American middle class wages haven’t been rising as rapidly as they once were, and a slowdown in productivity growth is probably an important cause.
In mature economies, higher productivity typically is required for sustained increases in living standards, but the productivity numbers in the United States have been mediocre. Labor productivity has been growing at an average of only 1.3 percent annually since the start of 2005, compared with 2.8 percent annually in the preceding 10 years. Without somehow improving productivity growth, living standards will continue to lag, this widely held narrative concludes.
Still, not everyone views the situation this way. For instance, Marc Andreessen, the Silicon Valley entrepreneur and venture capitalist, says information technology is providing significant benefits that just don’t show up in the standard measurements of wages and productivity. Consider that consumers have access to services like Facebook, Google and Wikipedia free of charge, and those benefits aren’t fully accounted for in the official numbers. This notion — that life is getting better, often in ways we are barely measuring — is fairly common in tech circles.
Until recently, this debate was inconclusive. It consisted mainly of anecdotes, with individuals describing how important advances like the Internet were — or were not — to them personally. But now Chad Syverson, a professor of economics at the University of Chicago Booth School of Business, has looked more scientifically at the evidence and concluded that the productivity slowdown is all too real. These results are outlined in his recent National Bureau of Economic Research working paper “Challenges to Mismeasurement Explanations for the U.S. Productivity Slowdown.”
Professor Syverson notes that a slowdown has come to dozens of advanced economies, more or less at the same time, which indicates it is a general phenomenon. Furthermore, the countries with smaller tech sectors still have comparably sized productivity slowdowns, and that is not what we would expect if a lot of unmeasured productivity were hiding in the tech industry.
An additional problem for the optimistic interpretation is this: The productivity slowdown is too big in scale, relative to the size of the tech sector, to be plausibly compensated for by tech progress. Basically, under a conservative estimate, as outlined by Professor Syverson, the productivity slowdown has led to a cumulative loss of $2.7 trillion in gross domestic product since the end of 2004; that is how much more output would have been produced had the earlier rate of productivity growth been maintained. To make up for this difference, Professor Syverson estimates, consumer surplus (consumer benefits in excess of market price) would have to be five times as high as measured in the industries that produce and service information and communications technology. That seems implausibly large as a measurement gap.
Keep in mind that while Facebook is free to its users, much of the tech industry consists of services that are measured and at least partly incorporated into G.D.P. For instance, the free Uber app arranges paid transportation. Even Facebook, Google and Wikipedia have a lot of their value expressed in G.D.P. figures, albeit indirectly. Those services make people eager to pay for smartphones, broadband connections and iPads, and those purchases are recorded as part of G.D.P.
From this standpoint, the measurement problem isn’t all that serious. It’s like going to an all-you-can-eat restaurant buffet. The chicken may be free, but the presence of so many quality items induces customers to pay more for the overall bundle, and that registers as a sale, regardless of the price of the individual items you put on your plate. Also, some of the value of Google and Facebook comes from the advertisements, which induce many purchases, which again are recorded in the national statistics. Consumers pay for many apps as well.
Or look at it this way: The tech economy just isn’t big enough to account for the productivity gap. That gap has caused measured G.D.P. to be about 15 percent lower than it would have been otherwise, yet digital technology industries were only about 7.7 percent of G.D.P. in 2004. Even if the free component of the Internet has become more important since 2004, it’s hard to imagine that it is so much better now that it accounts for such a big proportion of G.D.P.
Alternatively, consider the 2012 estimate of the gains from free Internet services made by Erik Brynjolfsson, professor of management at M.I.T., and Joo Hee Oh, assistant professor of management at the Erasmus University Rotterdam School of Management. They looked at how much time people spent on the Internet, and using that method they valued free Internet services at about $106 billion a year. That’s less than 1 percent of G.D.P., and again it doesn’t close the measured productivity gap. Professor Syverson considers other measures of Internet value as well, but all the different numbers keep circling back to the same conclusion: The productivity gap is real.
None of these remarks should be taken as a lack of appreciation for information technology. Arguably the Internet brought its biggest gains in the mid- to late 1990s, and in those years measured economic productivity was in fact very high. Information technology is still one of the most dynamic sectors of the American economy, and it probably will remain so, and grow yet more influential, even if its absolute impact is not as large as the optimistic revisionists suggest.
That still leaves us with a big productivity problem. We need to acknowledge that while many Internet entrepreneurs are economic heroes, statisticians are also pretty good at what they do, and we may simply need to accept some of the lessons embedded in their numbers. America’s productivity crisis is real and it is continuing. While information technology remains the most likely source of future breakthroughs, Silicon Valley has not saved us just yet.