Matt O’Brien had a very good piece on the silliness of the robots taking our jobs story. The basic point is that it is silly to worry about a possible future in which robots are taking our jobs, when we currently face a situation in which people don’t have jobs just because Congress won’t spend the money. I couldn’t agree more.

We can all see the really cool things that can be done by robots and advanced computers, but the fact is they are not doing it now. As Matt notes, productivity growth has been very slow in the last decade, the story of robots taking our jobs is one in which productivity growth is very fast.

There are two points worth adding to Matt’s comments. First, he refers to an often cited analysis that finds 47 percent of all jobs are at risk of being automated over the next twenty years. Sounds pretty scary, right? Well let’s imagine that all of the 47 percent of those at risk jobs gets computerized over the next two decades. (The study just identifies these as “at risk” jobs, a high proportion of which will be computerized, not all of them.)

This rate of computerization would translate into 3.1 percent annual productivity growth. That’s a hair higher than the 2.9 percent annual rate of productivity growth that we saw in the Golden Age from 1947–1973. That was a period of low unemployment and rapid real wage and income growth. If there is a reason that we should be scared in this story it is not because of the productivity growth, but rather an institutional structure that prevents most workers from benefitting from this growth.


This brings up the second point. Matt refers to another study that notes the large shift from labor to capital over the last 35 years. This study attributes roughly half of this shift to the “people who own the robots” to technology, the rest is not explained.

It is important to recognize that “owning robots” is a political issue, not an economic one. Specifically, people own robots because we give them patent monopolies. In most cases robots would be very cheap to produce if the government didn’t threaten to arrest people for not respecting patents. After all, they don’t require that much by way of materials, most of which are plentiful. As for the labor involved in building robots, I’m sure there are many robots that would be happy to do it.

So we end up with money going from the rest of us to people who own robots because the government has given these people a monopoly over the use of the technology. Suppose the government didn’t give a monopoly over the use of the technology. Suppose that we funded the research through a different mechanism or at least made the monopolies shorter and weaker. Then the folks who developed the robots would not have so much money, the robots would be cheaper, and the rest of us would be richer.

The best way to provide incentives for research is a longer story, but the key point here is that the incentive structure that makes the owners of robots rich at the expense of the rest of us is an explicit government policy. The robots didn’t write the rules, the folks in Congress and the White House did.