Sometimes I wonder whether the Department of Labor (DOL) has a clue about how to do statistical analysis! Today's DOL press release about the seasonally adjusted number of first time unemployment claims make no sense!
The DOL press release states that seasonally adjusted first time unemployment claims dropped from an unexpectedly high amount of 404,000 in the previous week to an unexpectedly low amount of 381,000 this week.
These "seasonally adjusted" numbers make no sense when compared to the actual non-seasonally adjusted data from the last 25 years. In fact, the previous week's non-seasonally adjusted first time unemployment claims had actually dropped to the lowest rate of the entire year! Yet the DOL announced a big worsening!
This week's "seasonally adjusted" numbers, as reported by the DOL were reported as a big drop in first time unemployment claims when, in actuality, there was a relatively big jump in the actual, non-seasonally adjusted claims, when compared to the last 25 years! What makes this even more ridiculous is the fact that this week's DOL data will likely be further adjusted upwards in next week's DOL press release. Therefore, this actual big jump in first time unemployment claims will likely be an even bigger resultant jump upwards when I look at next week's DOL press release!
In other words, no one should give any creedence whatsoever to the weekly seasonally adjuisted numbers in the DOL press release. Those bogus numbers are quoted on every media news report on television, on radio and in the newspapers. As a result of the DOL's invalid methodology in regards to the seasonally adjusted data, I can only recommend that the non-seasonally adjusted data, as compared to the last 25 years be viewed as valid.
The stupidities and absurdities by which mathematicians have rather excused than explained their mode of procedure, which remarkably enough always lead to correct results, exceed the worst and real fantasies of the Hegelian philosophy of nature.