Print this article | Return to Article | Return to CFO.com
America's productivity growth has slowed. Does that matter?
Economist Staff, The Economist
April 16, 2007
The good news about America's economy is that jobs are plentiful despite slower growth and the housing blues. Some 180,000 new jobs were created in March and the unemployment rate fell to 4.4 percent, three-tenths of a percentage point lower than a year ago. With employment and wage growth strong, consumers are unlikely to stop spending and throw the economy into recession.
That is not all cause for celebration, however. The drop in the jobless rate at the same time as the economy is slowing implies that the growth in productivity — the amount workers produce in an hour — is waning. If this proves to be a permanent shift, slower productivity growth bodes ill for inflation and living standards.
Few associate America with limping productivity. Central to its success over the past decade has been its "productivity miracle", the sudden acceleration in workers' efficiency in 1995. After advancing at a measly 1.5 percent per year for more than two decades, productivity growth soared to an average of 2.5 percent a year in the late 1990s and over 3 percent a year between 2002 and 2004.
This spurt set America apart from other rich countries. But between mid-2004 and the end of 2006, the growth in business output per hour outside agriculture, the most common gauge of worker efficiency, slowed to an annual rate of just 1.5 percent, on average. Judging by the recent jobs figures, its growth in the first few months of 2007 may be lower still.
Deciding how worrying this is depends on what lies behind the sluggishness. Productivity growth has two components: a long-term trend (set by the quality of the workforce, the pace of capital investment and the speed of innovation) and more volatile short-term fluctuations driven by the business cycle. Early in an expansion, for instance, productivity takes off temporarily as firms squeeze their existing staff harder before hiring new workers. As an economy slows, it tails off, because firms are loath to sack workers immediately.
This time, temporary factors are almost certainly playing the biggest role. Not only has the business cycle reached the point at which productivity growth usually slows, it also has several characteristics that may have exacerbated temporary productivity swings. One is the housing bust. Much of the recent weakness in output growth is thanks to the fall in building activity. Yet employment in construction and other housing-related industries has barely budged. It is hard to isolate workers' productivity in residential housing, but Jan Hatzius of Goldman Sachs estimates that it was 13 percent lower at the end of 2006 than a year earlier, whereas in the rest of the economy (outside farming) productivity rose by a healthy 2.8 percent. His analysis spells trouble for the economy's short-term health: once builders stop hoarding workers, the unemployment rate will rise. But there would be scant need to worry about a broad slip in productivity.
Unusually savage company cost-cutting early in this cycle is another reason why recent productivity swings have been so extreme. Robert Gordon of Northwestern University has long argued that much of the improvement in productivity after 2001 was a one-off event as firms tightened their belts after the bursting of the stockmarket bubble. In a new paper ("Explaining a Productive Decade," forthcoming from the Brookings Institution), Stephen Oliner, Daniel Sichel and Kevin Stiroh, all of the Federal Reserve, provide more evidence of this. They show that industries which saw the sharpest drop in profits between the late 1990s and 2001 also saw the largest gains in worker productivity after 2001, probably because they were the most reluctant to hire workers. Today's productivity dip may be the mirror image of that, as job growth catches up with output.
An odd business cycle makes it hard to gauge what has happened to America's underlying rate of productivity growth. So too do shifts in the sources of productivity growth. In the late 1990s workers' efficiency rose thanks both to rapid investment, particularly in information technology (IT), and to innovation, again mainly in IT. Hence the conventional view that America's productivity miracle was based on its ability to harness the power of computers.
Time to Upgrade the PC
After 2001, however, productivity gains had less to do with investment or information technology directly. Messrs Oliner, Sichel and Stiroh show that productivity growth was spread more broadly across industries and was partly the result of the reallocation of resources between firms.
Is this good or bad news? Optimists reckon the dispersion makes sense. Just as the efficiency-enhancing effect of steam power or electricity took years to seep through the broader economy, so the IT revolution is now changing business processes, and boosting productivity, throughout the economy. But worrywarts fret that the ingredients of the 1990s productivity boom are missing. Capital spending has been sluggish and is now falling. By some measures the pace of innovation in IT has slowed. John Fernald of the San Francisco Fed points out that the rate at which computer and software prices are falling relative to other overall prices — a crude gauge of IT innovation — has slowed from 8.75 percent a year in the late 1990s to 6 percent.
For the moment, the consensus among academic experts seems to be that America's trend rate of productivity growth is below the blistering pace of 2002-04, perhaps slightly lower than in the late 1990s, but well above the 1.5 percent average of the previous two decades. Messrs Oliner, Sichel and Stiroh, for instance, reckon trend productivity growth is around 2.25 percent. But no one is very sure. And if unemployment continues to fall while the economy remains weak, the more nervous the productivity gurus will become.