Productivity lost?

Any debate about the future growth rate of “total factor productivity” sounds like a caricature of academic economic discourse. To a layperson, economists may as well be arguing about the most efficient placement of angels on a pinhead. But working papers that try to divine trends in productivity growth are not just useful to other economists but also incredibly important for understanding the future health our economy. A recent study by an economist from the Federal Reserve Bank of San Francisco is one such paper.

But first, some clarifications on terms. When most people talk about productivity, they’re referring to labor productivity: how many cars can a worker assemble or how many tables she can wait on in an hour. When workers become more productive, they can produce more output in a set amount of time and their wages increase. But when economists look at the total economy, they also look at a different kind of productivity: total factor productivity—a measure of how efficiently both capital and labor are used to produce output.

John Fernald’s recent paper looks at both labor productivity and total factor productivity and their trends before, during, and after the Great Recession of 2007-2009. The San Francisco Fed’s senior research advisor argues that the slowdown in both labor and total factor productivity actually predates the recession, starting around 2004. The culprit in Fernald’s narrative is the disappearance of strong gains in productivity from information technology (IT). He shows that the decline in total factor productivity is concentrated in industries that create IT or use it intensely and not in the “bubble” industries of housing, finance, and natural resources. His results imply that productivity growth is back to the slower pace of the 1973-1995 era.

The consequences of this reduced productivity growth are significant. Fernald uses his total factor productivity figures to create an estimate of potential gross domestic product, how large economic output would be if all resources were fully utilized. His estimate is so much smaller than that of the Congressional Budget Office that it implies that about three-fourths of the apparent difference between current economic output and potential output is because of reduced potential GDP.

Fernald’s findings have consequences for the long-run prosperity of the United States as well as the exit strategy from the policies used to fight the Great Recession. His work would imply that our growth problem is secular, to use Greg Ip’s terminology. In other words, our economy has a supply-side problem that has reduced the rate the economy can sustainably grow at. This would mean the Fed should raise interest rates soon, but keep rates low in the long run as the natural rate of interest has dropped due to the decline in potential output

What’s frightening is that Fernald’s work doesn’t include the potential damage to long-run economic growth done by the Great Recession and the weak policy response. Total factor productivity growth could be even lower as the effects of the recession casts a shadow forward into our economic future as laid-off workers and idle equipment become less productive and unable to contribute to economic growth.

Of course, productivity growth could leap back up again and Fernald acknowledges this. Massachusetts Institute of Technology economists Erik Brynjolfsson and Andrew McAfee argue in “The Second Machine Age” that our economy is on the brink of high productivity growth due to new technology. The last jump up in productivity delivered courtesy of the IT revolution of 1990s surprised economists and the next one could as well.

Fernald’s work is not the last word on this topic and the paper will certainly be contested. But his paper is an example of the real world implications of seemingly obscure research, even time-series econometric studies of productivity trends.