Is There a Technology Trap?

If you took an introductory economics class in college, you may remember studying for your midterm and assessing the “comparative statics” effect of a change in price: the result can be decomposed into a substitution toward the lower price alternative, and an overall increase in affordable possibilities.

Almost all economic changes have these two aspects, something akin to a “substitution” effect, and something akin to an “income” effect. For example, suppose that the price of food falls. The substitution effect says I would buy more food. I’m also wealthier; the income effect means that I can use some of the savings on food for other things. Cheaper food means I spend more on the opera.

Proponents of the “technological unemployment” flavor of apocalyptical studies—much like their close cousins the environmental eschatologists—focus only on the substitution effect. Wages are falling, or the environment is getting worse, so the end is near. And it’s easy to think this way. If technology reduces our ability to find high-paying jobs, that’s bad, right? If population growth harms the environment, that’s terrible. Isn’t it?

The answer is, “Wait. You’re only looking at one part of the problem.” If the changes come with an increase in resources, either in the form of wealth or other means of coping with change, then those changes may not be as bad as they seem, and in fact may be positively good. Consider the sharp increase in the affordability of a wide variety of products and services made possible by dramatically cheaper labor. Sure, wages may fall, but if prices fall by more, the income effect bails us out. Environmental damage from sooty air was once taken as a sign of prosperity; sure, good air is now more expensive, but the income effect of greater command over goods and services bails us out. In fact, the increase in wealth can provide the means of adapting to the changes in jobs, or the changes in the environment, with enough left over to spend more on going to the opera.

It’s an ancient problem. As Suetonius wrote of the Roman emperor Vespasian:

He was the first to establish a regular salary of a hundred thousand sesterces for Latin and Greek teachers of rhetoric, paid from the privy purse. He also presented eminent poets with princely largess and great rewards, and artists, too, such as the restorer of the Venus of Cos and of the Colossus. To a mechanical engineer, who promised to transport some heavy columns to the Capitol at small expense, he gave no mean reward for his invention, but refused to make use of it, saying: “You must let me feed my poor commons.”

That’s pure substitution effect, there. It’s true enough that the effect of the invention would have been to “price” some labor out of the market, but that would have resulted in freeing up that labor to do other things, thereby increasing the wealth of the society.

Vespasian was wrong. Prosperity is not “good jobs.” Wealth is widely shared access to useful products and services, and suppressing innovation is not the way to get there. Interestingly, one of the people who recognized this aspect of “technological unemployment” was J.M. Keynes. In a 1930 essay, he argued:

We are being afflicted with a new disease of which some readers may not yet have heard the name, but of which they will hear a great deal in the years to come—namely, technological unemployment. This means unemployment due to our discovery of means of economising the use of labour outrunning the pace at which we can find new uses for labour.

But this is only a temporary phase of maladjustment. All this means in the long run that mankind is solving its economic problem… Thus for the first time since his creation man will be faced with his real, his permanent problem—how to use his freedom from pressing economic cares, how to occupy the leisure, which science and compound interest will have won for him, to live wisely and agreeably and well…

The most interesting prediction Keynes made was about the shape of the work life of the future. He combines two insights, neither of which are yet visible, at least not to the extent Keynes expected. The first was the elimination of scarcity on a wide scale; the second is the reaction of workers in seeking to “buy” more leisure with the increased income they receive from working. He did recognize that process might be slow, slow enough that it might be happening before we fully recognize it.

The key elements of Keynes’ argument must be that either (1) labor displaced by technology in one area, such as agriculture or manufacturing, will find well-paid applications in other sectors, or (2) people will simply work less, and substitute paid work for leisure, or construct communities of meaning around voluntary group activities. And that’s a useful way of briefly summarizing the argument in C. B. Frey’s timely book, The Technology Trap. Frey argues that there is no evidence of effect #2. In fact, the work hours of the most highly paid members of society are going up, not down. And the evidence on effect #1 is even less promising, with wages rapidly declining or jobs simply disappearing in sector after sector. It’s not just that Keynes was wrong, but that we are on the verge of a job crisis, according to Frey.

His core claim is that the nature of the technology now being deployed is different from analogous technological changes in the past. For many centuries, improved “capital” mostly meant better tools, which improved productivity of people such as blacksmiths or shoemakers, increasing their wealth. The Industrial Revolution, on the other hand, because of its economies of scale, created dramatically increased inequality. My own go-to observation on this difference comes from Upton Sinclair, who said: “Private ownership of tools—a basis of freedom when tools are simple—becomes a basis of enslavement when tools are complex.”

Except that, after a brief transition period that Karl Marx (and Upton Sinclair) thought portended the end times, the use of capital for mass production actually increased the wealth of even the poorest citizens. Think Adam Smith’s “woolen coat” logic, but applied to every aspect of our lives, ranging from clothing to food to transportation to entertainment. The displacement of labor by machines made people much better off, and created a wealth of opportunities for new and even better-paying uses of labor. The advantage of mass-production, as Frey sees it, is that it creates many niches for workers, unskilled and uneducated in many cases, to occupy positions in factories and other enterprises. So far so good.

But now, according to Frey, we are in an historically unprecedented “this time is different” setting. Once again, technological changes are immanent, and the consequent increases in inequality are rampant. But this time there is no hope of the ensuing benefits of mass-producing for mass-producing new jobs for the lower classes. Where the 19th century replaced people with machines but then required average people to run the machines, the new wave of innovation displaces a large number of unskilled workers and requires only a few highly skilled workers to replace them.

The difficulty, as I argued in my own recent book Tomorrow 3.0, is an increasingly problematic analogy: software and highly portable, connected “apps” are having the same effect on “service” jobs that robots, automation, and power tools had on manufacturing and agriculture. Frey’s key distinction, or so he thinks, is between technology that augments labor, making it more productive, and technology that substitutes for labor, making labor obsolete or redundant. If I have power tools, such as an electric drill and circular saw, I can be a much more productive (and therefore more highly paid) carpenter. If I am a tax accountant, or pharmacist, however, my entire job can be done more cheaply, and possibly better, by a software-driven expert system.

That distinction is not nonsense, but it’s not as meaningful as Frey appears to believe. Skill with power tools makes some workers more employable, and availability of software to do the entire job makes those same workers less employable, it’s true. But consider the example of ATM’s or automatic teller machines. Frey rightly notes:

As is evident by the existence of ATMs, we can easily write a set of rules that allows computers to substitute for bank tellers in accepting deposits and paying out withdrawals. Yet we struggle to define the rules for dealing with an unsatisfied customer. Naturally, banks have taken advantage of this… [A]s the handling of money has been automated, tellers have taken on nonroutine functions.

This would be evidence for Frey’s position—automation is reducing employment—if the result had been a sharp and persistent decline in bank teller employment. But the opposite is true: as the number of ATMs has exploded, banks have been able to expand the number of local branches because the costs of doing so are reduced by being able to automate the routine functions. As Eric Schmidt, head of Alphabet, said at a Columbia University forum in 2017:

There are more bank tellers now than ever because banks are more efficient… You’d have to convince yourself that a declining workforce and an ever-increasing idle force, the sum of that won’t generate more demand… That’s never been true.

One must be careful, of course. It is not literally true that the lower cost of ATM technology has directly caused an increase in bank teller employment. What’s true is that overall adjustments, including the increase in wealth caused by automation generally, have compensated for the narrowly defined displacement of workers in banking. And that’s why the Technology Trap argument is wrong, or at least not entirely right: Say’s Law is always operating in the background. The availability of inexpensive, educated, and productive labor is being increased by technological advances. This kind of glittering and highly profitable opportunity is available for entrepreneurs to take advantage of. It’s true that Frey cannot imagine what form these new developments will take—and neither can I. But the categorical claim, made now by many alarmists, that “this time is different; jobs are gone forever” is no more plausible now than when almost the same claims, and for the same reason, were made in 1848, in 1938, or in 1998.

The fact that there has been rapid adjustment in a few industries—everyone on my “side” cites the ATM example—doesn’t mean that there is no problem. The question is whether the problem is one of adjustment, with a period of sharp disruption followed by a new and hard-to-predict prosperity, or if it this time things really are different. Further, even a relatively “short-term” period of adjustment is a misleading way to characterize what will surely be a decade or more of substantial disturbance in labor markets. If you graduate from high school, and play by the rules, our system has for decades promised a life of relatively prosperity and a period of comfortable retirement. Frey’s observations, and detailed historical analysis, are useful for even those of us who cling to a more optimistic view in the long run.