The responses of less than 41,000 people were used to determine a major part of last month's U.S. unemployment rate, the Bureau of Labor Statistics tells Axios.
Why it matters: That's the lowest number in modern history and is one of many unusual developments in government data collection that have been impacting important readings for months.
What it means: The surprises in May's nonfarm payrolls report, which found there were only 21 million unemployed while 30 million Americans were collecting unemployment insurance benefits, were largely the result of oddities in data collection.
- A portion of the jobs report is determined by a household survey in which government workers interview people at their homes and determine whether any person over the age of 16 is "employed, unemployed, or not in the labor force" — the only three possible designations.
What's happening: The coronavirus pandemic has "depressed" survey responses since March, as BLS stopped conducting in-person meetings, restricting its ability to reach new households, Julie Hatch Maxfield, BLS associate commissioner for employment and unemployment statistics, tells Axios.
- "The first month of the sample we get a lot of information and that sets up the whole thing going forward," she says.
- This has taken the response rate from 82% in January to 73% in March to 67% in May.
- "Response rates probably will be depressed even when interviewers go back into the field," Maxfield notes.
What else: In May, BLS identified 9 million people who had lost their jobs but were counted as "not in the labor force" rather than unemployed because they hadn't been searching for a job in the last four weeks due to the pandemic.
- If those people were considered unemployed it would have taken the unemployment rate to 17.9%.
- A similar calculation would have put the unemployment rate at 19.8% in April and 7.5% in March, BLS says in a report about the coronavirus pandemic's impact on its data.
- A separate "misclassification error" categorized millions of workers who had been absent and likely lost their jobs as employed.
- Additionally, workers who were paid by their employer for any part of the pay period including the 12th of the month were counted as employed, even if they weren't actually at their jobs.
The bottom line: We may never know how many people lost their jobs as a result of the coronavirus pandemic.
- However, according to BLS, both the number of unemployment insurance claims and the number of unemployed people appear to have peaked in April and are declining.