"The length of the workweek declined by 0.1 hour to 33.1 hours, which is the aggregate hour equivalent of an additional loss of about 350k jobs. More importantly, the manufacturing workweek also declined (39.3 hrs vs. 39.5 previously)?"

Wouldn't we have to know if similar spikes down were measured similarly in the past and whether or not the length of the workweek declined in the same report which featured the spike down before we know what to make of the above?

I'm curious to see what Gartman says about this move, as he has been waiting and talking about this event for several months now. My feeling is, although Dennis is far from right most of the time, he's careful enough, especially on a statistic like this which is his cornerstone in timing the end of the recession, that he would know the impact of the decline in the work week.