The Myth of Technological Unemployment

If the nightmare of technological unemployment were true, it would already have happened, repeatedly and massively.

As a savvy reader, you already know that technological change is why the jobs in manufacturing are drifting away from Youngstown, Ohio. You know that most of the drift goes to other American cities, such as Houston or Chattanooga. You know that Appalachian jobs in coal mining are not coming back, because new techniques have permanently cheapened natural gas. You know that the Trump administration's scapegoat, foreign competition, bears little responsibility for any of this. And when foreign encroachment does happen, you know it's good, not bad, for most Americans.

Still, many reasonable people fret. Isn't technological unemployment a real and serious problem? Non-economists of a quantitative bent fret about what we're going to do when all the jobs go away—when, say, autonomous vehicles replace America's 3.5 million truck drivers.

Even some economists, brilliant ones, think we're in trouble. Robert Gordon of Northwestern suggests it in his recent book The Rise and Fall of American Growth (Princeton University Press). Tyler Cowen of George Mason University does too, in Average Is Over (Dutton). The great if misled John Maynard Keynes believed we would lose jobs from technology. So did the still greater David Ricardo.

They were wrong.

Otherwise sensible folk are, for some reason, terrified by robots. Yet the results of automation are good overall. Workers move from wretched assembly-line jobs to better ones standing in white coats monitoring the robots, at the higher wages made possible by the higher tech. Or, even better, they move to jobs outside the auto industry, earning pay that goes further because people can buy the radically cheaper stuff the robots now make.

If their new jobs are not higher paying, it's probably because the auto union managed to extract monopoly profits from the company, and therefore from consumers. Robert Reich, a reliable source of sweetly leftish errors of facts and ethics, declares that "the decline in unionization [of private companies] directly correlates with the decline of the portion of income going to the middle class." But paying selected workers on the assembly line more than they can earn elsewhere, at the expense of other, sometimes poorer, workers' ability to buy cars, is hardly an ethical formula for raising up the working class.

When a Ford plant installed robots, Walter Reuther, a long-ago president of the United Auto Workers union, is said to have asked a manager: "How are you going to get them to buy Fords?" But Reuther's argument is fallacious. Employees of car companies are a trivial share of the car-buying public. You can't create prosperity merely by having workers purchase from their own employers.

Reich has accused the following things of driving down American wages: "Automation, followed by computers, software, robotics, computer-controlled machine tools and widespread digitization." But such innovations have actually raised real wages, correctly measured, because a human supplied with a better tool can produce more outputs. And the point of an economy is production for consumption, not protection of existing jobs.

Consider the historical record: If the nightmare of technological unemployment were true, it would already have happened, repeatedly and massively. In 1800, four out of five Americans worked on farms. Now one in 50 do, but the advent of mechanical harvesting and hybrid corn did not disemploy the other 78 percent.

In 1910, one out of 20 of the American workforce was on the railways. In the late 1940s, 350,000 manual telephone operators worked for AT&T alone. In the 1950s, elevator operators by the hundreds of thousands lost their jobs to passengers pushing buttons. Typists have vanished from offices. But if blacksmiths unemployed by cars or TV repairmen unemployed by printed circuits never got another job, unemployment would not be 5 percent, or 10 percent in a bad year. It would be 50 percent and climbing.

Each month in the United States—a place with about 160 million civilian jobs—1.7 million of them vanish. Every 30 days, in a perfectly normal manifestation of creative destruction, over 1 percent of the jobs go the way of the parlor maids of 1910. Not because people quit. The positions are no longer available. The companies go out of business, or get merged or downsized, or just decide the extra salesperson on the floor of the big-box store isn't worth the costs of employment.

What you hear on the evening news is the monthly net increase or decrease in jobs, with some 200,000 added in a good month. But the gross figure of 1 percent of jobs lost per month is the relevant one for worries about technological unemployment. It's well over 10 percent per year at simple interest. In just a few years at such rates—if disemployment were truly permanent—a third of the labor force would be standing on street corners, and the fraction still would be rising. In 2000, well over 100,000 people were employed by video stores, yet our street corners are not filled with former video store clerks asking for loose change.

We could "save people's jobs" by stopping all innovation. You would do next year exactly what you did this year. Capital as well as labor would perpetually be employed the same way. But then we would perpetually have the same income. That's nice if you're doing well now. It's not so nice if you're poor or young.

Job protections for the old have in fact already created a dangerous class of unemployed youths in the world—50 percent among Greeks and black South Africans, for instance.

Helping the poverty-stricken is laudable. But we can't subsidize 1.7 million people a month. Nor is job retraining a good idea when directed from above: The wise heads in Washington don't know the future, and they'll end up teaching people to be machinists for companies that won't exist. Workers themselves know best how to retrain and relocate, as did the hundreds of thousands who moved to North Dakota during the brief oil boom there. We want the labor force to be as flexible as the capital force. And for that we need liberty, not government programs.

In the spirit of John Rawls, we should ask which society we'd rather enter at birth, without knowing where within that society we'd end up. One in which all jobs are protected, bureaucrats decide who gets subsidies and who doesn't, and the economy slides, as France has, into stagnation and high levels of youth unemployment? Or one in which labor laws are flexible, individual workers decide their own futures, and the economy lifts up the poorest among us?

Choose, and then quit worrying about technological unemployment.