Too bad Reuters didn't do a little research on the BLS web site. The method for calculating unemployment and the questions that are asked could have enabled the author to write something worth reading. Pretty easy for the average person to know the excuse is BS, or shall I say it is easy for the average FReeper to know.
To generate these numbers, they don't actually count paper claims filed. They call a bunch of random people on the phone and ask, "Hey, did you work last week?"
Given the way the question is typically phrased, some nimrods out there who ARE employed but DIDN'T go to work could skew the numbers.
This explanation was written about in several publications (and discussed here on FR) well before the numbers came out.