Tech's Next Surprise: Jumpstarting Productivity
A paradox has stumped economists for years: Why has productivity stalled in a nation bursting with technological innovation? They’ve mostly ruled out the most obvious potential explanation – mismeasurement – while considering several others. It's likely, though, that we're just witnessing a temporary lull: Tech advances drove productivity growth before 2010, and they're likely to do so again in a not-so-distant future.
Last year, Robert Gordon of Northwestern University argued in his book "The Rise and Fall of American Growth” that the "third industrial revolution," commenced in the 1970s, largely ended by 2005. All the major changes to business processes through advances in information technology -- email and electronic catalogs, desktop publishing and check-in kiosks at airports, bar code scanners and quick credit card authorization -- had already happened by then.
Today's innovations – such as mobile computing, robotics, 3D printing, artificial intelligence, driverless cars -- have little effect on how business works because they are either tangential to it or not developed enough to make a measurable contribution. Most importantly, any productivity gains they generate are being overshadowed – and will continue to be overshadowed, Gordon suggests – by losses due to education's slowing contribution to economic growth, weak population growth, and rising inequality.
This theory explains why productivity has slowed down amid so many exciting stories of technological advances and and hero innovators like Elon Musk. There are only two alternatives to Gordon's tech-pessimistic view: That innovation's contribution to productivity is mismeasured and it's actually high in the present, or that the effect is delayed and we'll see it at some point in the future.
In recent papers that hold sway over this debate, economists have argued that smartphones, Facebook and Google searches mostly provide non-market benefits that make little difference to productivity, and that the productivity slowdown is seen in many countries regardless of their enthusiasm about Silicon Valley innovation.
But a paper published last month by David Byrne of the Federal Reserve Board and two collaborators argues that mismeasurement is still important, just not in an obvious way. Byrne shows that multi-factor productivity, which relates output to different kinds of inputs such as labor, capital, energy, etc., increased faster in the tech industry than official statistics let on -- by 10.9 percent between 2010 and 2015, rather than by 3.1 percent, as official data indicate. Multi-factor productivity is often used as a proxy for innovation; the tech sector is innovating at a faster pace than the rest of the economy, but it doesn't have much effect on labor productivity -- yet.
"The apparent weak pace of innovation in the tech sector provides fuel for the story that little scope remains for the tech sector to boost aggregate labor productivity growth," Byrne and collaborators write. "We believe that these faster rates of growth in high-tech could presage a second wave of higher productivity growth spurred by the digital revolution."
In other words, innovation percolating in the tech industry hasn't yet boiled over into higher labor productivity for the economy as a whole.
That makes intuitive sense, too. Amazon is trying out a store without checkout counters thanks to innovations that are not reliable enough for a full rollout. Tesla is selling tens of thousands of electric cars with impressive features, but both that company and its traditional competitors have yet to bring them to the mass market. Uber is offering rides in self-driving vehicles -- and suppliers are lowering the costs of the technology needed to make such cars at scale -- but again, there's not enough accumulated experience for the technology to go mainstream. Thanks to neural networking technology, Google Translate's output is finally beginning to resemble sentences written by a human -- but it's still not enough to enable reliable professional interaction across languages.
Some technologies have come tantalizingly close to the threshold beyond which they'd revolutionize business processes the way pre-2005 innovations did. The development process has spawned significant new areas of tech, and competition in them is driving down prices and increasing the multi-factor productivity. These innovative areas are like coiled springs now. It's not quite clear how much energy will be released as they "uncoil": some of the technologies can hit their natural limits and never make a major contribution, others may hit regulatory barriers, but yet others may turn out to be major productivity boosters, capable of overcoming the headwinds described by Gordon.
The "retail apocalypse" in the U.S. shows Byrne's intuition may be correct. The revolution hasn't stopped at bar code scanners and credit card terminals -- it just took a while for the changes to sink in.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
To contact the author of this story:
Leonid Bershidsky at firstname.lastname@example.org
To contact the editor responsible for this story:
Mike Nizza at email@example.com