Tuesday, December 22, 2009

Statistical Limitations

As we approach the end of another calendar year for those following the Western calendar (those following the Eastern Orthodox calendar get to celebrate Christmas in a few weeks) we will start to see the "year end" statistics come out on all sorts of things.
 
Statistics are useful, both for what they reveal and for what they conceal.
 
Methodologies matter.  So do footnotes, if there are any.   I urge everyone to delve into the real statistics and look at the original sources, to get at the methodologies and find, read and understand the exceptions, generalizations and qualifications which put the few alert readers and scholars out there on notice that, well, the stats may end up being "way off" and there are plenty of reasons why.
 
The stats should be particularly interesting when looking at lagging economic indicators.   While the nominal unemployment rate (what is called "U-3") appears to be levelling off (perhaps wishful thinking? or political manipulation for the coming 2010 elections?), the wider "U-6" measures and the "underemployment" measures both have "more real" unemployment measures approaching or even well over 20 percent of the workforce.   Also consider that the stats finally show that the average duration of unemployment has increased during the recession by something like nine weeks; this indicates that not only has unemployment increased significantly, but that those suffering unemployment are being hurt far, far worse than anything in decades.
 
An analogy:  The number of sick in the current period has doubled when compared to the earlier "base" period.   But in the current period, the sick have pneumonia, versus the common cold in the earlier period.  
 
The severity and depth of this recession will have, in my opinion, major implications for the labor markets, capital markets, banking system and real estate markets -- and our wealth -- which have yet to unfold.
 

0 comments:

Post a Comment