Because the thing is, although jobs have continued to proliferate in the long run, these predictions weren’t exactly wrong in the short run: Machines do replace humans. In fact, replacing humans is often entirely the point. As the economic historian Robert C. Allen has shown, the spinning jenny was invented in England precisely because wages were high, and thus it was worth it to mill-owners to invest in a machine that would allow them to reduce the number of workers needed to make yarn. Even once the spinning jenny was invented, mill owners in France and India (where wages were relatively low) at first did not invest in them—they could just have the cheap humans do the work. Meanwhile, many of Thomas Mortimer’s sawyers surely lost their livelihoods. (As Keynes accurately and succinctly put it, “In the long run we are all dead.”)
But there is more to the upheavals than mere job loss, and even workers who managed to remain employed have found themselves affected. As Mokyr, Vickers, and Ziebarth describe, the concerns about the ways that technology was reshaping work were often not so much about the quantity of work available (with shortages leading to unemployment) but about the quality of that work—whether it was safe, whether it afforded workers sufficient autonomy, and whether it enabled them to have good lives. For example, they write of the “great anxiety” that people felt “in moving to factory work,” which for the first time separated workplace from home on a mass scale. Contrast that, they write, with today, when “people worry about the exact opposite phenomenon with the lines between spheres of home and work blurring.”
This history is much more valuable than some pat lesson about the foolhardiness of prognosticators. Even for those who were right, the fact that they saw the future accurately mattered very little, as no one could have known they were right at the time. What matters, then, isn’t whether early observers were right or wrong about the long term, but whether they were sufficiently empathetic in the short term. As the authors write, “While the predictions of widespread technological unemployment were, by and large, wrong, we should not trivialize the costs borne by the many who were actually displaced.”
Today, we are not much better at making reliable long-term predictions about the future of the economy than 18th-century onlookers. But there is plenty we can and do know about how technological change is shaping work today—who gets it, how they are compensated, where the work gets done.
Perhaps the biggest change, one that Mokyr, Vickers, and Ziebarth highlight, is the growth in so-called “nonemployer business,” which sometimes goes by the buzzier (and misleading) term “sharing economy,” and refers, in part, to gigs coordinated online via apps such as Uber, AirBnB, or Amazon’s Mechanical Turk. Other shifts include the greater “flexibility” many employees throughout the economy have with regard to where they work and their hours (which can mean unpredictable schedules based on demand and, as a result, unpredictable wages), and a recent doubling of the percentage of workers who primarily work from home (from 2.3 to 4.3 between 1980 and 2010). Many of these people—untethered from employment in the traditional, legal, W-2 sense—struggle, because the systems in place for protecting workers (unemployment insurance, worker’s comp, minimum wage rules, social-security taxes) don’t apply to them.
In other words, there is plenty to figure out with regard to how technology is reshaping the economy without looking to the far distant future. The question isn’t whether the machines are coming; they are already here, and have been for a long time.