Carl Benedikt Frey. The Technology Trap: Capital, Labor, and Power in the Age of Automation. Princeton: Princeton University Press, 2019
Economic historian Carl Frey deepens our understanding of the current technology revolution by comparing it to previous technology-driven transitions. Although the author is perhaps best known for his estimate that 47% of American jobs are vulnerable to automation, his general view is fairly optimistic. He does not doubt the long-run benefits of the Industrial Revolution, which has doubled per-capita income every 50 years since 1750. He expects similar benefits from the new digital technologies, eventually. But he is concerned about the loss of jobs due to automation and the resistance to change it may generate.
Two sides of technology
For many years, the conventional wisdom in economics was that new technologies create as many jobs as they destroy, and do so fairly quickly. More recent thinking distinguishes two different effects of technology, either of which may dominate.
The extent to which labor-saving technologies will cause dislocation depends on whether they are enabling or replacing. Replacing technologies render jobs and skills redundant. Enabling technologies, in contrast, make people more productive in existing tasks or create entirely new jobs for them.
A classic example of a replacing technology is the early industrial power loom, which replaced the hand loom for weaving cloth and put a lot of weavers out of work. An example of an enabling technology is an X-ray machine, which improves a physician’s ability to diagnose disease.
Which side of technology is more prevalent has implications for labor demand, wages, and the share of national income going to labor as opposed to capital.
If technology replaces labor in existing tasks, wages and the share of national income accruing to labor may fall. If, in contrast, technological change is augmenting labor, it will make workers more productive in existing tasks or create entirely new labor-intensive activities, thereby increasing the demand for labor.
Just because a technology is available does not mean that people will want to adopt it. That depends on how they expect it to impact their income. And since different groups can be affected differently, technological change depends on who stands to gain or lose and the distribution of power among competing interests.
Technology traps
Societies fall into a technology trap when they are unable to implement a potentially useful technology due to social resistance. Frey uses the term mainly in reference to preindustrial societies, although he fears that we could fall into a similar trap today. “One reason economic growth was stagnant for millennia is that the world was caught in a technology trap, in which labor-replacing technology was consistently and vigorously resisted for fear of its destabilizing force.”
During the period that he calls “The Great Stagnation,” the problem was not so much that innovations didn’t appear, but that people lacked the incentive to implement them, especially for purposes of saving labor in economic production. He characterizes the Renaissance as both a cultural movement and “a force of profound technological change,” but a period with “plenty of imagination, but little realization.” One reason why industrialization didn’t occur earlier than it did was that landed elites were living comfortably off cheap labor they controlled, and had no interest in seeing them go to town to work in a factory. In the late eighteenth century, 96% of the world’s population were slaves, serfs, servants, or vassals. Many monarchs also preferred the status quo to the uncertainties of a social upheaval.
The fear among the ruling classes that labor displacement would cause hardship, social unrest, and at worst a challenge to the political status quo meant that worker-replacing technologies frequently were resisted or even banned. This dynamic, in which the politically powerful had more to lose than they could gain from progress, kept the Western world in a technology trap where technologies that threatened people’s skills were forcefully resisted.
Who gained from industrialization?
When the Industrial Revolution did begin in Britain, it was in someone’s interest to make it happen.
The hegemony of landed wealth was challenged by the mobile fortunes of merchants, who came to form a new industrial class with growing political influence. The mechanized factory was deemed critical to Britain’s competitive position in trade and thus to merchants’ fortunes, which its government would do nothing to jeopardize.
The position of the merchants was strengthened by profits from the Atlantic trade, which were not as monopolized by royal trading companies as they were in other parts of Europe. Merchants also got some support from a stronger Parliament, after the Civil War and the Glorious Revolution. “In eighteenth-century England, the polity and judiciary, which had previously supported the cause of workers and guilds and opposed replacing technologies, began to side with the innovators.”
At first, industrialization benefited neither the artisans who had been spinning, weaving and sewing in cottage industries, nor the workers in the early textile mills. The main effect of innovations like the spinning jenny and the power loom was to replace skilled artisan labor with something cheaper. The new factory jobs paid less, required less skill, and were done by children about half the time. It took about seventy years, from about 1770 to 1840, before the British working class started to share the benefits from industrialization. But they were helpless to stop the process because of their weak political position. In 1769, Parliament made it a capital crime to destroy machinery, as some protesters had been doing.
The situation during the “Second Industrial Revolution,” when the United States emerged as the leading industrial power, was very different. Here Frey says that the new technologies, especially electric power and the internal combustion engine, “were predominantly of the enabling sort.” They raised worker productivity and created new jobs more than they replaced workers. He cites the work of another economic historian:
Alexander Field has argued that productivity growth in the period 1919–73 can be thought of as “a tale of two transitions.” The first involved the redesign of the factory to take advantage of the virtues of electricity, whereas the second constituted a shift toward the horseless age, as motorized vehicles revolutionized transportation and distribution.
In this case, the benefits did come to be shared with the workers in the form of higher wages, shorter hours, safer workplaces, and earlier retirement. Although the workers’ struggles to organize were often violent, they were focused on winning a better share of the benefits of higher productivity, not on destroying the machines that made it possible.
Technology in social context
One question that I had throughout the book was how economists make the distinction between replacing and enabling in practice. Since Frey uses those terms as adjectives describing technologies, the reader could easily get the impression that the effect of a technology is readily observable as soon as it is introduced. That may be true for some specialized machines, that either clearly do or clearly don’t replace what a worker is currently doing. More generalized technologies with many applications can have mixed effects, replacing some workers while enabling others, as Frey’s discussion of twentieth-century electrical machines makes clear. “Clearly, technology did cause some occupations to vanish–like those of lamplighters, elevator operators, laundresses and so on–yet these jobs employed only a fraction of the workforce relative to the new machine-aided occupations that emerged.”
Another example of mixed effects is the internal combustion engine. On the one hand, it gave the drivers of motor vehicles the power of many horses, enabling them to cover greater distances, move more goods, or plow more fields. It created the occupation of truck driver, which is still the largest single occupation in many states. On the other hand, the tractor and other farm machinery dramatically reduced the demand for farm labor.
Now as Frey points out, many of the laborers who left farm work in the twentieth century did so more voluntarily than the displaced artisans of early industrial times. They chose to leave because they could get higher wages in manufacturing. In that case, the same workers were replaced in the agricultural sector after they were enabled in another. Similarly, workers who left domestic service for manufacturing were replaced by electrical appliances. But that raises the question of why labor demand was so high in manufacturing. Was it simply in the nature of assembly-line technology to enhance rather than replace labor? Wasn’t it also because there was now a mass market for manufactured goods, supported by a system that routinely passed along the benefits of high productivity to consumers (as low prices) and to workers (as high wages). Without expanding markets, wouldn’t assembly-line technology replace many workers and not just empower them?
Another factor affecting whether workers are replaced or enhanced is their skill level. “One reason that the horseless age was not accompanied by a jobless age is that human workers, unlike horses, have the means of acquiring new skills, which allows them to take on tasks outside the realm of machines.” If the effect of a machine depends on what happens outside the realm of machines, then classifying the machine as replacing or enabling is no simple matter.
The conclusion I come to is that whether a technology is replacing or enhancing may not be at all obvious when it is first introduced. It depends on how its applications unfold over the course of many years, in a social context that includes things like corporate policies, markets, labor organization, and access to education.
What seems clear is that very early industrial technologies were more replacing than enabling, and that the potential of technology to empower workers was realized only gradually over the course of industrial history. Harnessing the power of nature with such innovations as the steam engine and electrical machinery had a lot to do with this. But rather than seeing technologies as either inherently replacing or enabling, I would call attention to the continued potential for both, as well as to the many decisions that influenced how technologies were actually used. Was it the assembly line itself that made Henry Ford raise wages and reduce car prices, or was it his vision of a path to a prosperous industry?
A new technology trap?
Frey’s concerns about a new technology trap arise from his observation that new technologies are more like those of the First Industrial Revolution than the Second; they are more replacing than enabling. (But again I ask: Is that a feature of the technology itself or of the social context in which we are using it?)
After reviewing many recent technological developments, including in machine learning, machine vision, sensors, various subfields of AI, and mobile robotics, my conclusion is that while these technologies will spawn new tasks for labor, they are predominantly replacing technologies and will continue to worsen the employment prospects for the already shattered middle class.
Like some other economists, Frey sees a similarity between the plight of workers today and that of early textile workers, whose skills and incomes were more replaced by machinery than enhanced by it. That accounts for a lot of the backlash against automation and global trade, which could impede technological change. “The mere existence of better machines is not sufficient for long run growth.” Growth will depend on “policy choices made in the short run.” At the very least, steps must be taken to ease the transition to a hi-tech society for workers who are experiencing dislocations.
Frey remains a long-run optimist, believing that eventually productivity growth will resume–it’s been sluggish lately despite the new technologies–and that more good jobs will be created involving tasks that are hard to automate. These will usually be the more creative, more skilled areas of human activity. Frey rejects as a “widespread misconception…that automation is coming for the jobs of the skilled.” From my perspective, that means that information technologies may not turn out to be as “predominantly replacing” as it now seems.
I will elaborate on many of these points in upcoming posts.
Beneath the Surface 