Spiking unemployment rates may not be telling full story in the US, and for the states

We know that we've seen the highest job loss since the Great Depression in America. But this recent article from Shahar Ziv in Forbes says
the unemployment may be much higher than the already-huge figure that's listed today.

The main reason why the 14.7 percent figure is underestimating the true rate of unemployment is a quirk in the BLS methodology. In gathering data for the unemployment survey, interviews are conducted and individuals are classified as either employed, unemployed, or not in the labor force based on their answers to a series of questions. As Betsey Stevenson, who was a member of the Council of Economic Advisers as well as the Chief Economist of the U.S. Department of Labor, highlighted, there was incorrect classification of many individuals in April. “Interviewers were told to classify people who were employed [but] absent from work due to COVID-related reasons as temporarily unemployed. Many did this incorrectly —correcting for this error raises the unemployment rate to nearly 20%,” she explained.

To its credit, the BLS realized and called out this technical misclassification in its report explaining that, “to maintain data integrity, no ad hoc actions are taken to reclassify survey responses.” The misclassification caused the BLS to understate the unemployment rate by roughly five percentage points, meaning the adjusted unemployment rate is really closer to 20%.

In looking at the state-by-state unemployment figures that came out on Friday, you can see that every state had significant jumps in their unemployment rates, but some had bigger increases than others.

You can see that Minnesota was an exception to this rule, as their unemployment rate rose to "only" 8.1%. And a trip to their workforce agency's website explains why.

In April about 76,294 people moved from being in the labor force - employed or unemployed - to being outside the labor force - neither employed or unemployed - resulting in a labor force participation rate of 67.6%, or 1.7 percentage points lower than in March (which itself fell 1.1 percentage points from February). This indicates that many people are choosing to drop out of the labor force during this period rather than hold a job or continue looking for work. This may include people on part-time work schedules or who haven’t worked recently but had been actively seeking employment in March....

“Employed but absent from work for other reasons”: The BLS national employment situation for April mentions that due to coding errors, many participants who were unemployed on temporary layoff were miss-categorized as “employed but absent from work”. As a result, the national unemployment rate should have been about five percentage points higher than the reported 14.7%. However, the CPS data for Minnesota did not reveal any unusual surge in the category of workers employed but absent from work.

By comparison, Wisconsin actually added to its labor force in April, going against a trend of a falling labor force over the last couple of years.

Wisconsin household survey, April 2020
Participation rate +0.2%
Labor force +6,700
Employed -334,300
Unemployed +341,000

If we had dropped 1.7% off of our participation rate like Minnesota did, that would have been another 88,000 people out of the work force, and our unemployment rate would have been 11.6% instead of 14.1%.

Which means that there is a lot of volatility with the unemployment numbers in these unprecedented times. Let's see whether Wisconsin gets dropouts of the work force in May because they aren't considered to be looking for work, or if Minnesota has their unemployment rate spike because respondents say they are.

Lots still to sort out, and given that millions of new unemployment claims continue to be filed, it seems likely that May will see even higher unemployment. But because of the subjective methods we use to ID whether someone is looking for work, it's not certain that we'll get the full story of just how bad things are until revised data comes in later.