----- Original Message ----- From: "Robert Seeberger" <[EMAIL PROTECTED]> To: "Killer Bs Discussion" <[EMAIL PROTECTED]> Sent: Sunday, January 04, 2004 7:24 PM Subject: Re: The New Math
> > I got it from > > > > "The Labor Department (news - web sites) reported Wednesday that new > > applications filed for unemployment insurance dropped by a > seasonally > > adjusted 15,000 to 339,000 for the week ending Dec. 27." > > > > Which was in his first post on the subject. I also saw the report > on the > > seasonal adjusted unemployment elsewhere. > > Well....I read that in the original post also, but I don't see how it > is relevant. > Without more information, "seasonally adjusted" could well be a > euphemism for "sanitized for your misdirection". <G> IOW, the phrase > by itself doesn't tell me what is being measured. But, it does mean something to me. Years ago, I read about how seasonal adjustments are made to data like these. It made sense then, and does now. Further, its worth noting that this is an accepted technique used by a wide group of ecconomists, of various political persuasions. > 200k people are absent in that measurement as compared to the general > stat and I am wondering what their status is that they don't count as > unemployed. Its not that. Its that regular seasonal adjustments must be discounted if one is to make sense of the trend. Think of the total unemployment as data = (long term trend + seasonal adjustment). One can fit a number of years worth of data to this, and determine each term. In order to obtain long term trend, one needs to subtract seasonal adjustment from data. > Please excuse my skepticism, but undefined statistics irritate me > because they leave one guessing at the meaning of the information they > are supposed to convey. But, it only takes a quick web search to understand what it is. Here is a decent explaination of what is done. http://www.dallasfed.org/research/houston/1997/hb9703.html > MY guess is that the "seasonally adjusted" numbers exclude the people > who are almost always unemployed at this time of year and/or those who > *are* employed at a time when they normally wouldn't be. > But that is purely a guess and I know this. Seasonal adjustments go both ways. It normalizes the year to take out the natural fluctuations that have no relevance to the underlying economy. For example, there are plants that shut down over the Christmas holidays. Those workers can file for unemployment for one or two weeks. That would give a spike in the unemployment figures for that week. This spike does not have the relevance to total unemployment that people being laid off permanently, because they will be back on the job shortly. Another example of this is the increase in employment in the summer and the decrease in the winter. Both need to be factored out to obtain a long term trend. Dan M. _______________________________________________ http://www.mccmedia.com/mailman/listinfo/brin-l
