Despite the surprisingly strong productivity figures for Q1 and Q2, however, the data also revealed the feared downward revisions in each of the prior three years that limited the scope of the presumed productivity acceleration. As was the case even prior to the report, however, the revisions were too small to fundamentally alter the premise that productivity growth rose substantially in the second half of the 1990s. Rapid growth in capital investment still occurred, and trend growth in the U.S. economy clearly accelerated relative to labor force growth to leave surprising strength in productivity. Whether we are in a "new economy" has always been suspect, but these underlying observations remain largely intact.
The BLS output measure used to calculate productivity has always closely tracked the "Real Nonfarm Business Less Housing Product" series in the GDP report. Revisions in growth here explain the market's prior knowledge that productivity growth would indeed be revised lower in the 1998-2000 period. Downward revisions revealed at the end of the July with the annual GDP revisions implied small downward revisions in the BLS data for 1998 and 1999 of 0.2% (versus 0.1% each for overall GDP), and then a sharp downward revision in the Q1 figure for 2000 of 3.2% that allowed a sizable 1.2% downward revision for the 2000 data overall. The quarterly gains for Q4 of 2000 and Q1 of 2001 were poised to be revised higher by 1.0% and 0.4%, respectively, given the GDP report.
One noteworthy revision in the late-July GDP report was the removal of the upside surprise to GDP growth for 2000 via an elimination of the upside Y2K surprise in Q1 of last year. The data now fully reveal the expected sharp inventory-led slowdown in Q1 following the Q4 output and inventory surge, and take some of the shine from the view that the U.S. is experiencing a "New Economy." Productivity previously posted a mind-boggling 4.3% surge in 2000 that was unexpected at the time. Now this surge is "only" 3.0%.
The results of recent research will not likely be substantially altered by the new figures, though projections of current trend-growth in GDP may be nudged lower. For instance, a recent NBER research paper by Baily and Lawrence finds "that the change in total factor productivity between 1995 and 2000 was largely structural and not cyclical," as labor productivity growth accelerated from a prior average of 1.4%. The acceleration may be less significant now, but the new productivity data still show annual gains of 2.1%-3.0% in each of the last five years. Since a large portion of the quarterly gyrations in productivity is cyclical rather than structural, recent weakness in productivity growth says little about whether we are or are not in a new economy. We will need to start a new cycle before we can evaluate the last one.
As we have also frequently noted, however, economists are loath to claim that we are in a "new paradigm," and the jury remains out on this question. On this front, note a NY Fed research paper last year by Rich and Rissmiller that fully explains the recent shortfall in U.S. inflation with surprising strength in the dollar and weak import prices. The new productivity figures will make it even more true that the current cycle may eventually be explained with models that look notably conventional. Englund is Chief Market Economist for Standard & Poor's