Friday, August 31, 2012

On Yglesias on Money agan

Yglesias has a post here.  He has a much more provocative tweet here

Michael Woodford knows how to create jobs. Does Ben Bernanke? 

Both refer to Woodford's Jackson Hole speech 

I have read only the concluding section of Woodford's paper.  In it he expresses no firm opinion about whether a nominal GDP target would have a significant (or even detectable) effect on GDP.  There is a lot about how QE without promises about future policy makes no sense, but he presents no evidence that statements about policy in the fairly distant future have any effect on anything.  The reason is simple, no central bank has attempted to cause an increase in expected inflation.

So far I am attempting to interpret Woodford.  I don't agree with everything which he implies.  He suggests that we have essentially no information about the effects of medium and long term policy guidance from central banks.  In fact we have a lot of information (see key weasel "an increase" in the paragraph above).  In the late 70s and early 80s monetary authorities tried very hard to convince people that they were absolutely determined to achieve low inflation.  If the expectations channel works even when other channels don't this would have caused low expected inflation even without direct effects through gigantic safe short term interest rates, severe recessions and record setting unemployment.  The evidence is overwhelming that monetary authority statements about their medium term targets and goals have never worked without policies which bypass the expectations channel.

As far as I know, neither you nor Woodford has explained why this historical evidence is not relevant and doesn't amount essentially to a refutation of our confident claims and his diffident claims that we don't have relevant data.

I see no justification for your tweet that "Woodford knows ..." I think this is a problem with epistemology.  What do you mean by "knows" ?  Doesn't it have something to do with proof (in mathematics) and evidence (in everything else).

No comments: