The Paradox of Productivity
In his nomination acceptance speech at last week's Republican convention, one of Donald Trump's central themes is that free trade has ravaged American manufacturing. He proclaiming himself the voice, not just of the common man or of the blue-collar worker, but specifically of "the laid-off factory workers and the communities crushed by our horrible and unfair trade deals." Blaming "bad trade agreements" (by which he clearly means free trade agreements) for the fact that "America has lost nearly one third of its manufacturing jobs since 1997," he promised to "make our country rich again" because we're going to "start making things again."
This picture of American manufacturing is not even remotely true—though there is a reason why, to a lot of people, it seems true. Refuting this persistent error tells us a lot about mechanization, automation, and the actual source of America's wealth.
First let's look at the facts. American manufacturing is not in decline, and we still make things. A lot of things.
American manufacturing output has been expanding all along, despite the usual recession-induced dips, and recently reached a new record level.
Manufacturing output index rose in Q2 to 130.74 (2009=100), the highest on record http://t.co/VBeGOifxBr pic.twitter.com/N6Zhpr2agU
— St. Louis Fed (@stlouisfed) August 11, 2015
America is still the second largest manufacturer in the world, only recently eclipsed by China. Even then, China is actually much less of a manufacturing nation compared to the US, in the sense that it produces far less per capita. That is, the average Chinese worker creates less value than the average American worker. Keep that in mind, because it's going to be important.
The point is that there is no real evidence that Chinese manufacturing has been taken away from US manufacturing. Both have been growing, and American manufacturing has been growing in line with the overall growth of our economy. While some specific things that used to be manufactured in America are now manufactured in China, America has never stopped making things.
So why do Trump supporters hold the profound conviction—which, as I have discovered, fire cannot burn out of them—that American manufacturing is dying because China is taking all of our jobs?
Well, first of all, if your factory is the one that got offshored, it's natural to assume that your specific plight must reflect the general condition of the American worker. This is an old point about economics: people extrapolate from what they can see, from their own particular condition and from anecdotes. But as Henry Hazlitt used to remind us, the job of actually understanding the economy involves looking beyond these specific circumstances to the wider truth.
There is a much more important reason, however, why people think manufacturing is in decline. Manufacturing employment is in decline. The actual output of goods is increasing, but fewer workers are required to make them. It's all captured in this graph:
What's actually going on is not that the Chinese are taking our jobs. It's that machines are taking our jobs. American manufacturing has become more and more high-tech and automatized, particularly with recent advances in robotics, which are only just getting started.
And here's the thing that's really, really hard for people to accept: this is a good thing. This is what we want. This is, and always has been, the real wellspring of American prosperity.
To understand this, let us begin with the Parable of the Spoons. This is an old story told in several different forms but usually attributed to the free-market economist Milton Friedman. Here's the version told by Stephen Moore:
At one of our dinners, Milton recalled traveling to an Asian country in the 1960s and visiting a worksite where a new canal was being built. He was shocked to see that, instead of modern tractors and earthmovers, the workers had shovels. He asked why there were so few machines. The government bureaucrat explained: "You don't understand. This is a jobs program." To which Milton replied: "Oh, I thought you were trying to build a canal. If it's jobs you want, then you should give these workers spoons, not shovels."
Just as an earthmover is a machine that takes jobs away from workers with shovels, so a shovel is a tool that takes jobs away from men with spoons. And the spoons presumably take jobs away from the extra men who would be needed to dig the earth with their bare hands. But notice something else. If the men perform the work with shovels, or spoons, or their bare hands, they accomplish exponentially less. And because they produce less, their labor is correspondingly far less valuable.
In the terminology of economists, increases in wages are driven by increases in productivity, the measure of how much a worker is able to produce in an hour of labor. When workers are more productive, when they accomplish more with the same amount of work, their labor is more valuable and they tend to get paid more.
Conversely, a worker with a spoon has a tiny, tiny fraction of the productivity of a worker with an earthmover, and a nation of workers with spoons would be living in abject poverty, far below the level of an advanced, industrial nation. And that's not economic theory. I'm referring to guys like this.
By the same token, the more jobs machines take over, the higher the productivity of the workers who run those machines. That often means they need to acquire higher-level skills—instead of just turning a wrench on an assembly line, they have to understand complex mechanical systems—but they also tend to get paid a lot more.
Let's go back to that observation about Chinese manufacturing. Their total output is now a bit higher than ours, but it's spread out over many more workers. Chinese workers have taken over some manufacturing jobs that used to be done in the US, but for the most part, they have taken over the low end of manufacturing. They have taken over jobs with lower productivity and lower pay. So when somebody like Donald Trump says he wants to bring back those jobs from China, he's actually promising his supporters low-end, low-paying, dead-end jobs. That's not exactly going to make America great again.
By contrast, if we really want greater productivity, we need more machines to take over our jobs. In fact, that is the argument recently made by one economist who is concerned that advances in robotics aren't coming fast enough for areas outside of manufacturing, causing growth in productivity and wages to stall out. The interviewer sums up the issue quite succinctly.
[A]fter the productivity of American workers grew by an impressive 2.9 percent per year between 1995 and 2005, it grew by just 1.3 percent between 2005 and 2015. That's far slower than the productivity of American workers grew, on average, during the 20th century. And slow rate of progress is a major reason that middle-class wages have been stagnant over the last decade....
For all the hype about robots taking peoples' jobs, there's been hardly any progress automating labor-intensive industries like retail, medicine, or construction over the past few decades. That might seem like a positive development if you're worried about losing your job to a robot. But it's a cause for worry if you're hoping for bigger paychecks over the next decade or two.
This is what you might call the Paradox of Productivity. If you want a better-paying job, you have to let a machine take it away from you. The paradox is that it seems as if you will then be out of a job and won't enjoy any of the benefits. The solution to the paradox is that you will have the opportunity to do a new job managing the work of the machines.
None of this is at all new. The same prosperous, mid-20th-Century manufacturing jobs everyone is so nostalgic about were themselves the product of machines "taking away jobs" from a previous generation of workers. The Industrial Revolution was history's greatest leap forward in human prosperity, and it consisted entirely of machines taking away people's jobs. At the time, a lot of people were unhappy about that. The Luddites famously smashed steam-driven power looms in the early 19th Century because they were convinced they were going to "take away the jobs" of people who were weaving cloth on hand looms. But what actually happened? A while back, I did a little calculation: "Between the Cartwright loom of 1785 and the Lancashire loom of 1842 (the first automatic power loom), a single worker went from supervising one machine to supervising six machines operating at twice the speed—a twelve-fold increase in his productivity." The result was an abundance of lower-priced cloth made by better-paid workers.
Extend that process forward another few hundred years, and you get something like this: recent gains in manufacturing productivity that result in "the average American factory worker today produc[ing] more output in an hour than his or her counterpart produced working almost a ten hour day in 1947."
Do you want to make America rich—not again, because we were much poorer in 1947—but do we want to be even richer than we are now? Then this is how we do it. Get machines to take our agricultural jobs, get them to take our manufacturing jobs—and now, in a new era of software and "artificial intelligence," get them to take our white-collar jobs.
That's the Paradox of Productivity in action: If you want a better-paying job, you'd better let a machine take it away from you.