Barkley Rosser's point is dead on, and very important.  There is 
indeed no other explanation for the extraordinary plunge in 
unemployment in Germany between 1933 and 1936 (not 1939, as Barkley 
suggests) than the large scale public spending of the German 
government on roads, infrastructure and similar capital projects 
(largely related only indirectly  to military objectives, as in the 
case of the Autobahnen), financed by public debt through the device 
of Mefo bills, undertaken from the outset of the Nazi regime.  This 
was precisely what Keynes's letter to the New York Times prescribed 
on December 31, 1933, and was widely recognized as such at the time, 
as Friedrich Baerwald's admiring account in the AER in 1934 makes 
clear.  Despite some divergences from what the General Theory would 
later advocate (higher taxes, enforced saving and, of course, 
increasing dirigisme over the entire period), there's no serious 
doubt that Hitler's policy in these crucial years represented the 
first serious experiment in "Keynesian economics," at least as 
Keynes's himself understood it at the time, and that this is true 
whatever Hitler's ultimate intentions in reviving the German economy 
might have been, and whether Hitler knew anything at all about 
Keynes's economics, or anyone else's.

Modern historians like Rochard J. Overy, concerned that any 
association of Keynes's name and Hitler's could only sully the 
reputation of a great liberal thinker, seize on the technical 
deviations of Hitler's policies and Keynes's prescriptions, obviously 
not significant enough to impede the rapid growth in GDP and drop in 
unemployment brought on by the deficit spending, to insist that 
Hitler was no "Keynesian."  It is certainly true that Keynes's was 
not a fascist, but it is less clear to what degree the aggregation 
intrinsic to his theory makes it susceptible to or compatible with 
collectivist, even "totalitarian" political theories and regimes, and 
for this reason it is worth discussing the significance of Hitler's 
experiment and its relation to Keynes's ideas in their broadest 
contours openly and in some depth.  Certainly Keynes himself thought 
so, as his curious Preface to the German Edition, written in 
September 1936, makes clear -- there, three years after the fact, he 
notes that "totalitarian" regimes can indeed make use of his ideas 
(indeed, he says, these regimes can't have much use for free markets 
and laissez faire, so his is a theory they can use) and cordially 
invites the theory-starved Germans to try it out.

Rich Adelstein