Decreases in American Jobs

On Oct. 6, the Economic Policy Institute reported, “Slowing Economy Generates Fewer Jobs”:

“The nation’s rate of job growth downshifted sharply last month, as employers added only 51,000 jobs, according to today’s report from the Bureau of Labor Statistics (BLS). This marks the lowest month for net job gains since the Gulf Coast hurricanes disrupted the labor market last fall. Even with a large upward revision to August’s job gains (188,000 — 60,000 more jobs than first reported), the average monthly gain this year has been 137,000, below last year’s monthly rate of 165,000.

“The unemployment rate ticked down to 4.6%, though the change was statistically insignificant. Employment in the household survey, which often moves quite differently than the payroll data noted above, jumped by 271,000. Here again, however, the Bureau reported that gains from this survey need to surpass 414,000 to indicate, with statistical certainty, that job growth was positive (the smaller sample and greater volatility from this survey leads to this large ‘confidence interval’)…

“Two other indicators of slowing demand in the job market come from the hours’ data. Average weekly hours were flat last month and have moved little this year. Combining the stagnation in weekly hours with the slowing rate of job growth, the BLS’ index of total hours worked by private sector, nonmanagerial workers fell in both August and September.”

Is There a Discrepancy?

According to Bloomberg, “U.S. Finds 810,000 More Jobs, Helping Explain Data Discrepancy”:

“The U.S. Labor Department found employers hired almost a million more workers in the year to March than previously estimated, bringing job gains more in line with what households had been saying all along.

“About 810,000 more jobs will be added to the payroll count when the figures are officially revised next February, the department estimated today. The proposed revision, the biggest since Labor started adjusting the numbers in 1991, would mean the economy created 2.6 million jobs from April 2005 through March, instead of the 1.8 million now on the books.

“The revision will go a long way toward solving the mystery of why households, in a separate Labor survey, were saying job gains during that period were much larger. Confirmation that employment was stronger may help explain why consumers haven’t buckled when faced with record fuel prices and a decline in the housing market.

“‘It explains the strength in incomes that we’ve seen and it explains why the unemployment rate has come down to a five-year low,’ said Conrad DeQuadros, a senior economist at Bear Stearns Cos. in New York. ‘There is some support to this economic expansion.’

“Once a year, the Labor Department revises its payroll figures after combing through tax records from the unemployment insurance program that covers practically all businesses. Those records are only available after a lag, explaining why it takes over a year to make the tabulations.

“Lack of Response

“The Labor Department said it is looking into why the expected revision this time around is three times larger than the average over the last decade. Part of the investigation involves determining whether the firms that are doing much of the hiring aren’t responding to their survey, said Kirk Mueller, chief of national benchmarking and special projects at the Bureau of Labor Statistics in Washington.

“Another possibility is that the department’s method of determining the formation and demise of businesses, called the birth/death model, may be faulty, he said.

“The divergence between the two employment surveys continued last month. The household survey showed a 271,000 gain in employment for September while payrolls were up a smaller-than-expected 51,000. Because the revisions suggest the poll of households may be capturing employment gains better than the payroll survey, the job market may be stronger than it looks, economists said.”

Overcoming my initial speechlessness at the idea that we added 810,000 more jobs, I have several questions. Five, actually.

Five Questions

  1. Assuming we did gain 810,000 jobs March to March, how does that affect the much ballyhooed productivity statistics? When will we see revisions?
  2. Did it really take 810,000 more people working to produce a GDP that is barely above the stall rate?
  3. If job expansion is so robust, why are real wages falling?
  4. Assuming we did gain those jobs, exactly what was the distribution of those jobs over time?
  5. What can we expect looking forward?

As interesting as I think Question 1 is, the key to what is going to happen economically is more likely behind door No. 5.

Let’s look at the “birth/death” numbers for 2006:


In Strike Four, I questioned the birth/death model, critical that it was adding jobs that simply did not exist. Now it seems that the Labor Department thinks 810,000 more jobs need to be added to totals that I thought were already exaggerated on the high end.

Pondering the latest birth/death revisions once again, I see construction jobs are being added by the model. 157,000 construction jobs were added between February-September, in the face of the massive housing slowdown. Is that plausible? Was construction that much understated in 2004, perhaps, that the model needed to add all of those jobs in 2005? Unless the revised numbers reflect one last push by builders to complete projects as all hell was breaking loose in sales, those numbers do not seem to make a lot of sense.

A Trend Change

Note that for the first time since January, jobs were being lost in leisure and hospitality. Jobs at restaurants such as Pizza Hut and Outback Steakhouse are included in that total. Is there any doubt this sector has been overbuilt? Still, the model suggests that 342,000 additional jobs (over the physical payroll data) were created between January-September 2006. Is that a plausible number? Actually a better question is is that a sustainable number?

Please consider the following snip about the birth/death model straight from the BLS:

“The most significant potential drawback to this or any model-based approach is that time series modeling assumes a predictable continuation of historical patterns and relationships and therefore is likely to have some difficulty producing reliable estimates at economic turning points or during periods when there are sudden changes in trend. “

Something is seriously wrong if 810,000 jobs have to be added on behalf of past activity. In addition, the economy has finally turned. The leading indicators think so; the yield curve thinks so; and, in fact, about the only things that do not think so are the stock market and the economic cheerleaders. Looking ahead, plan for more discrepancies in the numbers.

Some might suggest that commercial construction is still booming. Perhaps for now it is. But for how long? Commercial real estate tends to lag home construction by a year or so. As subdivisions build out, retailers and restaurants and gas stations start moving in. Strip malls multiply. Plans are made by Wal-Mart and Home Depot and Lowe’s and Pizza Hut to add stores. Those plans are set and that construction goes on, even if (as is happening now) housing is collapsing in its wake.

Housing where I live has pretty much slowed to a crawl, but we still have a huge new Wal-Mart going up nearby. That store will likely employ 500 people or more. Yet in the face of a continued slowdown, will growth of such stores continue at anywhere near the same pace? What do corporate executives think?

CEO Outlook

Investors.Com is reporting CEOs Gloomy On U.S. Outlook

“Nearly half (45.6%) of CEOs polled by the Business Council and Conference Board expect the economy to worsen over the next six months, versus 13.2% who think it’ll improve. In January-February, 28.4% expected better times, versus 16% who were downbeat. Most expect the economy to grow 2-3% in ’07.”

Are gloomy CEOs likely to be adding many jobs? More and more it seems the much ballyhooed “soft landing scenario” that commercial expansion will take over where the consumer left off is a bunch of nonsense.

Let’s now see if we can tackle Questions 4 and 5 above. Knowing that housing peaked in summer 2005, I suggest most of those jobs were filled by November 2005. Given the lag in commercial construction, the remainder of those jobs trickled in since then. Looking ahead, about the only possible bright spot is health care.

In other words, what was supposed to be excellent news might really be a nightmare. It took 810,000 additional jobs just to produce anemic real GDP growth of 2.6% during the second quarter. Subtracting CPI distortions, hedonics, imputations, etc., the reported 2.6% growth is likely at or under the stall rate. This brings to mind additional questions.

Still More Questions

  1. If the BLS was underestimating jobs created for March 2005-March 2006 by 810,000, then how much is it overestimating jobs by now if the economy has turned?
  2. If the Fed is basing some of its interest rates decisions on jobs, just how far behind the curve was it in tightening, and how far ahead of the curve is it now?
  3. If the Fed was aware that the economy was 810,000 jobs stronger than everyone thought, when did it know that, and why didn’t it comment?

Answers to the above questions would likely show the Fed has no business setting rates at all. The Fed is forever in a pointless tail-chasing exercise that causes it to wildly overshoot in both directions. Unfortunately, the Fed manages interest rates on a day-to-day basis, adding or draining reserves just to meet its own arbitrary targets, when, in fact, it does not know the correct interest rate any more than it knows the correct day-to-day price of orange juice.

If perchance the economy really did create an extra 810,000 jobs in some sort of housing/credit bubble blowoff top, that discrepancy will be unwound and then some in the upcoming recession.

Mike Shedlock ~ “Mish”