Tomorrow morning, when the Bureau of Labor Statistics reports the unemployment rate for March, don’t be surprised if it’s very low. I predict that it will be not much more than 4 percent and maybe even less. But that statistic will not really tell you much about the unemployment rate in March.
Here’s why. As the BLS explains on its web site:
Each month, highly trained and experienced Census Bureau employees contact the 60,000 eligible sample households and ask about the labor force activities (jobholding and job seeking) or non-labor force status of the members of these households during the survey reference week (usually the week that includes the 12th of the month).
Almost all the layoffs happened after March 14.
But the unemployment rate for April that will be reported on May 1, which, ironically, is known as International Workers Day? OMG!
READER COMMENTS
Patrick T Peterson
Apr 2 2020 at 3:04pm
OMG!!!! is about right.
Richard A.
Apr 2 2020 at 4:14pm
The monthly unemployment rate from the CPS strongly correlates with the percentage of those collecting unemployment insurance (UI). From the weekly UI data it is possible to make a good guess what the CPI unemployment rate would have been in the last 2 weeks of March had the CPS sampled the last 2 weeks.
BTW, many Americans falsely believe that the monthly unemployment figures come from UI data.
Mike Davis
Apr 3 2020 at 3:50pm
Good call. The official number was 4.4%. But as you point out, the truth is much worse.
Start with these numbers
• Over the past 2 weeks about 10 million people have applied for unemployment insurance compensation (3.3 million for the week ending March 20, 6.6 million for the week ending March 27).
• In February BLS reported total civilian labor force of 164.6 million with 5.79 million unemployed, giving an unemployment rate of 3.5%
Now make the reasonable assumptions that
• There has been very little new hiring in March
• The initial claims number understates the actual number of people who have become unemployed. (Not everyone qualifies for or elects to apply for unemployment. Also, the state offices have been overwhelmed and many people have been unable to apply.) We can’t know the actual number until after the BLS releases the preliminary household survey but I’ve got to think that there are at least 2 million newly unemployed who aren’t part of the 10 million.
• The total labor force is still close to what it was at the end of February. This will probably change as people drop out either by retiring, claiming disability or just becoming discouraged. We know what happened to participation during and after the Great Recession. But that hasn’t happened yet.
Put all that together and you get an unemployment rate of close to 11% (round numbers, 18 million unemployed divided by 165 million civilian labor force.
And just in case you aren’t feeling bad enough already, remember that in the Great Recession unemployment peaked at 10% in October of 2009, 5 months after the recession officially ended. In November, 1982 unemployment peaked at 10.8%, again after the end of the recession. This recession is going to be like no other and in case I don’t know enough about employment dynamics in a downturn to offer a solid prediction but it would be very, very surprising if the peak unemployment rate happens at the start of a recession.
David Seltzer
Apr 4 2020 at 10:46am
Mike, “In November, 1982 unemployment peaked at 10.8%, again after the end of the recession. ” Nominal one year T-bill rates averaged about 13.5%, inflation about 6.16% It seems the Phillips Curve was wrong. Interesting point. Thanks.
Comments are closed.