I’ve been reading Robert Hetzel’s excellent new history of the Fed, and came across this November 1992 quotation from Alan Greenspan:

We have seen that to drive nominal GDP, let’s assume at  4 1/2%, in our old philosophy we would have said that [requires] 4 1/2% growth in M2. . . .  I’m basically arguing that we are really . . . using a nominal GDP goal of which the money supply relationships are technical mechanisms to achieve that. [p. 426]

Later on, Hetzel discusses Greenspan’s views on the appropriate inflation goal:

Greenspan wanted price stability subject to the proviso that at low rates of inflation productivity growth would be high enough to restrain growth in labor unit costs to be consistent with moderate wage growth. He feared that with price stability wage compression coming from an inability to cut nominal wages would increase unemployment. [p. 435]

This led to a procedure that is essentially equivalent to targeting NGDP per worker:

However, for Greenspan, at low rates of inflation, the amount of inflation that the FOMC would tolerate would vary inversely over time with the rate of productivity growth.  [p. 437]

This is quite close to George Selgin’s “productivity norm”.

PS.  As of the fourth quarter of 1992 (when Greenspan made this statement), NGDP growth had averaged 8.29% over the previous 30 years.  How likely does it seem that NGDP growth would average 4.5% over the subsequent 30 years?  Not likely, unless you’re one of those people (like me) who believes that it is the Fed (not Congress) that determines the trend rate of NGDP growth.  

In fact, NGDP growth averaged 4.65% in the 30-year period after Greenspan’s remarks.  And Greenspan helped to make it happen.