Normally a couple of days after the release, I'll read about the tens or hundreds of thousands of people the government has declared to "no longer be in the labor force" thus allowing for the improvement in the unemployment rate.
This month was different.
The workweek for non-farm employees decreased by two tenths (0.2) hours in April. The article said this was the equivalent of losing 718,620 jobs.
But that number was based on math and logic errors.
Using the employed figure of 143,724,000 and multiplying that by 0.2 hours is the loss of 28,744,800 hours.
The author divided that number by a 40 hour week to arrive at the 718,620 jobs.
There are two problems with that number. First, the average work week is down to 34.4 hours. The other issue is comparing a week to a month.
First we'll convert those 28,744,800 lost hours to an average work week. April had 22 Monday to Friday business days, so we divide by 22 and then multiply by 5 to get a work week's number of hours. That's 28,744,800 / 22 * 5 = 6,532,909 hours.
Since the average work week is now 34.4, the equivalent number of jobs lost is 189,910 (6,532,909 / 34.4).
Not quite as impressive a headline as losing over 700k jobs, but it's still more of a loss than a gain. The official report said we gained 165k jobs when we lost 190k jobs' worth of pay. The net effect would be a loss of 35k jobs for April.
Here at The Financial State of the Union we've learned:
1. Never take a government report at face value - check the details.
2. Never assume the numbers are correct - check the math.
3. Always be on the lookout for bad assumptions or comparisons - in this case comparing a week to a month.