Isn’t that what they do each month anyway?
I’m just saying — between hedonics, seasonal adjustments, and the now notorious Birth/Death model, the establishment survey is of increasingly less utility for reality based investors. Now, we will get an ex-Katrina reading.
Anyway, here’s an excerpt from Dow Jones, regarding tomorrow’s hard to guess Non Farm Payrolls:
NEW YORK (Dow Jones)–Hurricane Katrina was the most costly natural disaster to hit the U.S. economy in terms of property destruction and disruption of economic activity.
In this regard, it’s not surprising that the Labor Department will make an unprecedented change in the way it compiles the jobs data for September that will show a negative effect, but may increase uncertainty about the economy’s true health.
The closely-watched payroll employment data are estimated each month from a sample of some 160,000 businesses and government agencies covering roughly 400,000 worksites. But the widespread disruption to normal activity – and most importantly the evacuation of New Orleans – raises the prospect that collection of this data for September will be faced with huge problems.
The Labor Department has explained how it intends to handle the data gathering in the affected area and has tried to indicate how these policies could affect the September jobs report.
As reported by Dow Jones Newswires on Friday, the Labor Department’s Bureau of Labor Statistics (BLS) will make special estimates for jobs in the Hurricane Katrina-affected region, which may result in understating employment. (ED: sure it will).
Where possible, the BLS will use employment data for an affected worksite from the headquarters outside the area if the local contact can’t be reached. But it pointed out in a special notice that "if there are sample units that BLS is unable to contact in the most heavily impacted disaster areas (Labor) will assume that the business is not operating and therefore has an employment level of zero. This carries some risk of overstating employment loss."
This is different from how it typically handles non-reporters, which ironically is often a special problem in September. Usually, if the payroll sample for a particular industry has 10,000 reporting firms in it but in a particular month the number falls to 9,900, BLS calculates the percentage change for the 9,900 that reported in both months and then extrapolates that change for the full 10,000 sample.
This is typically a problem in September because of late-reporting school districts and has historically resulted in a tendency for the month’s employment totals to be revised up.
The likely geographic concentration of non-reporting firms this September makes the usual adjustment for late-reporting firms inappropriate. Some jobs in the most affected areas certainly were lost.
In fact, there are eight counties in Alabama, 31 parishes in Louisiana and 47 counties in Mississippi the the Federal Emergency Management Agency designated as "most affected." There are 145,341 worksites, or 1.7% of the national total, located in these areas accounting for 2,424,194 payroll jobs, or 1.9% of the national total.
Expectations Of Large Declines
An extremely partial survey of nine economists who have made preliminary forecasts yields a median estimate of a decline of 168,000 in September payrolls in contrast to the increase of 187,000 averaged over the 12 months through August.
One economist who has made a forecast of 220,000 job losses for September is Dave Greenlaw of Morgan Stanley in New York, who explained how he arrived at that estimate. "We are guessing that about 400,000 individuals lost their jobs and did not receive a paycheck during the September survey period," he said. The firm’s expectation for payrolls growth ex-Katrina is 180,000 – about in line with the trend over the past year and a half. Greenlaw recognizes that any people who did remain on the payrolls – even if they did no work – would be counted among the employed.
To review: BLS says they don’t know how they are going to count people who aren’t in Louisiana; We know 400,000 people no longer work in the City of New Orleans, and we know that the BLS will make "special estimates" to adjust for Katrina.
We might as well throw out tomorrow’s number right now, as its going to be mostly worthless . . .
Katrina Will Cause Labor Dept To Guess Jobs
By John McAuley
Dow Jones Newswires, 09/26 3:42P