Minyan Mailbag - The Shedlock Deadlock
Note: Our goal in Minyanville is to remove intimidation from the financial markets and encourage an interactive dialogue among the Minyanship. We share this insight from Minyan Mike Shedlock with that very intent.
First let's take a look at jobs added or subtracted via the birth/death model. For those Minyans not familiar with the birth/death model a simple explanation is as follows:
Rather than gather statistics directly from companies as a sole source of employment data, the BLS believes that during a recession that more businesses are being destroyed in bankruptcies (business deaths), and in an expansion more new businesses are being created (business births). The BLS then adds or subtracts the number of jobs it presumes are being created by new companies or companies going under, to the reported employment totals.
Here is my take on the reported statistics:
It is logical to believe that new businesses are being created while others are going out of business, but it sure seems that the BLS continually errs on the high side during the year then attempts to make up for it in semi-annual revisions in July and January. January of 2005 followed the pattern of subtracting huge numbers of jobs just as they did last July and last January. The pattern suggests to me that too many jobs were assumed over the course of the year that never really existed at all. Admittedly figuring out how many jobs to presume is tough, but by now it would seem that it should occur to the folks at the BLS that this is not exactly a normal recovery.
Taking a look at this January's revisions we see:
The BLS added 297,000 jobs from August to December.
The BLS then subtracted 280,000 of those jobs in January.
In this Minyan's mind, many of those jobs were not added or destroyed. They just never existed in the first place.
Simple subtraction shows that 17,000 hypothetical jobs were created in the second half of the year due to new business births and deaths.
I can believe that number. Actually I can believe triple that number of jobs was created but certainly not 297,000 simply because I refuse to count selling junk on eBay (EBAY), multi-level marketing opportunities, financial planners with one client, and/or freshly minted real estate agents that just will never make it as having a job. Furthermore, most new businesses simply do not make it. How many start ups go under in a year or two? How many financial planners or real estate agents do we need? See: Califorina Dreamin for a more in depth look at real estate agents in particular. No doubt SOME of those people will indeed do well, but %wise I would guess the number to be quite small. In the end, people need money and just take what they can get: a job at Home Depot (HD) or Walmart (WMT).
The Birth/Death estimates for Construction are particularly glaring.
I doubt that lots of new construction businesses were created in the last 6 months. I also doubt that many construction businesses went under in the last 6 months either. Yet supposedly we created 14,000 construction jobs due to newly formed construction businesses since August but then in January 60,000 construction jobs died due to construction businesses going under. More than likely, on purpose or not, the BLS badly overestimated the number of construction jobs that were created in the prior 6 months (January-June 2004). In January 2005 they are just now figuring that out, (guessing) those jobs never really existed in the first place.
The Business & Professional Services category is interesting too.
We supposedly created 83,000 jobs from August to December and destroyed 115,000 of them in January. Perhaps the BLS decided that all those people selling financial planning services do not really hold a job. Then again if they did that, perhaps they should have killed 250,000 jobs. I am not sure where real estate jobs are placed. If independent real estate agents are in this category, perhaps the BLS should be prepared to kill tens of thousands of those jobs later this year or next.
Enquiring Minyans now turn their attention to payroll and household data numbers.
Payroll & Household Survey FAQs
Here is the January 2005 official release:
THE EMPLOYMENT SITUATION: JANUARY 2005
In the Household Survey we see that the civilian unemployment rate dropped to 5.2% in January from 5.4% in December. That drop is very questionable since the participation rate normally rises in an expansion. Here we are, three years into an "expansion" with the participation rate dropping from 66.7% in November of 2001 to 65.8% in January of 2005. The labor force is expanding due to immigration and births and the number of employed persons is NOT keeping up with that expansion, yet the unemployment rate drops. I do not buy it. Statistically, the household survey sample is too small, cell phone users are not contacted, and there are potentially other statistical errors. If a person says he has a job then that job is counted, without proof, and regardless of number of hours worked or the pay. Someone's new "job" might consist of selling junk on EBAY, a multi-level marketing opportunity such as Amway, or financial planning regardless of how many clients a person has. I suspect that the vast majority of those "jobs" fail quickly.
In the Establishment Survey we see that non-farm payrolls increased by 146,000 in January. December non-farm payrolls were revised down to 133,000. Many economists, including Greenspan consider the Establishment Survey a more accurate indicator than the household survey. I would subscribe to that viewpoint given the problems with the household survey mentioned above.
Each month I like to monitor Table A-12: Alternative measures of labor underutilization.
In this Minyan's eyes, unemployment is dramatically understated. Once people expire their unemployment benefits they are no longer considered unemployed. If they feel there are no jobs out there but have not looked this past month they are not considered unemployed. If they are working part time but want a full time job they are still considered fully employed. The rules of unemployment have changed a number of times over the years and I believe some of these changes were made on purpose, to make things appear far better than they are. At any rate, from Table A-12 we see that if we included discouraged workers and marginally attached workers, the unemployment rate would be 7.3%. If we include people working part time for economic reasons (people who want a full time job but can only find part-time), the unemployment rate jumps to 10.9%. Some people like to brag about our "low" unemployment rate compared to say Germany, but if we counted unemployment the same way Germany did, our rate would be much higher too. Our unemployment rate is realistically between 8-11%, not the glowing 5.2% that was reported. If you were wondering why things do not seem as good as is reported, it is because things are not as good as reported! One just needs to know where to look. The proof's right there, hidden away in Table A-12. Simple enough?
The Northern Trust discussed the January employment numbers as well. Enquiring Minyans probably want to take a look.
The Hoover Award:
Many people were wondering if Bush would win "The Hoover Award", the first president to lose jobs on his term since President Hoover.
The president escaped this rather inglorious fate, primarily because of the expansion of government jobs.
Here is how close it was.
Nonfarm payrolls increased by 146,000 to a seasonally adjusted record of 132.573 million, as employment broke the February 2001 peak for the first time. With the payroll increase and an upward revision to past data, job growth in President Bush's first term totaled 119,000, letting him avoid the stigma of being the first president since Herbert Hoover to preside over net job losses.
That "revision", if one believes it, saved Bush in the bottom of the 9th, from the ignominy of being forever joined at the hip with Hoover.
Enquiring Minyans just might be wondering about private sector job growth vs. total job growth as the latter is arguably a better measure of productive economic growth.
By my math, President Bush started with 110,996K private sector jobs (end of 2000/start of 2001) and ended with 109,863K jobs (preliminary). The preliminary numbers showed president Bush lost 1.133 Million jobs on his watch.
The January 2005 report revised the total number to110,862K jobs.
If we accept that number as fact, my math says that Bush started with 110,996K jobs and ended with 110,862K jobs which is a net loss of 134,000 private non-farm jobs on his watch. I suspect someone at the BLS missed this bit of history and may be tempted to play with these numbers later to prove that Bush did not lose private jobs on his watch. In the same regard, I am rather suspicious of the total job revisions that allowed Bush to escape winning "The Hoover Award" by a mere 119,000 jobs.
Let's see how Clinton did with private jobs.
Clinton inherited 89,940 thousand private jobs.
Clinton ended with 110,996 thousand private jobs.
By my math Clinton created 21,056 thousand jobs or 21.056 Million jobs.
Bush lost 134,000 private non-farm jobs on his watch
Clinton gained 21,056,000 private non-farm jobs on his watch.
To be fair, one cannot really credit all of those jobs to Clinton. He was fortunate to have come along at a time in history when the internet was being built and job opportunities were everywhere. Likewise, one can not totally blame Bush for the loss of jobs under his watch. His fate in history was to be elected right as a boom was ended.
No matter how you slice it, this recovery has been brutally anemic in terms of job creation in spite of the record stimulus that was thrown at it.
Now that the FED is attempting to take away the punch bowl, and with the long bond hinting at recession, I sense this recovery is on its last legs.
Minyan Mike Shedlock
Daily Recap Newsletter