The Non-Farm Payrolls Report: Air Brushing History

[girl gone mad]

Jesse's Café Américain
http://jessescrossroadscafe.blogspot.com/2012/02/non-farm-payrolls-report-air-brushing.html

Back in Stalinist Russia, they had whole departments of people that were responsible for rewriting history and documents in order to support the latest Party lines. When a particular person fell out of favor, for example, they not only altered the documents, but even went so far as to air brush them out of important historical photographs.

Today the US reported a remarkably high Non-Farm Payrolls number, well in excess of even the most optimistic estimates. 243,000 jobs added, and unemployment has dropped to only 8.3 percent. Isn't that good news indeed.

If one tracks the data closely, and keeps their own copies of the records, what we see instead are revisions, sometimes going back as far as ten years, that most greatly affect the 'seasonally adjusted' numbers, but also affect the raw numbers as well.

The Obama Administration, as well as the previous Administration, have been going back and tinkering with history, rewriting the numbers here and there, in most cases 'rolling jobs forward' to the current months to make the current headlines look better.

more...

This is troubling. The government chooses to distort reality rather than confront the truth. Meanwhile, our real unemployment problem goes unaddressed.

[liberalmuse]

1. Could you post a reputable source?

I've never heard of this blog...

[enlightenment]

2. Perhaps you could identify a reputable source?

While I too would like to see some examples of what the blogger is saying, I don't know if I can think of a 'reputable' source to verify.

Obviously, within his post, he identifies enough of what he is looking at in the BLS data to allow for independent verification, but I also realize that many will not want to do that work (I'm not sure I have the analytical skills to do it; statistics was never my strong suite).

It is, certainly, an interesting idea and not outside the realm of the possible, even if we don't like the implication.

[girl gone mad]

3. Jesse is a reputable source.

He is well known and well respected in the financial blogoshpere. He called the housing bubble and the financial crisis well in advance. He has a strong background in mathematics and business.

[Ghost Dog]

4. Yes. This is a serious independent financial analyst,

not given to histrionics, has been solid for years now.

[FogerRox]

6. IT should be common knowledge

U3 numbers are massaged.

[Po_d Mainiac]

11. Massaged?

U mean teased & stroked. Just like the methods used by the ladies down at stress relief parlor!

[Ghost Dog]

5. Actual jobs, not seasonally adjusted, are down 2.9 million over the past two months.

... Actual jobs, not seasonally adjusted, are down 2.9 million over the past two months. It is only after seasonal adjustments – made at the sole discretion of the Bureau of Labor Statistics economists – that 2.9 million fewer jobs gets translated into 446,000 new seasonally adjusted jobs...

/... http://www.zerohedge.com/news/trimtabs-explains-why-todays-very-very-suspicious-nfp-number-really-down-29-million-past-2-mont

[FogerRox]

7. NJ Star Ledger recently said there were 20 million looking for work in the US

We could easily use 20 million jobs.

[Yo_Mama]

8. Well, I don't read this guy

But I did read the linked post and he sure doesn't seem to know what he's talking about in this post.

EVERY January the population basis is adjusted for the Household Survey. The Household survey is used to calculate unemployment rates, the emp/pop ratio, and labor force participation ratio.

The Birth/Death model is constantly readjusted, but there was a change fairly recently to adjust it more often. Explained here:
http://www.bls.gov/ces/ces_quarterly_birthdeath.pdf

The B/D model is based on results of the quarterly Business Employment Dynamics survey. Home Page for that here:
http://www.bls.gov/bdm/

As you'll notice, the most recent survey release is from Q2 2011. That's because this is very high quality data compiled from actual job/firm records at the states. Some technical info about that here:
http://www.bls.gov/news.release/cewbd.tn.htm

The Birth/Death model is used in the Establishment Survey (the jobs number he references). The reason you need the (business) Birth/Death adjustment is THAT the survey is an actual business survey from a selected sample that is as representative as they can get it with what they know at the time. Because some businesses always are failing, but other businesses are always starting up, the Establishment Survey numbers would always be biased on the negative side if it weren't for the B/D adjustment.

BUT, since the B/D adjustment is based on a survey which is always from quarters ago, the B/D adjustment will lag reality during times of rapid economic change. So if the economy is rapidly improving, the B/D adjustment will be too low, and if the economy is beginning to contract, the B/D adjustment will be too high.

When they get more recent information, they revise the adjustment and go back and recalculate all their imputed figures. There is no other way to do it.

There is no fakery here. When both these surveys are going the same way, it's a pretty good guess that that's what is actually happening. You can't compare January household survey numbers to December household survey numbers well, because every year the population adjustment (from Census) produces a discontinuity. That's why BLS doesn't disclose month-to-month changes in the Household Survey for January.

All of this stuff, and much more exhaustive detail, is fully disclosed on the BLS website. The comment about not disclosing B/D numbers is most confusing - here they are:
http://www.bls.gov/web/empsit/cesbd.htm

If you want the historical CES stuff, here it is:
http://www.bls.gov/ces/tables.htm

If you want the long-term historical data for the Establishment Survey data, go to this page and scroll down to the end:
http://www.bls.gov/news.release/empsit.nr0.htm

If you want the decadal benchmark adjustments, here they are:
ftp://ftp.bls.gov/pub/suppl/empsit.compaes.txt

None of this data is being futzed. These surveys are samples, and so they are not 100% precise. Movements over a few months mean more than movements over 1 month. Yet if you are concerned about data quality, every month BLS published the significance tests:
ftp://ftp.bls.gov/pub/suppl/empsit.tab1.txt

Absolutely nothing about these numbers are being distorted. The current administration is not lying about anything.

[fasttense]

9. I looked it up on the BLS

And yes there is a noticably decline in the workforce participation rate.

I find it interesting that when they adjusted, it lowered the rate instead of increasing the rate. You would think with continued population growth, you would have a continuing workforce participation rate increase. Then perhaps my significant other suggested more retirements. So, we looked up participation rates by age. There is no decline in workforce participation rate in the 55 and old crowds.

Here are some links to look at. The charts are very helpful.

http://data.bls.gov/timeseries/LNS11300000

http://www.bls.gov/emp/ep_table_303.htm

If the participation rate continues to decline, we can reach 3% unemployment without creating a single job.

[Demeter]

10. This isn't news to the Marketeers on DU

The books have been cooked at least since Reagan. I don't think Nixon ever thought to do it.

I don't expect to see anyone who writes for a living (Krugman, Friedman, etc) to admit to it--after all, they serve a master.

But Jesse, Shadowstats, Bill Black, Michael Hudson, etc. have no reason to lie...they are just royally pissed off that the government is trying to conceal reality for political gains, making a big mess even bigger.