Rear View Mirror Jobs

2006-08-04

Today's "question of the day" comes from John Succo at Minyanville. Following are his questions and thoughts up to the label "Mish:".

As everyone stares at the employment number I ask why? 71% of all new jobs this year are accounted for in the birth death adjustment.

So all we have to do is decipher the birth death adjustment to see what is really going on. So exactly what is it? We don't know.

The BLS will not allow anyone, ANYONE, to see how they calculate this number.

So everyone clamors over a number that gives us no real information and only has meaning because other people attach meaning to it.

Meanwhile, housing is looking weaker and weaker and housing really is driving almost all economic activity...at least it has in the recent past. And housing is rolling over with rates here even if the Fed pauses.

Hovnanian (HOV), an important housing stock, just cut numbers dramatically citing lower sales and higher cancellations. That is a fact. So the company is having trouble NOW with rates HERE. In addition, the time lag essentially means that part of that trouble began at even lower rates.

Doesn't that mean that a pause by the Fed in rates is too late? What good would a pause do if housing is slowing dramatically with rates here? It would take a rate cut to help them marginally from here.

Bulls trading futures are buying because others are buying.

Mish:

MarketWatch reported U.S. nonfarm payrolls rise 113,000 in July.

The U.S. economy added fewer-than-expected 113,000 nonfarm jobs in July and the unemployment rate rose to 4.8%, its highest level since February, the Labor Department said on Friday. The payrolls number was weaker than the 143,000 gain economists surveyed by MarketWatch were expecting. "Overall, too soft to hike," said Ian Shepherdson, chief U.S. economist at High Frequency Economics. Shepherdson said the Fed was now done raising rates.

The birth/death model took away 57,000 jobs in July. Those not familiar with the Birth/Death Model should read "Strike Three" for an explanation or the rest of this discussion may not make much sense.

Absent the Birth/Death adjustment, the jobs number was 170,000 rather than the reported 113,000. Yet, nearly 70% of employment gains this year came from birth/death model adjustments by the BLS.

This is the fourth miss in a row vs. job expectations. I talked about the last miss in "Strike Three".

Is it that the numbers are irrelevant, or is it that the interpretation of those numbers is fraught with a ton of problems. Is the economy adding more jobs than is shown by the hard payroll data itself, or less? Is the BLS assuming one thing but the economy proving another?

Yes, I agree with professor Succo that looking at the raw figures from the government is meaningless. But is the process of trying to figure out trends to get a feel for what is really happening useless or not? I can not say for sure but I suspect he might agree with me that it is not.

There is no doubt that the BLS model will miss the turn. Employment numbers will lag at both the top and the bottom. They admitted as such themselves.

But as goes housing, jobs will follow. So when I see marginal growth in jobs at a time housing is falling off the cliff, I am going to be very suspicious of those numbers. There can be no question about it. Housing has fallen off a cliff.

When I looked at today's numbers I expected to see huge negative revisions in the "Birth Death Model" but they were not there.

Here is the chart of Birth/Death assumptions: (click on the image for a larger view)



Housing peaked in July or August of 2005. Yet leading up to that peak we see the BLS assumed a total of 501,000 jobs for the months of April thru July. This year, in the midst of an enormous housing slump, the BLS assumed the creation of 600,000 jobs. Not only that, but the birth/death model only subtracted 57,000 jobs this July vs. 72,000 last July.

This is obviously total garbage. We may not know if 2005 was understated or if 2006 was overstated but we do have a glimpse that at least one set of the BLS's numbers is pure nonsense. No matter how you look at it, there is a problem.

I strongly suspect that the BLS model is looking in the rear view mirror when it comes to jobs. They are currently assuming that the economy is growing faster now in terms of job creation vs. a year ago (not for one month but for a 4 month average), at a time when GDP has fallen to 2.5% and housing starts and home sales have fallen off a cliff. Either that or their model has gone totally haywire.

We may not know what the real numbers are, but we do knows know that relative to last year, the last 4 months of birth/death assumptions from the BLS are insanely optimistic. That to me is useful information.

Had the numbers shown massive declines in the job assumptions for July, it would have been one thing, but the actual numbers shown above for April through July (year over year) are clearly a joke.

Still the reflex action of the market has been to buy bad news. It happened again today (at least at the open) just as Succo commented "Bulls trading futures are buying because others are buying."

This will work until it doesn't. The longer it "works" the bigger the eventual blowup will be. At some point, job losses will matter simply because they have to. People out of work will not be buying much regardless of whether or not interest rates pause, or even decline.

Mike Shedlock / Mish
http://globaleconomicanalysis.blogspot.com/