I don’t believe ‘em (jobs numbers)

NY Post ^ | August 3, 2012 | John Crudele

[lasereye]

I’ve long believed the Labor Department’s monthly employment statistics are horribly inaccurate. So bad, in fact, that they are hardly worth compiling.

But I never thought the numbers were fudged — until now.

The government reported yesterday that 163,000 jobs were created in July and that the unemployment rate rose by 0.1 percentage points to 8.3 percent.

While Wall Street’s first reaction to the job report was very positive, that’ll change in the weeks ahead.

Why? Because the strength of the report probably eliminated any chance that Ben Bernanke’s Federal Reserve would begin a new money- printing operation, known as quantitative easing, before the presidential election.

What really matters to you and me is that the economy is creating jobs again, right?

As I’ve explained before, Labor each month guesses at the number of jobs that it thinks are being created by newly formed companies that it really can’t prove actually exist. This is called the birth/death model, in case you want to impress your cocktail-party friends.

So far this year, 400,000 of these phantom jobs have been added to Labor’s count.

Yesterday’s report included the addition of 52,000 phantom jobs. In July 2010, by comparison, Labor subtracted 18,000 jobs because small, invisible companies were dying and killing jobs.

In July 2011, Labor subtracted 38,000 jobs. On average, 20,000 phantom jobs are subtracted in July.

This year, the economy is now officially growing much more slowly than it was last summer and Labor suddenly adds 52,000 jobs to the July count that it can’t prove exist.

Let me put this into a different perspective: When the economy was still booming in the summer of 2007, Labor subtracted 57,000 phantom jobs.

Could the fact that this is an election year have anything to do with this remarkable discrepancy?

[Perdogg]

ping

[avacado]

163,000 jobs created but 195,000 fewer people employed in July. Something ain’t right!

Dept of Labor:
http://www.bls.gov/news.release/empsit.a.htm

[napscoordinator]

It doesn’t matter about number of jobs. 8.3 is what is going to kill any chance of an Obama reelection. Of course we will be sattled with an idiot Romney so we all lose anyway. Lord 2013-2017 is gonna suck!!!!!

[MachIV]

The industry I’m in continues to decline with one plant closing in a few months that I know of. Somewhere some media source planted the idea that we were in a recovery somewhere a few years ago. There has been NO recovery. Period.

[KSCITYBOY]

btt

[Iron Munro]

Obama Smoke and Mirrors add "Virtual Jobs"

Seasonal And Birth Death Adjustments Add 429,000 Statistical "Jobs"

(largest seasonal adjustment for July in the past decade, possibly ever)

Happy by the headline establishment survey print of 133,245 which says that the US "added" 163,000 jobs in July from 133,082 last month?

Consider this: the number was based on a non seasonally adjusted July number of 132,868. This was a 1.248 million drop from the June print.

So how did the smoothing work out to make a real plunge into an "adjusted" rise? Simple: the BLS "added" 377K jobs for seasonal purposes.

This was the largest seasonal addition for a July NFP print in the past decade, possibly ever, as the first chart below shows. But wait, there's more: the Birth Death adjustment, which adds to the NSA Print to get to the final number, was +52k. How does this compare to July 2011? It is about 1000% higher: the last B/D adjustment was a tiny +5K!

In other words, of the 163,000 jobs "added", 429,000 was based on purely statistical fudging.

[Pearls Before Swine]

I think the numbers aren’t very meaningful either, but for a different reason. I think BLS has to make adjustments—and without a detailed look in their models, I have no idea what they should be, or whether they’ve been distorted in a particular month or quarter. But, the adjustments over time have proved to be larger, and sometimes much larger, than the resulting jobs number.

Any time this happens, you are looking for a “signal” as to what is happening in the presence of overwhelming noise. It’s like trying to listen to music on the radio from a very weak station. You can’t make it out, or tell what is going on.

So, what’s clear is that the jobs number is weak—but what you can’t do is conclude definitively that it is significantly better or worse than the month before, or indicative of any change in trend.

Did the BLS cook the numbers, or is the method such that the numbers are not very meaningful except six months after the fact when you have enough of them to discern a real trend? Does it make any difference? The numbers still suck.

I think the author is right to look at other statistics, like the total and percentage number of the workforce employed, to get a clearer picture.

I also think that the market didn’t go up because of the jobs number. Europe was up bigtime before the number was announced, because it sounds as if the ECB is going to find a way to buy questionable bonds from deadbeat countries, in a manner similar to the way the Fed “expanded its balance sheet” buying crap mortgage backed securities.

[mrsmith]

The market went up but volume was low, below average.

Absent growth,”new” jobs are still created by- death, retirement. There was some talk of how the numbers were messed up by the auto industry not having their usual layoffs.

[TomGuy]

But I never thought the numbers were fudged — until now.

Crudele has had his head in the sand for months.

Anyone who has watched the revised upward week after week and the unexpected week after week and the number of people magically disappearing from the work force since 2009 knows something stinks -- and it isn't dead fish.

[Cletus.D.Yokel]

:: In other words, of the 163,000 jobs “added”, 429,000 was based on purely statistical fudging. ::

There has got to be a less confusing way of stating this fact. As written, I try to reconcile how 429,000 is a ^subset^ of 163,000.

The 163,000 is a subset of the total job numbers to which 429,000 was arbitrarily added.

[Starboard]

The weekly report on the number of Americans filing applications for unemployment benefits that comes out every Thursday morning is a complete farse.

As an example of their perverted math, BLS reports an increase in the number of jobless claims (e.g., 10,000) then the following week revises it UPWARD (e.g., 15,000) to make it easier for the ensuing new weekly number (e.g., 12,000) to print as a ‘DECLINE”. Even though 12K is obviously an increase over 10K, its magically converted into a drop after the original 10K was conveniently revised upward.

Whatever the new weekly number happens to be, simply “adjust” the old number higher to make sure you can report a decline. This manipulation goes on week after week thereby allowing the MSM to almost always report a never ending “improvement” in jobless claims.

[Starboard]

Agree. The economy is bad and getting worse and America is rapidly losing its competitive edge. Remember the farsical charade by the administration (and their pals in the media) known as “Recovery Summer” back in 2010? This was nothing but a political fabrication to dupe the populace into believing that the economy was rebounding. The administration knew all along this was complete BS.

[Starboard]

“In other words, of the 163,000 jobs “added”, 429,000 was based on purely statistical fudging”

---

The BLS has been totally compromised under this administration. The monthly print would have been horrible if not for all the “adjustments”.

A poll today that was reported on Fox Business showed that nearly two thirds believe we are currently in a recession. You can’t fool all of the people all of the time.

[JimRed]

The current quote of 8.3% is a flat out lie. Counting those whose unemployment compensation ran out who are not yet employed, which the government figures ignore, it is more like 15%.

[Starboard]

“Did the BLS cook the numbers”

---

Let’s just say a lot of ordinary (non-statistical) people wonder about the huge and ongoing disparity between the monthly prints and U6 as well as the recurring “upward revisions” in the Thursday weeklies that are obviously tortured. Issues like these call into serious question the integrity of the BLS.

[Pearls Before Swine]

Let’s just say a lot of ordinary (non-statistical) people wonder about the huge and ongoing disparity between the monthly prints and U6...

I'm still not clear whether the Great Depression definition of unemployment was more like U6 than U3, myself. Now, THAT redefinition would be a switch of substantial importance. Did it occur, and when?

[blam]

The Employment Rate In The United States Is Lower Than It Was During The Last Recession

[khelus]

The BLS collects two sets of data on employment.

• The household survey from which the unemployment rate is calculated. It consists of data from ~60,000 households.
• The payroll survey from which the number of jobs created is taken. It consists of data from businesses.

The household survey is considered much more reliable than the payroll survey. However it is open to manipulation through the use of seasonal adjustment and removing people from the civilian labor force.

The payroll survey is open to manipulation through seasonal adjustment and the Birth and Death Model.

From this months payroll survey - Table B-1:
# jobs, not seasonally adjusted
June 134,072,000
July 132,860,000
=
-1,212,000 fewer jobs in July

After running -1,212,000 through 2 statistical tools: the Birth and Death model and seasonal adjustment

-1,212,000 becomes +163,000

[khelus]

a short version:

http://www.freerepublic.com/focus/news/2914558/posts?page=19#19