NFP Payroll: Special Hurricane Accounting Edition

Given the longevity of the streak, I once again will be taking "The Under" in the Employment Situation Survey today.

However, some of the recent data "adjustments" seen out of BLS make it hard to take any of the headline data seriously. My admonishments regarding the Fog of Katrina are coming all too true. Therefore, I am hard pressed to say whether or not NFP will be yet another example of creative accounting a/k/a Special Hurricane Accounting.

Lookie here: we know that most of the working population of New Orleans have been displaced after the Katrina. That’s somewhere between 250,000 and 400,000 employees no longer on the job, depending upon how  many were successfully relocated. We also know that the September NFP numbers showed only a modest drop of 35,000 workers.

What accounted for the difference? According to the BLS, for the purposes of the last NFP report, "a person is counted as employed if they are on a payroll during the survey week." In other words, you were still officially employed, even if only ’cause
your employer couldn’t physically track you down to lay you off.

This month, consensus is for +125,000 new jobs.  Those numbers simply do not compute — unless there has been an offsetting 1/4 million jobs created in NOLA.

How is this possible? Special Hurricane Accounting. Consider: 

Q3 Employment Cost Index: (ECI) Wages and salaries, adjusted for inflation, are a negative -2.3% in the quarter. The 12-month percent changes in ECI, unadjusted, is the slowest growth rate on record.

Q3 GDP (advance):  The big driver’s of GDP? Auto sales (GM and Ford giving
cars away at cost), and a big uptick in spending by Uncle Sam. Neither are thought to be sustainable.

But lets dig into some of the components:

-Final Sales (a "demand" indicator) dropped  from 5.6% in Q2 to 4.4% in Q3.   

-For the first time ever, Personal savings (savings as a percentage of disposable income) in Q3 dropped to NEGATIVE 1.1%.

Here’s the kicker: The Commerce Department could not separate any special effects from Katrina and Rita, which it said are embedded in the data.

Q3 Productivity: was much stronger-than-expected 4.1%, compared with 2.1% in Q2.

Sounds great, until you dig below the headlines: "The effects of Hurricanes Katrina and Rita, which struck the Gulf Coast during the third quarter, could not be separately identified in the productivity and costs measures."

Productivity is a function of Hours worked and output. We had slightly less output, but a huge decrease in hours worked, thanks to the job losses in the gulf. That made productivity gains appear large, when in fact it was function of so many less workers clocking in down south.

So you can chuck out the Productivity surprise as a BLS "not measurable."

Personal Income:   Consensus was for 0.3%, but the actual data was for 1.7%; How come? Hurricane insurance payments. They offset the lost "rents and lower proprietor income" and then some. 

Do note that Same Store Sales were robust, without much in the way of Hurricane distortions. So its not all negative data.

Bottom line on today’s crapshoot: Noobody knows, the data is confusing at best, and whatever number comes out today is likely to have as little meaning as any ever out of the Commerce Department. 

Add to all this that Treasury Secretary Snow is scheduled as a morning guest on CNBC — he tends not to do that when the data stinks the joint up — you have the makings of an upside surprise.

I consider myself a "Reality Based Strategist" (RBS). So I will stick with "The Under" — on principle alone . . .

Print Friendly, PDF & Email

What's been said:

Discussions found on the web:
  1. Steve commented on Nov 4

    “Productivity is a function of Hours worked and output. We had slightly less output, but a huge decrease in hours worked, thanks to the job losses in the gulf. ”

    Hmmm, does this mean that Gulf workers are far less productive that the rest of the economy. Freeloaders!, or maybe its the drive through Cocktail shacks.

  2. Barry Ritholtz commented on Nov 4

    Its not reality, but BLS data that made productivity look so high.

    The output was measured;

    The hours worked were a wild (and excessive) guess.

    Viola! Increased productivity

  3. Tom Bozzo commented on Nov 4

    +56,000… the ‘under’ remains a good bet.

  4. Emmanuel commented on Nov 4

    Hmm…the dollar is way stronger now than in the past two years when the current account deficit was considerably smaller.

    The stock market seems to be headed up despite the two major hurricanes that were supposed to wipe out so much GDP growth, which advanced at a rapid pace despite everything (at least with error-prone estimates).

    Maybe we shouldn’t think too much and just accept conventional market wisdom and Bush administration rhetoric. Maybe this is the best economy of our lifetimes. Deficits (and dissavings) don’t matter; just relax, spend your little hearts out like the little buying automatons we’ve been conditioned to be.

    Everything is alright. Why are Bush’s approval ratings so low, particularly on the economy? Let’s all be Jolly Bulls instead of Angry Bears. To paraphrase Dr. Strangelove, maybe it’s time we stopped worrying and learned to love Neo-Lafferite Voodoo Economics.

    It’s hard to stick with our convictions when presented with so much contrary evidence (no matter how distorted it may be)!

  5. Steven Engelhardt commented on Nov 4

    Productivity is a function of hours worked and output. We had slightly less output, but a huge decrease in hours worked, thanks to the job losses in the gulf.

    For reference, let me post the relevant section from the NY Times article “Productivity Growth Nearly Doubled in the Third Quarter“:

    Output jumped 4.2 percent, and hours worked edged up 0.1 percent.

    Your claim seems to be that the BLS is overestimating the number of hours worked because it is incorrectly counting unemployed, displayed NOLA Katrina victims as working. But how did you arrive at the “slightly less output” statement?

  6. Barry Ritholtz commented on Nov 4

    Productivity, which is calculated by dividing output by hours worked, rose at an annual pace of 4.1 percent from July through September, up from 2.1 percent in the second quarter

    4.1% blew away forecasts.

    In the last 2 months of Q3 (Aug/Sept), large parts of very vital areas of the US were blown off the face of the map.

    Do you think that the number of hours worked could have been pressured lower by the Katrina/Rita storms?

    The BLS had this to say:

    …”The effects of Hurricanes Katrina and Rita, which struck the Gulf Coast during the third quarter, could not be separately identified in the productivity and costs measures.” …

    http://www.bls.gov/news.release/prod2.nr0.htm

  7. cm commented on Nov 5

    I’m increasingly suspicious of the BLS numbers (and, for that matter, most macro numbers like GDP and CPI). Mind you, I’m not suggesting outright fudging (on the part of BLS). But from a methodology point of view, I think the BLS “algorithms” and modeling work only under certain “steady-state” assumptions, and shocks like Katrina, that on top of it affect the reporting infrastructure, can throw the numbers wildly off track, even without disingenuity.

    Who believes the number of NOLA unemployed is so low? Were the occupants of the superdome, and all other hurricane exiles, all welfare cases? Come on.

    Maybe a number of people are technically still on some payroll. Have they been paid? I think that is one question in the (hypothetical) CPS survey.

Posted Under