----------------- HES POSTING ----------------- 
I presume that Sam Bostaph means to hold Clive Granger up as object lesson in how
mainstream economics goes wrong.  If so, it would help to get the facts right.  Granger's
Nobel citation was for his work on cointegration and not for his work on Granger-
causality, the only reasonable referent for the post-hoc-ergo-propter-hoc remark.
 
Granger-causality is frequently misused in the way that Sam Bostaph suggests.  I was once
ready to condemn it wholesale for that reason. However, Granger himself understands
perfectly that a finding of Granger-causality does not warrant causality in its workaday
uses. Careful students of econometrics heed Granger's own warnings that the notion deals
with incremental predictability.  What is more, Granger-causality is a specie of the genus
probabilistic causality, which has a rich philosophical tradition behind it.  While I have
argued that probabilistic causality -- in virtually all of its forms -- is inadequate as a
general account of causality, it does carry important lessons that any serious account of
causality must accommodate.  For more on all these points, see my book:  Causality in
Macroeconomics, Cambridge, 2001.
 
Cointegration is a really important idea that has radically altered macroeconometrics.
The prize is well deserved. But not to end on a sour note, I would favor an Ignobel Prize
for the principle of composition  -- except that it would be hard to narrow the field down
to three worthy recipients.
 
Kevin Hoover 
 
------------ FOOTER TO HES POSTING ------------ 
For information, send the message "info HES" to [log in to unmask]