FINALLY, a confession!
I’ve been writing in this column for years that the way the Labor Department calculates the number of jobs in this country is deceptive.
In 11 of the 12 months, the government adds massive numbers of jobs — sometimes more than 100,000 — that it thinks, but can’t prove, exist.
This is because the Labor Department uses something called the birth/death model, which assumes that no matter how bad the economy is, there are itty-bitty, newly-formed companies — which can’t be reached by government surveyors — that are creating jobs.
I’ve said that this assumption is ridiculous — that these companies probably don’t exist. And non-existent companies can’t create jobs.
Not only is the public fooled by this practice, but policymakers are being led astray. So the Labor Department, essentially, lied again when it reported last Friday that only 263,000 jobs disappeared from the economy during September.
As shocking as that figure was, the more truthful number was actually worse because there were 34,000 of these phantom birth/death jobs included in the count. Nearly 1 million of these non-existent jobs have been added to the government’s count since the beginning of 2009, mostly this past spring.
As I’ve said before, these bogus jobs are what caused the employment numbers to look slightly better in March, April, May and June, and created the notion that the economy was somehow on the mend.
Right after Friday’s report came out, Bloomberg News called Chris Manning, the national benchmark branch chief at the Labor Department’s Bureau of Labor Statistics, and asked about the 34,000 probably non-existent jobs.
“In this period of steep job losses, the birth/death model didn’t work as well as it usually does,” Manning told Bloomberg. “To the extent that there was an overstatement in the birth/death model, that is likely to still be there.” No freakin’ kidding! This year alone, this model has added over 700,000 jobs that don’t exist to the government’s count.
The Labor Department is not only still using this model, but it nearly doubled the number of phantom jobs for this September compared with the same month last year.
Bloomberg was the only media outlet listening when this column started broadcasting the errors of the monthly job figures.
And while this glitch in the jobs report is now hot stuff in Wall Street circles, no other news organization has even bothered to look into this use of statistical razzmatazz.
Now that I’ve gotten that off my chest, let me tell you what’s next.
Over the next few months the birth/death model will be kinder and the monthly jobs report should be helped by these silly assumptions.
Then, on Friday, Feb 5., Labor will report the number of job losses for the month of January.
This is the one month in every year that the birth/death model removes jobs from the economy — lots of them.
Unless someone fixes the government’s broken system for guessing at job growth, that figure will be disturbingly bad.
And the entire discussion about the depth of this recession will turn especially glum early next year.
Mark down that date — Feb. 5! It should be a real sad affair.
*
As an aside to the above topic, The New York Times Magazine a few weeks ago used thousands of words to explore “How Did Economists Get It So Wrong?” There’s a lot of academic hokum in the piece, written by distinguished economist Paul Krugman, about Panglossian Finance and the “anti-Keynesian coun terrevolution.”
I’ll now give economists worldwide the only excuse they need to defend their calls.
When you put garbage statistics (see above) into your equations, you get garbage results. Economists didn’t have a chance to get it right.
“They are all using the same mathematical models and relying on poor quality statistics. They aren’t going to get anything reliable out of that,” said John Williams, an economist who runs ShadowStats.com and who did correctly call the downturn.
And, by the way, this column also had it right.