Is Technology Stagnating?

Without breakthroughs in innovation more powerful than we've recently seen, long-term growth in U.S. output is likely to disappoint.
This is the digital revolution.

This is the third of four posts about VoxEU's e-book on secular stagnation. The first was an overview of the secular-stagnation debate; the second looked at the decline in the equilibrium rate of interest.

Last time, I discussed the core of the secular-stagnation threat -- the prospect of a long-run excess of saving over investment, together with the fact that nominal interest rates can't go negative to eliminate the excess. This combination can cause a persistent gap between potential output and actual output -- a recession that doesn't go away.

This would be bad enough, but the outlook is even grimmer if, setting aside the output gap, potential output stagnates in its own right. Robert Gordon's essay for the VoxEU e-book argues that growth in potential output has already slowed more than most economists recognize, and it is unlikely to pick up.

Gordon is well-known as a technology pessimist, but in his essay for VoxEU he resists the label. The technology slowdown in the U.S., he argues, is already decades old. His main point isn't that technological progress will slow further (though he thinks it might), merely that there's no reason to expect it to rise from its existing low track.

This weak trend, he says, will then encounter four additional "headwinds": stagnant population growth, the defects of the U.S. education system, rising income inequality (which diverts the benefits of growth, such as they are, to a rich minority), and rising public debt (which will require higher taxes to service). It's the headwinds, rather than any new technology slowdown, that do the damage.

Gordon estimates that U.S. potential output is growing by only about 1.5 percent a year -- less than most other forecasters assume. On this view, as he notes, the gap between actual and potential output "is currently quite narrow." Gordon, in other words, is talking about a different kind of secular stagnation: not permanent recession, but persistently slow growth even at full employment.

Let's suppose Gordon is right about the headwinds. If he is, long-term growth at the more comfortable rate of 2 percent or more a year would require faster long-term growth in productivity -- a technological acceleration. Is that plausible? It's a large claim, and Gordon finds the idea "utterly unconvincing." For technology optimists to be proved correct, the power of innovation over the next half-century would have to exceed what's happened since the 1970s:

The mainframe era that eliminated routine clerical jobs of endlessly retyping contracts, bills, and legal briefs; the invention of the personal computer that allowed many professionals to write their papers without the aid of a secretary; the invention of game-changing technologies in the retail sector including the ATM machine, barcode scanning, self checkout, and airline automated check-in kiosks; Amazon and e-commerce; wiki and the availability of free information everywhere; the obsolescence of the hard-copy library catalogue, the auto parts catalogue, the print dictionary and encyclopedia.

Joel Mokyr, another contributor and a leading historian of technology, puts the other case. The computer revolution, he says, "has led to the re-invention of invention." And it's just getting started.

Digital technology is everywhere, from molecular genetics to nanoscience to research in medieval poetry. Quantum computers, still quite experimental, promise to increase this power by orders of magnitude. In much recent writing, the importance of information and computer technology on output and productivity has been stressed, and it is clearly of great importance. What needs to be kept in mind, however, is that the indirect effects of science on productivity through the tools it provides scientific research may, in the long run, dwarf the direct effects.

If that's true, why don't you see more of it in the numbers? Because, Mokyr says, the numbers are misleading. Digital goods and services are harder to count than objects that roll off production lines. Also, once developed, digital products can be supplied at very low cost. That's great for living standards but doesn't add much to gross domestic product.

The disagreement between Gordon and Mokyr is hard to resolve. It's possible, and quite likely, that both are right.

The still-unrealized possibilities of the digital revolution do seem boundless. Measured in qualitative terms -- by the actual and prospective changes in our ways of living -- innovation is surely not faltering, and could indeed be accelerating. Yet, even if it is, Gordon could be right that measured productivity and GDP will grow more slowly than generally assumed. And, unfortunately, that matters: Slower growth in measured GDP will cause economic and political stresses even if the statistic fails to capture all the blessings of innovation.

I consider myself a technology optimist, but Gordon's calculations aren't easily dismissed. Without breakthroughs in innovation more powerful than we've seen for many years, and regardless of what happens to interest rates and the output gap, long-term growth in U.S. output is likely to disappoint.

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.