>Lastly, is the household survey figure for employment seasonally adjusted?
>I always assumed it was, but it's so different I was wondering.
Yes it is. (The widely reported one is, that is. You can get the unadjusted numbers too if you like.) The differences between the household and establishment surveys have been unusually wide, with the HH showing much greater employment growth - causing right-wingers to say that it's the "real" story. But the HH survey uses a much smaller sample (60,000, vs. 300,000 for the establishment). The BLS always says the establishment survey is more reliable. Every year, the monthly estimates are benchmarked against unemployment insurance records, which cover nearly 100% of the (legal) workforce. With next month's release, the BLS will revise its estimate of March 2003 employment down by 145,000, or 0.1%, based on the benchmarking technique. (They made that general announcement several months ago, but the real numbers will be released on Feb 6; for more, see <http://www.bls.gov/web/cesspec.htm>.) That should have put to rest claims by Larry Kudlow & the WSJ edit page that the HH survey was telling the real truth, but it only quieted them for a few weeks; Kudlow was making that argument again on CNBC about 8:25 last Friday morning. He looked kind of silly five minutes later, but the man has no shame.
Doug