At the founding of the country, 9 out of 10 Americans lived on a farm, but as of 2022, fewer than 2% of Americans made their living farming. Explaining this transition to an intelligent observer in the late 18th century would be difficult. While they might not use the words 'mass unemployment', that would definitely be on their minds. But more than anything, they wouldn't believe that such a small number of people could feed the whole nation. In the 18th century, humanity was just beginning to escape the Malthusian trap, where any agricultural surplus would be consumed away by a growing population that was at the very edge of subsistence. There's a reason the Church of England's 1662 Book of Common Prayer has three separate prayers for rain, fair weather, and plenty; it took nearly all of society to feed all of society, and just barely at that.
The resulting story of economic development, initially in the West before diffusing, is one of technological advances increasing farm productivity. Industrial farming equipment, artificial fertilisers, and pesticides dramatically increased how much a single farmer could produce. A bad wheat harvest today likely won't be noticed by much of American society. But with far fewer farmers needed to feed the nation, why didn't we face mass unemployment?
An excerpt from Brad DeLong's Slouching Towards Utopia gives some clues:
Four percent of Americans had flush toilets at home in 1870; 20 percent had them in 1920, 71 percent in 1950, and 96 percent in 1970. No American had a landline telephone in 1880; 28 percent had one in 1914, 62 percent in 1950, and 87 percent in 1970. Eighteen percent of Americans had electric power in 1913; 94 percent had it by 1950.
In nations that reached escape velocity from the Malthusian trap, expectations rose for living standards. If a nation could routinely produce enough food to sustain itself, it could start to bring more and more consumer goods to the masses. But manufacturing these goods required the human labor feed by newly-mechanised agriculture. Before kinks had been worked out, it was beyond the capability of the late 19th century American economy to provide automobiles, indoor plumbing, home telephones, or electricity to anyone more than a privileged few. But the labor story of mechanised agriculture repeated itself: at first manufacturing was highly labor-intensive, but moving up the learning curve made labor more productive and brought manufactured consumer goods within the reach of more people.
There's no law of nature that requires economic development to take this path. You could certainly set up a society that maintains a 19th century standard of living where people work single-digit hours per week, but very few societies make this choice. Humans seem to have no intrinsic limitations to our wants; our expectations expand to fill our ability to meet them. But the goods and services at the very frontier of what an economy is able to produce will be more labour intensive, so technological frontiers change what the newly labor-intensive industry is.
Many rich countries struggled with the transition to more service dominated economies. International trade complicates this story, but only somewhat. In any country where manufacturing is done, higher manufacturing productivity means fewer manufacturing jobs, but these gains drive higher demand from rich-world consumers for services at the frontier of what the economy can deliver. Put another way, as hard as it would be to explain to someone from the late 18th century how few people work on farms in 2025, it would be even harder to explain to them how almost 11 million Americans were employed in an industry that the federal Bureau of Labor statistics describes as 'Professional, Scientific, and Technical Services'. Or over 23 million in 'Healthcare and Social Assistance; Private', let alone nearly 3 million in 'Information'.
This is ultimately what makes me skeptical that this decade's dizzying advancements in artificial intelligence will result in mass unemployment. It will certainly be an adjustment, and likely a painful one for many industries. But if our past record is anything to go off of, insatiable human wants will adjust for the productivity savings, and we will move on to other labor-intensive desires.
U.S. Department of Agriculture, Economic Research Service. Ag and Food Statistics: Charting the Essentials - Ag and Food Sectors and the Economy U.S. Department of Agriculture, National Agricultural Statistics Service. History of Agricultural Statistics. The Book of Common Prayer 1662: Statutory Services. U.S. Bureau of Labor Statistics. Employment by Major Industry Sector.