In late October 1999, the Bureau of Economic Analysis (BEA) released a portion of its 1999 Comprehensive Revision of the NIPA that included two improvements that significantly revised the historical path of productivity. First, the BEA revised the pre-1995 NIPA data to include the effects of the new BLS geometric weighting formula. The BLS introduced this improvement to the CPI in January 1999 and estimated that it would lower the future annual growth rate in the CPI by about 0.2 percentage point. The BEA estimates this change to have a 50.0 percent “feed-through” effect on the aggregate annual GDP price index growth rate. Thus, in its NIPA revisions, the BEA lowered the annual historical growth rate in the aggregate GDP price index by about 0.1 percentage point (0.2 * 50%). And, since this improvement does not alter the historical path of nominal GDP, the annual historical growth rate in the real GDP, and therefore productivity, were increased by about 0.1 percentage point. Second, the revised NIPA data also shifted private business and government investment expenditures on computer software from an intermediate to a final good. This improvement increased both nominal and real GDP growth, and therefore productivity, particularly over the last twenty years. However, this revision also lowered the GDP Chain-Type Price Index growth rate relative to the CPI growth rate.
NIPA stands for National Income and Product Accounts.
Is this saying that the historical GDP growth has been adjusted for the new CPI? And that the effect is to make the historical GDP growth higher?
And what is the meaning of "GDP Chain-Type Price Index growth rate?" Is this trying to say that expected price growth is lowered by the CPI change?
And what validity is there in counting software as a "final good"? Indeed, what does "intermediate good" mean?
Bemusedly yours,
-- John K. Taber