Computers used at a shale gas site in Washington: “the way in which technologies are applied to production processes is important”
© Keith Srakocic/ PA Images
Robert Gordon’s remarks in the April issue of Prospect outlined a pessimistic view of US growth. His theory about why US productivity has stagnated is fascinating and quite worrying. Persistent poor productivity levels (economic output relative to input) would imply that the United States is on course for very low rates of economic growth. I can imagine many pessimistically-inclined readers immediately jumping on Gordon’s ideas, seeing in them a justification for their own hunches. But I do not accept his pessimistic outlook for economic growth. I have four reasons for worrying less.
First, as authoritative as Gordon is on the topic, do we really know what we are measuring with productivity data anymore? Many senior economists will probably be horrified to read me saying this—or perhaps pity me as a lesser mortal. But in a modern economy, measuring the output from a service sector business is very difficult, much more so than counting up the number of cars rolling off a factory production line. So can we be so sure of ourselves when measuring productivity in a modern economy? If we cannot, then conclusions about long term growth based on these productivity measurements begin to feel uncertain.
In any case, do normal people really care about productivity? Certainly, economists know of its importance, or at least think they do. But given how difficult it has been to anticipate changes in productivity, and how intangible the idea itself actually is, can it really be that important? People care about their jobs, incomes and lives—not especially their productivity. Gordon suggests that many modern inventions have turned out to be very unproductive, which may lead to economic problems as a result. But as that problem becomes apparent, policymakers will either choose, or be forced, to deal with it.
Now as I say, economists in general will probably pity me for raising such issues, but I say it partly because Gordon might well be wrong. How can anyone know in advance whether technological innovations are going to change all of our productivity?
Yet a second, stronger point follows closely from this. While the adoption of modern technologies might not be as productive as the advent of the motor…