While some economists who will remain nameless are constantly crowing about the big productivity numbers the US economy has been putting up recently, an interesting op-ed in today's Financial Times puts a valuable new spin on the data.
But starting in 1995 something extraordinary happened. America's productivity growth speeded up while Europe's slowed. A host of research studies attributed most, if not all, of the US's surge to its dominance in making computers and developing software. Intel and Microsoft are familiar names that symbolise US dominance in the information and communications technology (ICT) sector. . . .Aha!
Two weeks ago the productivity puzzle suddenly deepened when the US government revised upwards its productivity numbers for the past two years and provided its first release for the second quarter of 2003, which came in at an unbelievable 5.7 per cent. The underlying long-run trend of productivity growth is currently running at about 2.8 per cent a year, fully double the pre-1995 growth rate.
So we face a new paradox. Those research studies that attributed the 1995-2000 revival of US productivity growth to the ICT investment boom of the late 1990s have some explaining to do. After 2000, the ICT investment boom collapsed along with the stock market but productivity growth accelerated. If ICT growth collapsed but productivity growth increased ever faster, something else besides ICT investment must have been behind the American miracle.
The General crunched some numbers and the results are quite fascinating. The below chart shows quarterly growth rates in private non-residential fixed investment (blue line) and worker productivity (red line) since 1992. The growth rates are seven quarter centered moving averages, and the productivity line is lagged nine quarters.
[If you can't see the image, clicking here should work]
Throughout the roaring '90s, investment and productivity gains are shadow one another in an amazingly tight way. The nine quarter lag sort of surprises me, but there the data is.
As we approach 2000, investment growth starts to fade, and by 2000:IV (remember, these are 7-quarter centered moving averages) investment begins to actually fall. What is stunning, of course, is how productivity begins to take off like a rocket just as investment begins to wane. These post-2000 gains -- that is, the really big productivity gains -- are not built on technology.
What then are they built on? The FT piece today suggests it's
"intangible" productivity-yielding activities. These comprise such things as reorganising and reinventing business practices, and both formal and informal training of computer and software users. . . . especially [in] retail trade, which just happens to be where the US's productivity showing is strongest. America's retail productivity performance has all been achieved in stores newly built since 1990, not in existing stores.Yes, ladies and gentlemen, "modern retailing" is the answer.
The new stores are the "big boxes" such as Wal-Mart, Home Depot and Best Buy, large new buildings set up on greenfield sites at interstate highway junctions, in suburbs and, increasingly, in inner cities. As these new stores reap the rewards of their size, openness and accessibility and drive smaller stores out of business, they bolster the average productivity of the US retail sector as a whole.
If true, this analysis fits nicely into the exploitation argument the General has been pushing all along. After all, Wal-Mart is synonymous with "exploitation," both of its workers and its suppliers. "Modern retailing" gives corporations big profits and gives consumers cheap goods. It also supplies low wages and low benefits, battles union activity, acts as a huge accelerator on the US current account deficit, destroys smaller businesses (which notably funnel much more of their revenue into local communities), undermines democracy and aggravates deflation.
"How beauteous mankind is! O brave new world, that has such people in it!"