A worker watches as automated robots assemble automobile parts.Simon Dawson/Bloomberg/Getty Images
You’ve probably seen the scary headlines: Prepare yourself, human! A robot is coming for your job. From burger-flipping bots to software lawyers, a new breed of supermachines, we are told, is driving a “jobocalypse” that will push Homo sapiens into obsolescence.
In their book “Race Against the Machine,” Erik Brynjolfsson and Andrew McAfee, both professors of management at MIT, argue that since fewer people have jobs and wages are falling even as productivity and profits rise, automation must be a big factor driving unemployment. They predict it will only get worse as we approach “The Second Machine Age” (the title of their follow-up book), in which expanding computing power will create machines that could replace us at an exponential rate.
How worried should you be? The jobless rate now stands at 6.7 percent — not the terrible double digits of a couple of years ago, but still not great, especially if rates of underemployment and labor nonparticipation are considered. Economists have a lot of theories about why we have had a “jobless recovery,” but most agree that something other than technology is the key culprit. In 2011, a Wall Street Journal survey found that, of economists polled, the majority blamed high unemployment on a weak demand for goods and services. Many economists (including new Fed Chair Janet Yellen) theorize that jobs will return with higher demand, whether or not a Roomba is vacuuming the floor.
If job decreases were really caused by waves of automation, the unemployment rate should have ticked up during the 1990s, when you probably started the decade with a typewriter and finished it with a laptop. But that is not what happened. The U.S. had a buoyant economy and enjoyed robust demand, which led to a low unemployment rate for most of the decade — starting at 5.6 percent and ending even lower, at 4.2 percent. News stories tended to cheer technology, while cultural products, from Errol Morris' documentary “Fast, Cheap & Out of Control” to “Star Trek: The Next Generation,” celebrated robots that complement human work. (Interestingly, “The Terminator,” which depicts the robot as the ultimate threat, came out in 1984 — right in the aftershocks of an unemployment explosion.)
Whether we consider technology a blessing or a curse seems to depend on how the economy is doing.
Fast-forward to the aughts, another period of turbocharged technological innovation. In December 2007, the jobless rate was only 5 percent. Exactly two years later, the rate had nearly doubled. Did a surge of technology suddenly paper the country in pink slips? No, a giant financial crisis shrank demand as people stopped spending money. As the business sector rapidly shed jobs, Washington pursued austerity policies that further reduced aggregate demand and threw many government workers onto the jobless rolls.
Larry Mishel of the Economic Policy Institute has pointed out that blaming technology distracts us from how financial crises and political responses affect employment. “Journalists and policymakers who think they’re getting at the deeper story by focusing on robots are actually missing the big picture,” wrote Mishel. “Robots may not have been even near the site of the crime, so don’t blame them so quickly.”
Yet today, machine-bashing is back. Far from being a friend of the worker, the robot is a pitiless monster that never needs a bathroom break. Lashing out at robots is understandable. Science-fiction writer Isaac Asimov even had a name for the fear that a robot will replace you: the “Frankenstein complex.” Thomas Friedman titled his Jan. 12 column for The New York Times “If I Had a Hammer,” referring to what people might fantasize doing to a machine that is outperforming a human. History, however, shows that this fear may be misplaced.
The Luddite lesson
Even among those who reject robot hysteria, some warn that technology is harming the middle class. MIT’s David Autor recalls the Luddites, 19th century English textile workers who destroyed labor-saving technology because they feared for their jobs. Autor admits that many 19th century Britons benefited from the introduction of mechanized (or power) looms; unskilled laborers got stable jobs as loom operators and a growing middle class gained the means to afford mass-produced fabrics. But he suggests that a group of middle-skilled textile workers probably didn’t benefit on the whole. The same thing, he says, is happening to a similar subset of workers today, and the middle class is suffering as a result.
We have a dangerously lopsided economy wherein the productivity gains from improving technology are funneled to those at the top.
William Lazonick, an economist at the University of Massachusetts–Lowell and an expert in the history of the business corporation, disagrees. As he sees it, technological advances boost productivity, and, depending on how those productivity gains get divvied up, can bring us better, more fulfilling jobs, just as happened in Britain during the 19th century.
According to Lazonick, textile workers who operated power looms in factories were not unskilled, as Autor suggests, and fared better than the cottage handloom weavers whose work the machines ultimately displaced.
The middle class benefited from technology as factory weavers got unions — and, by the latter half of the 19th century, decent wages. Machines and humans complemented each other, and Britain became “the workshop of the world.” In the 1920s and ’30s, the British lost textile markets when the Japanese introduced even more advanced power looms. As a result, Japanese textile workers enjoyed more plentiful jobs, higher wages and better work conditions. (The inventor and leading producer of the Japanese automatic loom was the Toyoda Automatic Loom Co., out of which sprang a very familiar auto company that would help transform Japan into one of the richest nations in the world.)
So if history doesn’t support the idea that workers should fear machines, what about the argument that this time it’s different? Don’t the ATMs and self-checkout scanners popping up on every corner increase unemployment?
No, they don’t. Consider the larger picture: If Bob’s Home Goods decides to replace retail workers with machine scanners, the store might save money and the savings may be passed on to you, the shopper. Now you have money to spend on other things, which boosts the economy. On the other hand, Bob might decide to use some of the profits to open a new store, which requires the hiring of a whole new staff. If the growth of Bob’s enterprise generates more tax revenues, then the government can hire more people to provide social services. Bob might also channel a portion of increased profits into savings instruments that, through the financial system, can help kick off or expand other businesses. Now we’ve got more jobs overall, not fewer, despite some displacement of workers by machines.
If government and business are smart, they will retrain workers replaced by machines to do different jobs that require thinking and social skills. Remember, no matter how fancy the machine, somebody needs to build it, package it, market it, sell it, explain it to the customer and fix it when it breaks down. That should mean more jobs as we hit the next wave of technology, and perhaps more interesting ones.
Here is the real problem: We have a dangerously lopsided economy wherein the productivity gains from improving technology are funneled to those at the top. Instead of sharing profits with workers, executives pay themselves sky-high salaries. Rather than devoting resources to retraining employees or creating new products, they implement stock-buyback plans in order to enrich themselves. They shirk taxes and cheat the taxpayers who invested in the very technology the company uses to succeed.
We should focus instead on changing the behavior of the executives who control corporate resources and who make decisions to introduce new technologies. We could demand a fairer tax system, ban stock buybacks and regulate executive compensation. Our government could help to retrain workers when companies won’t do it and use our tax dollars to invest in technology where people really need it. Health care and education, for example, are both areas that require technology and humans to work together.
Unfortunately, some politicians are using the rhetorically inflated threat of automation to dissuade workers (and voters) from demanding such changes. They warn that if a waiter fights for a better wage, the company will simply buy an iPad to replace him. But even scarier than an iPad-waiter are companies that stop investing in technology and turn their attention to avoiding taxes, manipulating stock prices and cheating workers of their fair share of the profits.
Don’t fear the robot. Fear the Homo sapiens who owns the robot.