The Bureau of Labor Statistics reported today that the U.S. unemployment rate fell to 6.3% in April, the lowest level since the collapse of Lehman Bros. in September 2008, and the largest one-month drop in the jobless rate in three years.
Don’t believe it. The data come from a survey of about 60,000 households. In normal times, it’s pretty volatile. In April, it was absolutely bonkers. Many of the statistics in the survey are extreme outliers, producing values far outside the normal ranges.
It’s simply not believable that the labor market contracted so violently according to one survey, while at the same time a separate survey of business establishments showed strong payroll growth of 288,000.
Consider this: The jobless rate fell because 733,000 fewer people were considered unemployed. That’s the largest one-month decline since 1949, when unemployment fell by 920,000 when two large national strikes concluded.Nothing like that happened in April.
The number of people in the labor force plunged by 806,000, the largest since last October when the federal government shutdown pushed a lot of people out. Nothing like that happened in April.
The big reason for the decline in the labor force was a sharp drop of 417,000 in the number of people who have been previously employed and who started a job or started looking for a job in April.
That suggests that the extreme declines in April reflected the expiration of extended unemployment benefits to millions of people. The theory is that people who’ve been out of work for longer than six months are more likely to tell the BLS that they are actively looking for work if they are receiving a benefit check. Once the checks stop, so does the illusion that they’ll find a job, and they stop looking.
But extended benefits were cut off beginning in late December. If the expiration of benefits was causing hundreds of thousands of people to drop out of the labor force, it should have showed up in the data in January, or February. It didn’t.
Economist Joe LaVorgna of Deutsche Bank tweeted that “Unemployment rate fell because of 806k drop in labor force, a lagged effect from expiration of unemployment insurance.”
The question is why would the effect be lagged?
You probably shouldn’t pay too much attention to the household survey data for April. The Federal Reserve certainly won’t. The Fed will wait for a few months to see if the April anomalies continue or are revised away.
Most likely, the extreme moves in April are statistical noise. In any given month, the standard error for the household survey is plus or minus 300,000, the BLS says. How how is that figure.........???????