A little silliness here yesterday: I praised the accuracy of the potential output data series, real and nominal, based on similarity to the inflation series. That's silly, because they have to be similar. After CBO has "real potential GDP" figured, to get the nominals they can just factor-in the inflation numbers. Those numbers are known. So when I come along later and subtract the real rate from the nominal, I'm left with inflation numbers that differ from the originals only by rounding errors.
Then, for the ten-year futurecast, they just draw a nice, flat line about as close to the "target" rate as you could get if you drew it by hand.
So while I was having some fun with poutput and inflation, Paul Krugman was having some fun with unemployment:
Estimates of how low U can go seem always to be a bit below the current level of unemployment.
Well, Krugman squeezed all the fun out of it by the end, there. But he's right: If there is no inflation, then the NAIRU must be less than the unemployment. So then, they guess a lower NAIRU number.
Point of interest (maybe): FRED identifies the U.S. Congressional Budget Office as the source for the NAIRU numbers. Not the FOMC.