims last week, according to the Labor Department, underscoring the slow road ahead for economic recovery.
Throughout the COVID-19 crisis, there’s been a proliferation of data as people pay greater attention to the monthly jobs reports and unemployment claims during this time of economic upheaval. In a series of recent tweets, President Donald Trump engaged in one his favorite pastimes: touting the strength of the economy. He cited job numbers, the stock market and the housing market.
Some representations of data that have been floated around by news organizations, President Donald Trump and the Bureau of Labor Statistics depict job gains and losses in a line graph format, suggesting a massive economic recovery. (Trump’s chart also incorrectly shows a massive plunge in jobs occurring in late 2019, as opposed to the beginning of 2020.)
The U.S. economy saw a record 20.7 million job losses in April this year, but an increase of 2.7 million jobs in May (revised upward from the 2.5 million the Labor Department initially reported), about 4.8 million in June and about 1.8 million in July.
Alberto Cairo, author of the book “How Charts Lie” and a journalism professor at the University of Miami, explained that a line chart isn’t the proper method to represent this type of data set.
“The line chart exaggerates the difference, because you were down there and now you’re here, and that represents a huge change. But the change is not that big,” Cairo said. “It is more clearly represented through a bar graph with a zero baseline, and bars going down or going up.”
Connecting a line between the April and May data essentially creates the illusion that the U.S. recovered more than 22 million jobs.
This line chart from the St. Louis Fed, based on total nonfarm employees, also more accurately reflects how job gains have failed to bring the workforce back to pre-COVID levels:
While there was an increase of 1,800 employees added back to the workforce between June (more than 137,800 workers) and July (about 139,600 workers), showing the total amount of employees allows us to see that the latest figure is much lower than February 2020’s height of more than 152,400 workers.
As another example, Todd Gabe — an economics professor at the University of Maine — said that U.S. employment in restaurants and lodging dropped by 49% between February and April. Then, between April and July, employment increased by 48%.
But while the percentage increase between April and July was almost identical, that doesn’t mean job gains in this sector were back at where they were in February.
When you look at total numbers, there were 14.4 million jobs in February, meaning that in April, there were only 7.4 million jobs. A 48% jump between the months of April and July, from 7.4 million, only brings you back to 11 million.
“We’re still 24% lower in July than February,” Gabe explained. “Our economy is not used to such large changes over such a short time period. What we’re looking at now is unprecedented.”
Cairo also pointed to watch out for Trump’s use of charts, when defending his performance, that only show data during his presidency.
“That’s not enough,” Cairo noted. “If you want to see the trend, you need to show several presidencies in order to compare yourself to previous presidents.”
Economic growth, for example, began under President Barack Obama and continued under President Donald Trump until the COVID-19 crisis hit. However, Trump has taken credit for achievements — like job growth — that occurred prior to taking office.
“So when a politician gives you a big number or a very small subset of data, always try to zoom out and put the data in context. A data point lacking context is always meaningless,” Cairo said. “A chart is not an illustration. A chart is an argument. And in order to understand an argument, you need to read that argument and think about that argument.”
Within the jobs report, the COVID-19 crisis has also made it tricky to gauge metrics like wage growth.
Heidi Sheirholz, senior economist and director of policy at the Economic Policy Institute, said that year-over-year wage growth was massive in April, when average hourly earnings stood at 7.9%. (Hourly earnings have previously hovered above 3% in previous months.)
But Sheirholz pointed out that wasn’t because workers were seeing meaningful raises — it’s because so many low-wage workers were losing their jobs.
Now we’re seeing declines, with July posting 4.78% wage growth.
“And it’s not because people were seeing big cuts. It’s because we’re adding a whole bunch of low-wage workers back,” Sheirholz said.
Jeff Tucker, an economist at Zillow, said one data point in the housing sector that people should be careful with is the share of delinquent mortgages, which is at an all-time high. About 3.6 million homeowners are in forbearance, representing 7.2% of loans, according to the Mortgage Bankers Association.
But because of the CARES Act, homeowners can get a forbearance due to financial hardships caused by COVID-19.
“That stat at this moment is a really big deal. But it’s a very, very different picture than saying more than 7% of mortgages are delinquent, which in any other time in history would be a major red flag,” Tucker said. “Just grabbing that delinquency number can create a misleading impression of how many distressed mortgages are on the market.”
Tucker added that the big question after those 12 months are up will be: “How many of the homeowners in forbearance are able to get back on their feet and get current with their mortgages again?”
Heidi Sheirholz from the EPI said that while data can be misinterpreted, it’s crucial for our understanding of the economy and for good economic policy.
“Smart people, armed with high quality data can absolutely help policymakers make much, much better decisions for the country,” she said.
Cheers to trustworthy journalism!
Give just $7/month to get your own KaiPA glass.