Unemployment number decline is all about seasonal adjustments
A lot of people are questioning the unemployment rate of 9.7% in the face of a –20,000 non-farm payroll print. How could we be losing jobs and have the unemployment rate drop? It would seem people are dropping out of the labor force.
However, I have now parsed the household survey data and most of the data seems reasonable. The labor force participation rate actually ticked up slightly (both seasonally adjusted and unadjusted) – as did the number of people not in the labor force who wanted a job (unadjusted only). This is what we would expect.
What sticks out is the seasonal adjustment for the number of persons employed and unemployed.
In December 2009, there were 15.267 million people unemployed on a seasonally-adjusted basis. This ticked down to 14.837 in January 2010, a fairly large drop of 430,000. Meanwhile the unadjusted numbers go the other direction – massively. In December 2009, the number of unemployed persons was 14.740 million. This rose 1.4 million to 16.147 million. Therefore, we saw a swing of over 1.8 million between what the unadjusted and the seasonally adjusted data are saying about who’s unemployed. The number of people employed increased by over 500,000 on a seasonally-adjusted basis, while it decreased by over 1.1 million on an unadjusted basis. That’s a swing of 1.6 million.
Bottom line: the unemployment rate downtick has nothing to do with people dropping out of the workforce; it is an statistical aberration due entirely to seasonal adjustments in the household survey in the number of people employed and unemployed.
If you dig a little deeper into the data, the discrepancy is not as wide as it looks. January is a huge month as far as seaonal adjustments are concerned. Firms let go many workers this time of year so the difference between the SA and the NSA number will be large. Look at last year from Dec. to Jan and compare with this year. What you see is that the NSA jump was about 1.5M more than the SA jump for that time period. This year it was a little wider at 1.8M. True the SA was down and the NSA was up which is different from last year. But to trump the change between the two distorts the picture. You should compare last year’s net change between the two series to this year’s net change. That gives you a truer picture of what is going on.
bondscoop. Looking at last year only confirms that seasonality is a big part of this as the data tables above indicate. You will have noticed how both SA and NSA data last year were showing unemployment rising dramatically. So, that’s not the right analysis.
A better analysis is to look at average historical Dec-Jan seasonal adjustments – something time constraints don’t permit me to do. But, having eye-balled this I can tell you these numbers a re outliers. And I await a trend confirmation with February and March data.
A reader of Mish’s ran the numbers on household survey seasonal adjustments. It shows this last month is an outlier as I indicated. Here they are:
https://globaleconomicanalysis.blogspot.com/2010/02/bls-seasonal-adjustments-gone-haywire.html
I tend to agree with Atrios that this report is bad news for the economy because it will perpetuate the illusion that all is well and no more extraordinary measures are needed to get the economy back up to speed….
I think the context for these seasonal layoffs has to be considered when evaluating their overall importance. It probably doesn’t make nearly as much difference during an expansionary period when unemployment is low and consumer credit is flowing that a lot of people lose their (temporary, in many cases) jobs in January. But in the present context of proto-depression, with high unemployment, declining consumer credit, etc., it could matter a great deal that suddenly a significant number of people who had been in the workforce the previous couple of months now no longer have a paycheck and therefore can’t buy as many things. In this respect, the BLS casually applying basically the same seasonal adjustment formula used for expansionary periods creates a misleadingly positive impression about the health of aggregate demand.