What To Do If We Get A Productivity Surge

Farhad Manjoo has a very enlightening series of articles about the possibility of a surge in applied robotics in the future that’s built around a framing conceit that I’m not thrilled with. Basically it’s asking whether or not you (or perhaps we) should be “afraid” that robots will take our jobs away.

I think a better way of looking at this is to say that productivity is increasing all the time in the American economy. And technologies that increase productivity always do displace some workers. But for context, one should note that the past 30 years have been a stretch of unusually slow productivity growth while the 35 years of postwar prosperity featured extremely rapid productivity growth. People sometimes have the sense that current joblessness is related to productivity growth, but that’s just not there in the numbers. What’s more, there’s really no use in fearing or not fearing a possible future technology-driven surge in productivity. The question to ask is what should we want to do if one happens.

One thing that’s obvious is that you have to continue to get macroeconomic stabilization right. A productivity surge plus deficient demand is a recipe for an awkward deflationary period in which many people will suffer. Something like that happened during the Long Depression of the 19th century, and it’s as avoidable as it is undesirable.

That’s pretty basic. My less basic worry is that an automation-driven surge in productivity will put intense pressure on natural resources. In a classical three factors model, you’d say production is composed of labor inputs plus capital inputs plus land and a robot surge would cause all the returns to go to land. In the modern world, we’re not just looking at farmland but at the whole resource suite. That’s fossil fuels and fresh water and urban space and the atmosphere’s ability to absorb greenhouse gas emissions and all the rest. None of these factors are managed very well currently, which is known to be a problem, but a productivity surge will make poor management of our resources are much bigger deal. A related issue is that intellectual property rents become a bigger and bigger deal the more productivity you have.

The last leg of the stool here is that a surge in productivity should increase the amount of time we spend not-working. It’s not sketched into the fabric of the universe that people have a five-day workweek. The weekend is a social invention. Conceivable future people should have a three-day weekend every week. Or maybe just more holidays and more vacation days. If the productivity surge isn’t paired with a big increase in life expectancy, we should start retiring sooner. Not only is leisure valuable on its own terms, but increased leisure will reduce some of the environmental pressures associated with higher productivity. What’s more, one very exciting trend we’ve seen in the past decade is an increase in the use of the Internet to engage in non-market production. Wikipedia, funny cat videos on YouTube, etc. Just because people aren’t working doesn’t mean they’ll be doing nothing all day. We’ve currently marketized a decent share of the work that wives used to do on a non-market basis, and the future we may un-marketize a lot of the entertainment and research functions that are currently marketized.