The `Revolution' In U.S. Productivity May Be Overstated

It's a popular thesis--the idea that long-term U.S. productivity growth is finally on a roll and will cruise through the 1990s at the 2% to 3% annual clip it hit in the 1950s and 1960s. But economist David A. Wyss of DRI/McGraw-Hill sees "no reason to rejoice over a revolution in productivity."

Wyss agrees that the record pace of real equipment investment in the current upswing, along with continued downsizing and work restructuring, has helped foster productivity gains. He also concedes that rising output per hour has accounted for far more of the economy's growth than in similar expansions.

Wyss notes, however, that productivity growth itself in the current recovery appears close to normal (chart). And though the government says real business equipment spending has soared at a near-record 44.3% clip since early 1991, it reports that actual outlays in current dollars are up only 34%--the lowest amount for any recovery since 1960.

The big reason for this, of course, is that quality improvements attributed to computers by the government's numbers grinders have caused their prices to fall sharply. That means less cash buys more real goods, pushing up equipment's contribution to economic growth.

Wyss questions such quality improvements, which are based on computers' raw processing powers. As the government measures it, he says "my IBM ThinkPad is `bigger' than the $3 million IBM 7094 mainframe on which I first learned to program." But that 7094, he notes, was used exclusively to crunch numbers and ran 24 hours a day.

By contrast, Wyss's PC is usually zipped up in a travel case and is most often crunching words, not numbers, when in use. "To a large degree," he says, "my PC has replaced the $129 typewriter I used in college. Valuing it as the equivalent of the mainframe vastly overstates its impact on productivity."

The picture is even less inspiring if you factor in the shorter lives of computers. "Because equipment is being scrapped a lot faster, you need more investment just to keep the capital stock steady," says Wyss. "Net investment as a share of gross domestic product has dropped sharply since the 1960s."

The upshot, say Wyss, is that productivity growth is likely to slow as the expansion matures, as it has in prior cycles. And while he expects it to average about 1.6% in the 1990s, compared with 1.1% in the 1980s, that's still far below its blazing pace in the 1960s.

    Before it's here, it's on the Bloomberg Terminal.