Manufacturing Jobs and the Rise of the Machines

Harvard Business Review

The story of how technological progress is affecting employment — whether, in other words, the robots are eating our jobs — is clearly an important one. But who's telling it correctly? I believe that technological unemployment (and underemployment) is a real and growing phenomenon.

But since Erik Brynjolfsson and I appeared on 60 Minutes in January for "March of the Machines," a story that examined the labor force implications of advanced digital technologies like robots and other forms of automation, we've been accused of being unclear on the concept.

For example, the Association for Advancing Automation said in response that we "are missing the bigger picture" by not recognizing that American companies are "successfully implement[ing] automation technologies instead of going out of business or sending manufacturing overseas." They add: "American manufacturing's embrace of robotics will ensure a new manufacturing renaissance in this country."

If the A3, or anyone else, thinks that lots more manufacturing jobs will accompany this renaissance, they're just dead wrong. The facts are too clear, and they all point in the other direction. For example:

  • Manufacturing employment has been on a steady downward trend in the U.S. since 1980 (it increased some after the end of the Great Recession, but this boost appears to be leveling out).
  • Manufacturing jobs have also been trending downward in Japan and Germany since at least 1990 and, as I wrote earlier, in China since 1996.
  • Manufacturing employment decline is a global phenomenon. As a Bloomberg story summarized: "Some 22 million manufacturing jobs were lost globally between 1995 and 2002 as industrial output soared 30 percent. ... It seems that devilish productivity is wreaking havoc with jobs both at home and abroad."

Rob Atkinson, president of the Information Technology and Innovation Foundation, is another of our detractors. He takes the argument up a level across industries. Even if total manufacturing employment goes down because of automation, he writes, other industries will pick up the slack by employing more people. This is because:

"...most of the savings [from automation] would flow back to consumers in the form of lower prices. Consumers would then use the savings to buy things (e.g., go out to dinner, buy books, go on travel). This economic activity stimulates demand that other companies (e.g., restaurants, book stores, and hotels) respond to by hiring more workers."

Fair enough, but what if those other companies are also automating? One of the most striking phenomena of recent years is the encroachment of automation into tasks, skills and abilities that used to belong to people alone. As we document in Race Against the Machine, this includes driving cars, responding accurately to natural language questions, understanding and producing human speech, writing prose, reviewing documents and many others. Some combination of these will be valuable in every industry.

Previous waves of automation, like the mechanization of agriculture and the advent of electric power to factories, have not resulted in large-scale unemployment or impoverishment of the average worker. But the historical pattern isn't giving me a lot of comfort these days, simply because we've never before seen automation encroach so broadly and deeply, while also improving so quickly at the same time.

I don't know what all the consequences of the current wave of digital automation will be — no one does. But I'm not blithe about its consequences for the labor force, because that would be ignoring the data and missing the big picture.

Press spacebar to pause and continue. Press esc to stop.

Bloomberg reserves the right to remove comments but is under no obligation to do so, or to explain individual moderation decisions.

Please enable JavaScript to view the comments powered by Disqus.