LA Times, May 6, 2015
Are robots displacing millions of workers? Many people seem to think so. Recently, for instance, the New York Times ran an op-ed claiming that “the machines are getting smarter, and they're coming for more and more jobs.” On Tuesday the Wall Street Journal sounded the alarm that “robots are taking over corporate finance departments.” The story goes that we can look forward to an ever greater problem of unemployment as technological advancement allows machines to replace a growing percentage of the workforce.
It's a scary story. But, first of all, it's not new. Second, although robots are “coming for” some jobs, that fact does not explain the current economic situation. Robots are a distraction from very real problems.
On the first point, concerns about machines creating mass unemployment date back centuries. The term “Luddite” refers to textile workers in the early 19th century who smashed the machines that threatened to make their skills redundant. When I was young, Kurt Vonnegut's “Player Piano” was a must-read tale of a society in which machines did all the work and people were unnecessary. So we have seen this script many times before.
Turning to the evidence, if technology were rapidly displacing workers then productivity growth—the rate of increase in the value of goods and services produced in an hour of work—should be very high, because machines are more efficient. In the last decade, however, productivity growth has risen at a sluggish 1.4 percent annual rate. In the last two years it has limped along at a pace of less than 1 percent annually. By comparison, in the post-World War II “Golden Age,” from 1947 to 1973, productivity grew at an annual rate of almost 3 percent.
And let's not ignore that the rapid productivity growth of the Golden Age was associated with low unemployment and strong wage growth. So even if we did see a big uptick in productivity growth, there is no reason to assume that this would be bad news for (human) workers.
The other scary robots story that won't seem to die is that even if technology isn't hurting everyone and making all humanity redundant, it is responsible for growing inequality.
The economist David Autor popularized this “hollowing out of the middle” narrative, which holds that many of the jobs that used to support a middle-class living standard, like manufacturing and clerical work, are disappearing because of technology. As a result, less-skilled workers are scrambling for an ever-smaller number of jobs and falling behind their better-educated peers.
But the data don't support that theory either. If Autor's story were true, we would expect to see higher demand for high-skilled jobs relative to less-skilled jobs. That's not happening. Instead, there's been a modest decline in the relative demand for almost all occupations above the bottom 25% to 30%. In the years since 2000, most of the job growth has been in occupations at the bottom end of the wage distribution, with the share at the middle and top both declining.
Inequality is real, but it has more to do with the wages certain occupations command than the myth that machines are lowering the quantity of work that's available for the less-skilled.
Still, the robots story is important because it perpetuates another myth: that inequality is something that just happened, when in fact it's the consequence of specific policies.
If robots are the problem, we could feel bad about it and maybe look to help those who are on the losing side of the great human-android war. But few would want to be Luddites standing in the path of technological progress. Opposing mechanization seems futile. On the other hand, if the rise in inequality over the last 35 years is due to government action or inaction, then we might try to bring about change. The list of policies that have led to inequality is long. It includes a trade policy designed to whack the middle class; Federal Reserve Board policy that fights inflation at the expense of jobs; a bloated financial sector that relies on government support; and a system of labor-management relations that is skewed against workers.
If we really want to address the causes of inequality we have to get over the robots are taking our jobs story, move beyond handwringing or Luddism, and accept the harsh reality that people, or their policies, anyway, are to blame—not technology.