Does lack of tech-driven unemployment mean US innovation is moribund?

“Is this the best the US economy can do?” is perhaps the most pressing economic question there is. (Right along with it: “How do we create faster, sustained growth whose benefits are more widely shared?”)

March 2013. Artificial Intelligence Laboratory of the University of Zurich (AI Lab). REUTERS/Michael Buholzer.

So it is a real treat to read an insightful debate between two key thinkers on this issue, such as the one presented by [email protected] between Northwestern University economists Robert Gordon and Joel Mokyr. One sees stagnation, the other acceleration. The techno-pessimist vs. the techo-optimist. Both give great summaries of their positions. Gordon thinks the computer revolution has already worked its magic on the economy in terms of faster productivity and growth. (And readers of his magisterial “The Rise and Fall of American Growth,” know he’s hardly gobsmacked by the potential of Big Data, AI, or self-driving cars.)

Mokyr, on the other hand, is greatly enthused by the feedback loop between scientific invention (armed with evermore sophisticated tools) and business innovation. Technology and science driving each other. Well, all that and the prospect of billions of minds with internet access “to the union of all of the scientific journals and magazines that exist in libraries.” (Also check out his new book, “A Culture of Growth: The Origins of the Modern Economy.”)

This is a definite “read the whole thing” sort of thing. But I did want to highlight one bit that gets to the issue of technological unemployment. Gordon:

I would like to start from a paper that was written in 2013 by two economists from Oxford, England who predicted that in the next 20 years, 47% of the jobs would be eliminated. And we’re now four or five years into their 20 years and nothing that they have predicted has happened yet. Let’s look at some of the job categories that they suggest will be greatly reduced.

They predicted that over the next 20 years, 55% of the jobs of airline pilots would be replaced. Well, we’ve seen no move whatsoever by the federal regulators to allow commercial airlines to be flown with one pilot instead of two. They suggested that 86% of real estate agent’s jobs would be eliminated, but we still see humans involved in virtually every real estate transaction. They said that 92% of the jobs of retail clerks would be eliminated. Yes, we are moving towards e-commerce, but so far, as of last year e-commerce only made up 8% of retail trade. So there is a long way to go before all of the jobs of those who work in the retail sector are eliminated.

As you mentioned, there are techno-optimists and techno-pessimists, I always like to reassure my audiences when I come up with the story of slow growth that I expect to continue in the future, that this is good news for jobs because what we’ve seen in the last eight years, the creation of 16 million new jobs, is not a fluke. We will continue to see jobs created rather than destroyed.

I think that’s a legit point. And I have before raised questions about that Oxford study, which is the basis for the most extreme job-loss forecasts. When you look at the task content of jobs rather than occupation, estimates tend to show less than full automatability, raising more a “race with the machines” scenario. So maybe tech unemployment isn’t the best way to gauge innovation.

(Two things: First, there is reason to believe that innovation diffusion, not creation is the big problem. Second, something does seem to happening now in retail. )

Of course, tech creates jobs, too! One example: economist Michael Mandel estimates that “app economy” employment in the US totaled nearly 2 million as of December 2016. There is a happy ending here that involves mass employment, not unemployment.