Much has been said about the January Non-farm payrolls number, which
rose by 113K on expectations of a 180K increase, most of which has been
focused on the US atmospheric conditions during the winter. There is a
problem with those numbers: they don't really exist (as for the
non-impact of "the weather" on jobs we showed previously that the number of people "not at work due to weather"
as calculated by the BLS itself. this winter was lower than 2008, 2009,
2010, 2011 and 2012 - so much for historic winter weather).
So what really happened in January?
For the real answer we have to go to the BLS' non-seasonally adjusted data series. It is here that we find that in January, some 2.870 million real, actual jobs were lost, not
gained. Putting this further in perspective, the number of NSA jobs
losses in January 2014 was greater than in January of 2013, 2012, 2011
and tied that of 2010. In fact only during the peak of the depression in
January 2009 was there a greater NSA drop in the first month of the
year when 3.691 million jobs were lost.
So how does a loss of 2.9 million jobs become a "gain" of 113K jobs
in the same month? Simple: through the magic of seasonal adjustments.
Here's the rest of the article.