By John Crudele
The Census Bureau had trouble compiling the nation’s jobless rate in March and, because of this, I’m told, questionable shortcuts were taken.
So when the Labor Department announces the new unemployment rate on Friday at 8:30 a.m. — and experts are predicting the figure will remain at 5.5 percent — take it with a giant grain of salt.
Census, which is part of the Commerce Department, tallies the unemployment rate after conducting thousands of household surveys. It then projects what it gets from those surveys into a nationwide result.
As you probably already know, the Federal Reserve is among the many organizations and companies that carefully monitor the unemployment rate and make important decisions based on that number.
So it’s a big nuisance if this figure is unreliable.
But that’s exactly what it is — unreliable to the point of being nearly useless at best and fraudulent at worst.
Just so you have it straight, let me go over the list of players I am mentioning in this column: Labor hires Census, which is part of Commerce, to conduct interviews that lead to the figure we know as the official unemployment rate.
If you’re the type who can’t get enough government gibberish, you should know that this figure is also called the U-3 unemployment rate, which is the “total unemployed as a percentage of the civilian labor force.”
Here’s how the U-3 survey works. Census randomly picks the homes that are to be surveyed. Labor requires that 90 percent of those surveys be successfully completed. Most of the homes are getting repeat surveys, so not everyone is a first timer.
Before I started reporting on the method and lack of honesty in taking these surveys back in 2013, Census was easily able to reach that 90 percent goal.
After The Post investigation, Census started coming up short. That led to a congressional inquiry and a probe by Commerce’s inspector general.
Unable to cheat once a light was shown on its shady methods, all six Census regions showed declining survey success rates. March’s numbers were way down.
Census was only able to complete 85.59 percent of March unemployment surveys nationwide, sources tell me. And that below-the-threshold number was only reached after Census extended its survey two extra days.
The survey should have ended on Tuesday, March 24, but was extended until Thursday, March 26, because the success rate was only 80.38 percent as of that Monday night.
I’m not an expert on statistics, but this awful performance would seem to invalidate the results of a survey that relies on consistency to measure the ebb and flow of employment.
When the survey finally ended on March 26, not one of the six Census regions had hit the 90 percent goal. The Atlanta region came closest at 89.19 percent, sources said. New York, which has always been a laggard, had a miserable 81.27 percent.
So how much faith will you have in Friday’s unemployment number?
But wait, there’s more.
A source tells me that one of the six regions had to fudge its numbers to make its results look just a little lousy.
I’m told the region that cheated did so by changing a considerable number of interviews from what are called Type A to Type B.
Type A interviews are those that couldn’t be completed, so they count against the 90 percent goal.
Type B interviews also haven’t been completed, but for a legitimate reason — like a house at the address picked randomly was vacant or the person who is supposed to be living there has a primary residence somewhere else. Type B interviews — even when fraudulently classified as such — are not counted against the 90 percent goal.
The cheating region, I’m told, even had to bring in someone new from outside its area to oversee the March survey. Not everyone, apparently, was willing to change Type A’s to B’s.
Labor didn’t respond to my request for a comment.
The other number that’ll be released on Friday is total nonfarm payrolls.
Experts think that the number of jobs in this country increased by 260,000 in March. That would be a good increase — albeit considerably less than the 295,000 increase during February.
Labor collects the data for this survey itself, so there’s no issue with Census quality control.
But there is an issue with rogue seasonal adjustments that have made job growth look stronger in recent months. To add to the confusion, Labor starts adding very optimistic guesstimates to its job totals in spring for companies it thinks — but can’t prove — are just coming into being.
Add all this together and I don’t have a clue as to whether Friday’s job totals will be better or worse than the experts expect. Flip a coin.