>Dan Sichel says (and I agree) that Gordon's cyclical adjustment
>factors are too large, and are masking an acceleration of trend
>productivity growth outside of durable manufacturing.
But Sichel & OIiner make no cyclical adjustment at all, and it's well-established that productivity is pro-cyclical. If Gordon over-adjusts, he's doing too much of the right thing; not adjusting at all is the wrong thing.
Don't want to give away the contents of A New Economy?, but the Fed's HITEK2 index - part of the industrial production series that comprises computers, communications equipment, and semiconductors - is rising at a 50% annual rate. The X4HTK2 index - all manufacturing except HITEK2 - is rising at a 2% rate. Since employment in both sectors is pretty flat, these are pretty close approximations of the labor productivity in both. HITEK2 is about 10% of total industrial production, but it's driving the results for the rest of manufacturing.
Does it really mean that much for human welfare that real computer output is rising at a 50% annual rate? Does it make any intuitive sense at all? Is it right that a 10-15% increase in the nominal value of high-tech output is inflated, thanks to the wacky price indexes, into a 50% real increase?