For all those who continue to doubt the statistically quetionable methods of our Labor Department, as well as for all others who mock those who doubt the veracity out of anything coming out of the BLS, the following chart should provide much needed closure. The top section of the chart below demonstrates weekly prior revisions in initial claims for all of 2010. Readers may be surprised to discover that beginning in April, of 2010, continuing through today, there have been 22 out of 23 consecutive upward prior weekly revisions! In other words, the BLS has a definitive mandate to underrepresent the "current" weekly data and to allow it to catch up with reality once it has become "prior", and thus no longer market moving, when in reality should the BLS present true data it would have likely missed estimates on more than half the occasions it has "beaten" and caused ridiculous market spikes like the one experienced earlier. Furthermore, combining all individual weekly data, demonstrates that the BLS has underrepresented initial claims by roughly 80,000 year to date. The chart pretty much leaves no room for doubt as to the BLS "trans-statistical" approach to quantifying data. As for the lower chart, it shows the same thing but with continuing claims, which have been revised upward pretty much consecutively for the entire year.
And there you have it - magical BLS statistics in action.
h/t John Lohman